MyArxiv
Computation and Language 95
☆ ALTA: Compiler-Based Analysis of Transformers
We propose a new programming language called ALTA and a compiler that can map ALTA programs to Transformer weights. ALTA is inspired by RASP, a language proposed by Weiss et al. (2021), and Tracr (Lindner et al., 2023), a compiler from RASP programs to Transformer weights. ALTA complements and extends this prior work, offering the ability to express loops and to compile programs to Universal Transformers, among other advantages. ALTA allows us to constructively show how Transformers can represent length-invariant algorithms for computing parity and addition, as well as a solution to the SCAN benchmark of compositional generalization tasks, without requiring intermediate scratchpad decoding steps. We also propose tools to analyze cases where the expressibility of an algorithm is established, but end-to-end training on a given training set fails to induce behavior consistent with the desired algorithm. To this end, we explore training from ALTA execution traces as a more fine-grained supervision signal. This enables additional experiments and theoretical analyses relating the learnability of various algorithms to data availability and modeling decisions, such as positional encodings. We make the ALTA framework -- language specification, symbolic interpreter, and weight compiler -- available to the community to enable further applications and insights.
☆ TP-Eval: Tap Multimodal LLMs' Potential in Evaluation by Customizing Prompts
Recently, multimodal large language models (MLLMs) have received much attention for their impressive capabilities. The evaluation of MLLMs is becoming critical to analyzing attributes of MLLMs and providing valuable insights. However, current benchmarks overlook the problem of prompt sensitivity - minor prompt variations may lead to significant performance fluctuations. Thus, inappropriate prompts may obscure the models' capabilities, underestimating the models' performance. Moreover, different models have different preferences for different prompts, and thus, using the same prompt for all models will cause evaluation bias. This paper analyzes this deficiency in existing benchmarks and further introduces a new evaluation framework named TP-Eval, which introduces a prompt customization method to reduce evaluation biases and tap models' potential. TP-Eval will rewrite the original prompts to different customized prompts for different models. In particular, we propose some well-designed modules for prompt customization tailored to the scenario of MLLM evaluation. Extensive experiments demonstrate the effectiveness of our approach to uncovering models' capabilities, and TP-Eval should benefit the community in developing more comprehensive and convincing MLLM evaluation benchmarks.
☆ CLEAR: Character Unlearning in Textual and Visual Modalities
Machine Unlearning (MU) is critical for enhancing privacy and security in deep learning models, particularly in large multimodal language models (MLLMs), by removing specific private or hazardous information. While MU has made significant progress in textual and visual modalities, multimodal unlearning (MMU) remains significantly underexplored, partially due to the absence of a suitable open-source benchmark. To address this, we introduce CLEAR, a new benchmark designed to evaluate MMU methods. CLEAR contains 200 fictitious individuals and 3,700 images linked with corresponding question-answer pairs, enabling a thorough evaluation across modalities. We assess 10 MU methods, adapting them for MMU, and highlight new challenges specific to multimodal forgetting. We also demonstrate that simple $\ell_1$ regularization on LoRA weights significantly mitigates catastrophic forgetting, preserving model performance on retained data. The dataset is available at https://huggingface.co/datasets/therem/CLEAR
☆ LongRAG: A Dual-Perspective Retrieval-Augmented Generation Paradigm for Long-Context Question Answering EMNLP 2024
Long-Context Question Answering (LCQA), a challenging task, aims to reason over long-context documents to yield accurate answers to questions. Existing long-context Large Language Models (LLMs) for LCQA often struggle with the "lost in the middle" issue. Retrieval-Augmented Generation (RAG) mitigates this issue by providing external factual evidence. However, its chunking strategy disrupts the global long-context information, and its low-quality retrieval in long contexts hinders LLMs from identifying effective factual details due to substantial noise. To this end, we propose LongRAG, a general, dual-perspective, and robust LLM-based RAG system paradigm for LCQA to enhance RAG's understanding of complex long-context knowledge (i.e., global information and factual details). We design LongRAG as a plug-and-play paradigm, facilitating adaptation to various domains and LLMs. Extensive experiments on three multi-hop datasets demonstrate that LongRAG significantly outperforms long-context LLMs (up by 6.94%), advanced RAG (up by 6.16%), and Vanilla RAG (up by 17.25%). Furthermore, we conduct quantitative ablation studies and multi-dimensional analyses, highlighting the effectiveness of the system's components and fine-tuning strategies. Data and code are available at https://github.com/QingFei1/LongRAG.
comment: EMNLP 2024 Main
☆ Key Algorithms for Keyphrase Generation: Instruction-Based LLMs for Russian Scientific Keyphrases
Keyphrase selection is a challenging task in natural language processing that has a wide range of applications. Adapting existing supervised and unsupervised solutions for the Russian language faces several limitations due to the rich morphology of Russian and the limited number of training datasets available. Recent studies conducted on English texts show that large language models (LLMs) successfully address the task of generating keyphrases. LLMs allow achieving impressive results without task-specific fine-tuning, using text prompts instead. In this work, we access the performance of prompt-based methods for generating keyphrases for Russian scientific abstracts. First, we compare the performance of zero-shot and few-shot prompt-based methods, fine-tuned models, and unsupervised methods. Then we assess strategies for selecting keyphrase examples in a few-shot setting. We present the outcomes of human evaluation of the generated keyphrases and analyze the strengths and weaknesses of the models through expert assessment. Our results suggest that prompt-based methods can outperform common baselines even using simple text prompts.
comment: The 12th International Conference on Analysis of Images, Social Networks and Texts (AIST'2024)
☆ MiLoRA: Efficient Mixture of Low-Rank Adaptation for Large Language Models Fine-tuning EMNLP 2024
Low-rank adaptation (LoRA) and its mixture-of-experts (MOE) variants are highly effective parameter-efficient fine-tuning (PEFT) methods. However, they introduce significant latency in multi-tenant settings due to the LoRA modules and MOE routers added to multiple linear modules in the Transformer layer. To address this issue, we propose Mixture of Low-Rank Adaptation (MiLoRA), a novel and efficient LoRA variant. MiLoRA differs from previous MOE-style LoRA methods by considering each LoRA module as an expert and employing a prompt-aware routing mechanism. This mechanism calculates expert routing results once before generating the first new token and reuses these results for subsequent tokens, reducing latency. Extensive experiments and analysis on commonsense reasoning tasks, math reasoning tasks, and widely used LLM evaluation benchmarks demonstrate that MiLoRA consistently outperforms strong PEFT baselines with comparable tunable parameter budgets. Additionally, MiLoRA significantly reduces latency in multi-tenant settings compared to previous LoRA-based methods.
comment: Accepted by EMNLP 2024 Findings. arXiv admin note: substantial text overlap with arXiv:2405.18203
☆ GraphTeam: Facilitating Large Language Model-based Graph Analysis via Multi-Agent Collaboration
Graphs are widely used for modeling relational data in real-world scenarios, such as social networks and urban computing. Existing LLM-based graph analysis approaches either integrate graph neural networks (GNNs) for specific machine learning tasks, limiting their transferability, or rely solely on LLMs' internal reasoning ability, resulting in suboptimal performance. To address these limitations, we take advantage of recent advances in LLM-based agents, which have shown capabilities of utilizing external knowledge or tools for problem solving. By simulating human problem-solving strategies such as analogy and collaboration, we propose a multi-agent system based on LLMs named GraphTeam, for graph analysis. GraphTeam consists of five LLM-based agents from three modules, and the agents with different specialities can collaborate with each other to address complex problems. Specifically, (1) input-output normalization module: the question agent extracts and refines four key arguments from the original question, facilitating the problem understanding, and the answer agent organizes the results to meet the output requirement; (2) external knowledge retrieval module: we first build a knowledge base consisting of relevant documentation and experience information, and then the search agent retrieves the most relevant entries for each question. (3) problem-solving module: given the retrieved information from search agent, the coding agent uses established algorithms via programming to generate solutions, and in case the coding agent does not work, the reasoning agent will directly compute the results without programming. Extensive experiments on six graph analysis benchmarks demonstrate that GraphTeam achieves state-of-the-art performance with an average 25.85% improvement over the best baseline in terms of accuracy. The code and data are available at https://github.com/BUPT-GAMMA/GraphTeam.
☆ Cross-lingual Transfer of Reward Models in Multilingual Alignment
Reinforcement learning with human feedback (RLHF) is shown to largely benefit from precise reward models (RMs). However, recent studies in reward modeling schemes are skewed towards English, limiting the applicability of RLHF in multilingual alignments. In this work, we investigate the cross-lingual transfer of RMs trained in diverse languages, primarily from English. Our experimental results demonstrate the strong cross-lingual transfer of English RMs, exceeding target language RMs by 3~4% average increase in Multilingual RewardBench. Furthermore, we analyze the cross-lingual transfer of RMs through the representation shifts. Finally, we perform multilingual alignment to exemplify how cross-lingual transfer in RM propagates to enhanced multilingual instruction-following capability, along with extensive analyses on off-the-shelf RMs. We release the code, model, and data.
☆ Together We Can: Multilingual Automatic Post-Editing for Low-Resource Languages EMNLP 2024
This exploratory study investigates the potential of multilingual Automatic Post-Editing (APE) systems to enhance the quality of machine translations for low-resource Indo-Aryan languages. Focusing on two closely related language pairs, English-Marathi and English-Hindi, we exploit the linguistic similarities to develop a robust multilingual APE model. To facilitate cross-linguistic transfer, we generate synthetic Hindi-Marathi and Marathi-Hindi APE triplets. Additionally, we incorporate a Quality Estimation (QE)-APE multi-task learning framework. While the experimental results underline the complementary nature of APE and QE, we also observe that QE-APE multitask learning facilitates effective domain adaptation. Our experiments demonstrate that the multilingual APE models outperform their corresponding English-Hindi and English-Marathi single-pair models by $2.5$ and $2.39$ TER points, respectively, with further notable improvements over the multilingual APE model observed through multi-task learning ($+1.29$ and $+1.44$ TER points), data augmentation ($+0.53$ and $+0.45$ TER points) and domain adaptation ($+0.35$ and $+0.45$ TER points). We release the synthetic data, code, and models accrued during this study publicly at https://github.com/cfiltnlp/Multilingual-APE.
comment: Accepted at Findings of EMNLP 2024
☆ Dependency Graph Parsing as Sequence Labeling EMNLP-2024
Various linearizations have been proposed to cast syntactic dependency parsing as sequence labeling. However, these approaches do not support more complex graph-based representations, such as semantic dependencies or enhanced universal dependencies, as they cannot handle reentrancy or cycles. By extending them, we define a range of unbounded and bounded linearizations that can be used to cast graph parsing as a tagging task, enlarging the toolbox of problems that can be solved under this paradigm. Experimental results on semantic dependency and enhanced UD parsing show that with a good choice of encoding, sequence-labeling dependency graph parsers combine high efficiency with accuracies close to the state of the art, in spite of their simplicity.
comment: Accepted at EMNLP-2024
☆ A Time-Aware Approach to Early Detection of Anorexia: UNSL at eRisk 2024
The eRisk laboratory aims to address issues related to early risk detection on the Web. In this year's edition, three tasks were proposed, where Task 2 was about early detection of signs of anorexia. Early risk detection is a problem where precision and speed are two crucial objectives. Our research group solved Task 2 by defining a CPI+DMC approach, addressing both objectives independently, and a time-aware approach, where precision and speed are considered a combined single-objective. We implemented the last approach by explicitly integrating time during the learning process, considering the ERDE{\theta} metric as the training objective. It also allowed us to incorporate temporal metrics to validate and select the optimal models. We achieved outstanding results for the ERDE50 metric and ranking-based metrics, demonstrating consistency in solving ERD problems.
comment: In Conference and Labs of the Evaluation Forum (CLEF 2024), Grenoble, France
☆ Zeitenwenden: Detecting changes in the German political discourse
From a monarchy to a democracy, to a dictatorship and back to a democracy -- the German political landscape has been constantly changing ever since the first German national state was formed in 1871. After World War II, the Federal Republic of Germany was formed in 1949. Since then every plenary session of the German Bundestag was logged and even has been digitized over the course of the last few years. We analyze these texts using a time series variant of the topic model LDA to investigate which events had a lasting effect on the political discourse and how the political topics changed over time. This allows us to detect changes in word frequency (and thus key discussion points) in political discourse.
comment: 7 pages, 6 figures
☆ ExpertFlow: Optimized Expert Activation and Token Allocation for Efficient Mixture-of-Experts Inference
Sparse Mixture of Experts (MoE) models, while outperforming dense Large Language Models (LLMs) in terms of performance, face significant deployment challenges during inference due to their high memory demands. Existing offloading techniques, which involve swapping activated and idle experts between the GPU and CPU, often suffer from rigid expert caching mechanisms. These mechanisms fail to adapt to dynamic routing, leading to inefficient cache utilization, or incur prohibitive costs for prediction training. To tackle these inference-specific challenges, we introduce ExpertFlow, a comprehensive system specifically designed to enhance inference efficiency by accommodating flexible routing and enabling efficient expert scheduling between CPU and GPU. This reduces overhead and boosts system performance. Central to our approach is a predictive routing path-based offloading mechanism that utilizes a lightweight predictor to accurately forecast routing paths before computation begins. This proactive strategy allows for real-time error correction in expert caching, significantly increasing cache hit ratios and reducing the frequency of expert transfers, thereby minimizing I/O overhead. Additionally, we implement a dynamic token scheduling strategy that optimizes MoE inference by rearranging input tokens across different batches. This method not only reduces the number of activated experts per batch but also improves computational efficiency. Our extensive experiments demonstrate that ExpertFlow achieves up to 93.72\% GPU memory savings and enhances inference speed by 2 to 10 times compared to baseline methods, highlighting its effectiveness and utility as a robust solution for resource-constrained inference scenarios.
comment: Mixture-of-Experts, Inference, Offloading
☆ SimRAG: Self-Improving Retrieval-Augmented Generation for Adapting Large Language Models to Specialized Domains
Retrieval-augmented generation (RAG) enhances the question-answering (QA) abilities of large language models (LLMs) by integrating external knowledge. However, adapting general-purpose RAG systems to specialized fields such as science and medicine poses unique challenges due to distribution shifts and limited access to domain-specific data. To tackle this, we propose SimRAG, a self-training approach that equips the LLM with joint capabilities of question answering and question generation for domain adaptation. Our method first fine-tunes the LLM on instruction-following, question-answering, and search-related data. Then, it prompts the same LLM to generate diverse domain-relevant questions from unlabeled corpora, with an additional filtering strategy to retain high-quality synthetic examples. By leveraging these synthetic examples, the LLM can improve their performance on domain-specific RAG tasks. Experiments on 11 datasets, spanning two backbone sizes and three domains, demonstrate that SimRAG outperforms baselines by 1.2\%--8.6\%.
comment: Work in Progress
☆ ELAICHI: Enhancing Low-resource TTS by Addressing Infrequent and Low-frequency Character Bigrams
Recent advancements in Text-to-Speech (TTS) technology have led to natural-sounding speech for English, primarily due to the availability of large-scale, high-quality web data. However, many other languages lack access to such resources, relying instead on limited studio-quality data. This scarcity results in synthesized speech that often suffers from intelligibility issues, particularly with low-frequency character bigrams. In this paper, we propose three solutions to address this challenge. First, we leverage high-quality data from linguistically or geographically related languages to improve TTS for the target language. Second, we utilize low-quality Automatic Speech Recognition (ASR) data recorded in non-studio environments, which is refined using denoising and speech enhancement models. Third, we apply knowledge distillation from large-scale models using synthetic data to generate more robust outputs. Our experiments with Hindi demonstrate significant reductions in intelligibility issues, as validated by human evaluators. We propose this methodology as a viable alternative for languages with limited access to high-quality data, enabling them to collectively benefit from shared resources.
comment: 11 pages, 1 figure, 3 tables
☆ Value Residual Learning For Alleviating Attention Concentration In Transformers
Transformers can capture long-range dependencies using self-attention, allowing tokens to attend to all others directly. However, stacking multiple attention layers leads to attention concentration. One natural way to address this issue is to use cross-layer attention, allowing information from earlier layers to be directly accessible to later layers. However, this approach is computationally expensive. To address this problem, we propose Transformer with residual value (ResFormer) which approximates cross-layer attention through adding a residual connection from the values of the the first layer to all subsequent layers. Based on this method, one variant is the Transformer with single layer value (SVFormer), where all layers share the same value embedding from first layer, reducing the KV cache by nearly 50%. Comprehensive empirical evidence demonstrates that ResFormer mitigates attention concentration problem in deeper layers and enhances representation across most layers, outperforming the vanilla Transformer, DenseFormer, and NeuTRENO in training error as well as downstream tasks. SVFormer trains significantly faster than the vanilla Transformer and performs better than other methods like GQA and CLA, with performance influenced by sequence length and cumulative learning rate.
☆ Scaling Diffusion Language Models via Adaptation from Autoregressive Models
Diffusion Language Models (DLMs) have emerged as a promising new paradigm for text generative modeling, potentially addressing limitations of autoregressive (AR) models. However, current DLMs have been studied at a smaller scale compared to their AR counterparts and lack fair comparison on language modeling benchmarks. Additionally, training diffusion models from scratch at scale remains challenging. Given the prevalence of open-source AR language models, we propose adapting these models to build text diffusion models. We demonstrate connections between AR and diffusion modeling objectives and introduce a simple continual pre-training approach for training diffusion models. Through systematic evaluation on language modeling, reasoning, and commonsense benchmarks, we show that we can convert AR models ranging from 127M to 7B parameters (GPT2 and LLaMA) into diffusion models DiffuGPT and DiffuLLaMA, using less than 200B tokens for training. Our experimental results reveal that these models outperform earlier DLMs and are competitive with their AR counterparts. We release a suite of DLMs (with 127M, 355M, and 7B parameters) capable of generating fluent text, performing in-context learning, filling in the middle without prompt re-ordering, and following instructions \url{https://github.com/HKUNLP/DiffuLLaMA}.
comment: 25 pages. Code: https://github.com/HKUNLP/DiffuLLaMA
☆ SpeakGer: A meta-data enriched speech corpus of German state and federal parliaments
The application of natural language processing on political texts as well as speeches has become increasingly relevant in political sciences due to the ability to analyze large text corpora which cannot be read by a single person. But such text corpora often lack critical meta information, detailing for instance the party, age or constituency of the speaker, that can be used to provide an analysis tailored to more fine-grained research questions. To enable researchers to answer such questions with quantitative approaches such as natural language processing, we provide the SpeakGer data set, consisting of German parliament debates from all 16 federal states of Germany as well as the German Bundestag from 1947-2023, split into a total of 10,806,105 speeches. This data set includes rich meta data in form of information on both reactions from the audience towards the speech as well as information about the speaker's party, their age, their constituency and their party's political alignment, which enables a deeper analysis. We further provide three exploratory analyses, detailing topic shares of different parties throughout time, a descriptive analysis of the development of the age of an average speaker as well as a sentiment analysis of speeches of different parties with regards to the COVID-19 pandemic.
comment: 10 pages, 3 figures
☆ Understanding Layer Significance in LLM Alignment
Aligning large language models (LLMs) through fine-tuning is essential for tailoring them to specific applications. Therefore, understanding what LLMs learn during the alignment process is crucial. Recent studies suggest that alignment primarily adjusts a model's presentation style rather than its foundational knowledge, indicating that only certain components of the model are significantly impacted. To delve deeper into LLM alignment, we propose to identify which layers within LLMs are most critical to the alignment process, thereby uncovering how alignment influences model behavior at a granular level. We propose a novel approach to identify the important layers for LLM alignment (ILA). It involves learning a binary mask for each incremental weight matrix in the LoRA algorithm, indicating the significance of each layer. ILA consistently identifies important layers across various alignment datasets, with nearly 90% overlap even with substantial dataset differences, highlighting fundamental patterns in LLM alignment. Experimental results indicate that freezing non-essential layers improves overall model performance, while selectively tuning the most critical layers significantly enhances fine-tuning efficiency with minimal performance loss.
☆ Understanding When Tree of Thoughts Succeeds: Larger Models Excel in Generation, Not Discrimination
Tree of Thoughts (ToT) is a reasoning strategy for Large Language Models (LLMs) that employs a generator to suggest reasoning steps and a discriminator to decide which steps to implement. ToT demonstrates strong performance on reasoning tasks, often surpassing simple methods such as Input-Output (IO) prompting and Chain-of-Thought (CoT) reasoning. However, ToT does not consistently outperform such simpler methods across all models, leaving large knowledge gaps on the conditions under which ToT is most beneficial. In this paper, we analyze the roles of the generator and discriminator separately to better understand the conditions when ToT is beneficial. We find that the generator plays a more critical role than the discriminator in driving the success of ToT. While using even a smaller model as the discriminator, scaling the generator leads to notable improvements in ToT performance, whereas scaling the discriminator with a fixed generator yields only marginal gains. Our results show that models across different scales exhibit comparable discrimination capabilities, yet differ significantly in their generative performance for ToT.
comment: Code: github.com/mainlp/tot-eval
☆ OmniFlatten: An End-to-end GPT Model for Seamless Voice Conversation
Full-duplex spoken dialogue systems significantly advance over traditional turn-based dialogue systems, as they allow simultaneous bidirectional communication, closely mirroring human-human interactions. However, achieving low latency and natural interactions in full-duplex dialogue systems remains a significant challenge, especially considering human conversation dynamics such as interruptions, backchannels, and overlapping speech. In this paper, we introduce a novel End-to-End GPT-based model OmniFlatten for full-duplex conversation, capable of effectively modeling the complex behaviors inherent to natural conversations with low latency. To achieve full-duplex communication capabilities, we propose a multi-stage post-training scheme that progressively adapts a text-based large language model (LLM) backbone into a speech-text dialogue LLM, capable of generating text and speech in real time, without modifying the architecture of the backbone LLM. The training process comprises three stages: modality alignment, half-duplex dialogue learning, and full-duplex dialogue learning. Throughout all training stages, we standardize the data using a flattening operation, which allows us to unify the training methods and the model architecture across different modalities and tasks. Our approach offers a straightforward modeling technique and a promising research direction for developing efficient and natural end-to-end full-duplex spoken dialogue systems. Audio samples of dialogues generated by OmniFlatten can be found at this web site (https://omniflatten.github.io/).
comment: Work in progress
☆ Leveraging the Domain Adaptation of Retrieval Augmented Generation Models for Question Answering and Reducing Hallucination
While ongoing advancements in Large Language Models have demonstrated remarkable success across various NLP tasks, Retrieval Augmented Generation Model stands out to be highly effective on downstream applications like Question Answering. Recently, RAG-end2end model further optimized the architecture and achieved notable performance improvements on domain adaptation. However, the effectiveness of these RAG-based architectures remains relatively unexplored when fine-tuned on specialized domains such as customer service for building a reliable conversational AI system. Furthermore, a critical challenge persists in reducing the occurrence of hallucinations while maintaining high domain-specific accuracy. In this paper, we investigated the performance of diverse RAG and RAG-like architectures through domain adaptation and evaluated their ability to generate accurate and relevant response grounded in the contextual knowledge base. To facilitate the evaluation of the models, we constructed a novel dataset HotelConvQA, sourced from wide range of hotel-related conversations and fine-tuned all the models on our domain specific dataset. We also addressed a critical research gap on determining the impact of domain adaptation on reducing hallucinations across different RAG architectures, an aspect that was not properly measured in prior work. Our evaluation shows positive results in all metrics by employing domain adaptation, demonstrating strong performance on QA tasks and providing insights into their efficacy in reducing hallucinations. Our findings clearly indicate that domain adaptation not only enhances the models' performance on QA tasks but also significantly reduces hallucination across all evaluated RAG architectures.
comment: Initial Version fine-tuned on HotelConvQA
☆ Latent Structures of Intertextuality in French Fiction
Intertextuality is a key concept in literary theory that challenges traditional notions of text, signification or authorship. It views texts as part of a vast intertextual network that is constantly evolving and being reconfigured. This paper argues that the field of computational literary studies is the ideal place to conduct a study of intertextuality since we have now the ability to systematically compare texts with each others. Specifically, we present a work on a corpus of more than 12.000 French fictions from the 18th, 19th and early 20th century. We focus on evaluating the underlying roles of two literary notions, sub-genres and the literary canon in the framing of textuality. The article attempts to operationalize intertextuality using state-of-the-art contextual language models to encode novels and capture features that go beyond simple lexical or thematic approaches. Previous research (Hughes, 2012) supports the existence of a literary "style of a time", and our findings further reinforce this concept. Our findings also suggest that both subgenres and canonicity play a significant role in shaping textual similarities within French fiction. These discoveries point to the importance of considering genre and canon as dynamic forces that influence the evolution and intertextual connections of literary works within specific historical contexts.
comment: 13 pages, 6 figures. Computational Humanities Research Conference 2024
☆ Local Contrastive Editing of Gender Stereotypes EMNLP 2024
Stereotypical bias encoded in language models (LMs) poses a threat to safe language technology, yet our understanding of how bias manifests in the parameters of LMs remains incomplete. We introduce local contrastive editing that enables the localization and editing of a subset of weights in a target model in relation to a reference model. We deploy this approach to identify and modify subsets of weights that are associated with gender stereotypes in LMs. Through a series of experiments, we demonstrate that local contrastive editing can precisely localize and control a small subset (< 0.5%) of weights that encode gender bias. Our work (i) advances our understanding of how stereotypical biases can manifest in the parameter space of LMs and (ii) opens up new avenues for developing parameter-efficient strategies for controlling model properties in a contrastive manner.
comment: Accepted at EMNLP 2024
☆ MojoBench: Language Modeling and Benchmarks for Mojo
The recently introduced Mojo programming language (PL) by Modular, has received significant attention in the scientific community due to its claimed significant speed boost over Python. Despite advancements in code Large Language Models (LLMs) across various PLs, Mojo remains unexplored in this context. To address this gap, we introduce MojoBench, the first framework for Mojo code generation. MojoBench includes HumanEval-Mojo, a benchmark dataset designed for evaluating code LLMs on Mojo, and Mojo-Coder, the first LLM pretrained and finetuned for Mojo code generation, which supports instructions in 5 natural languages (NLs). Our results show that Mojo-Coder achieves a 30-35% performance improvement over leading models like GPT-4o and Claude-3.5-Sonnet. Furthermore, we provide insights into LLM behavior with underrepresented and unseen PLs, offering potential strategies for enhancing model adaptability. MojoBench contributes to our understanding of LLM capabilities and limitations in emerging programming paradigms fostering more robust code generation systems.
☆ Dialectal and Low Resource Machine Translation for Aromanian COLING 2025
We present a neural machine translation system that can translate between Romanian, English, and Aromanian (an endangered Eastern Romance language); the first of its kind. BLEU scores range from 17 to 32 depending on the direction and genre of the text. Alongside, we release the biggest known Aromanian-Romanian bilingual corpus, consisting of 79k cleaned sentence pairs. Additional tools such as an agnostic sentence embedder (used for both text mining and automatic evaluation) and a diacritics converter are also presented. We publicly release our findings and models. Finally, we describe the deployment of our quantized model at https://arotranslate.com.
comment: 16 pages, 3 figures, 6 tables, submitted to COLING 2025
☆ CogSteer: Cognition-Inspired Selective Layer Intervention for Efficient Semantic Steering in Large Language Models
Despite their impressive capabilities, large language models (LLMs) often lack interpretability and can generate toxic content. While using LLMs as foundation models and applying semantic steering methods are widely practiced, we believe that efficient methods should be based on a thorough understanding of LLM behavior. To this end, we propose using eye movement measures to interpret LLM behavior across layers. We find that LLMs exhibit patterns similar to human gaze across layers and different layers function differently. Inspired by these findings, we introduce a heuristic steering layer selection and apply it to layer intervention methods via fine-tuning and inference. Using language toxification and detoxification as test beds, we demonstrate that our proposed CogSteer methods achieve better results in terms of toxicity scores while efficiently saving 97% of the computational resources and 60% of the training time. Our model-agnostic approach can be adopted into various LLMs, contributing to their interpretability and promoting trustworthiness for safe deployment.
☆ Beware of Calibration Data for Pruning Large Language Models
As large language models (LLMs) are widely applied across various fields, model compression has become increasingly crucial for reducing costs and improving inference efficiency. Post-training pruning is a promising method that does not require resource-intensive iterative training and only needs a small amount of calibration data to assess the importance of parameters. Previous research has primarily focused on designing advanced pruning methods, while different calibration data's impact on pruning performance still lacks systematical exploration. We fill this blank and surprisingly observe that the effects of calibration data even value more than designing advanced pruning strategies, especially for high sparsity. Our preliminary exploration also discloses that using calibration data similar to the training data can yield better performance. As pre-training data is usually inaccessible for advanced LLMs, we further provide a self-generating calibration data synthesis strategy to construct feasible calibration data. We conduct experiments on the recent strong open-source LLMs (e.g., DCLM, and LLaMA-3), and the results show that the proposed method outperforms commonly used calibration data and can effectively enhance strong pruning methods (e.g., Wanda, OWL).
comment: under review
☆ An Adaptive Framework for Generating Systematic Explanatory Answer in Online Q&A Platforms
Question Answering (QA) systems face challenges in handling complex questions that require multi-domain knowledge synthesis. The naive RAG models, although effective in information retrieval, struggle with complex questions that require comprehensive and in-depth answers. The pioneering task is defined as explanatory answer generation, which entails handling identified challenges such as the requirement for comprehensive information and logical coherence within the generated context. To address these issues, we refer to systematic thinking theory and propose SynthRAG, an innovative framework designed to enhance QA performance. SynthRAG improves on conventional models by employing adaptive outlines for dynamic content structuring, generating systematic information to ensure detailed coverage, and producing customized answers tailored to specific user inquiries. This structured approach guarantees logical coherence and thorough integration of information, yielding responses that are both insightful and methodically organized. Empirical evaluations underscore SynthRAG's effectiveness, demonstrating its superiority in handling complex questions, overcoming the limitations of naive RAG models, and significantly improving answer quality and depth. Furthermore, an online deployment on the Zhihu platform revealed that SynthRAG's answers achieved notable user engagement, with each response averaging 5.73 upvotes and surpassing the performance of 79.8% of human contributors, highlighting the practical relevance and impact of the proposed framework. Our code is available at https://github.com/czy1999/SynthRAG .
comment: 10 pages, 6 figures
☆ Towards a Similarity-adjusted Surprisal Theory EMNLP 2024
Surprisal theory posits that the cognitive effort required to comprehend a word is determined by its contextual predictability, quantified as surprisal. Traditionally, surprisal theory treats words as distinct entities, overlooking any potential similarity between them. Giulianelli et al. (2023) address this limitation by introducing information value, a measure of predictability designed to account for similarities between communicative units. Our work leverages Ricotta and Szeidl's (2006) diversity index to extend surprisal into a metric that we term similarity-adjusted surprisal, exposing a mathematical relationship between surprisal and information value. Similarity-adjusted surprisal aligns with information value when considering graded similarities and reduces to standard surprisal when words are treated as distinct. Experimental results with reading time data indicate that similarity-adjusted surprisal adds predictive power beyond standard surprisal for certain datasets, suggesting it serves as a complementary measure of comprehension effort.
comment: EMNLP 2024 main conference proceedings
☆ Quantifying the Risks of Tool-assisted Rephrasing to Linguistic Diversity
Writing assistants and large language models see widespread use in the creation of text content. While their effectiveness for individual users has been evaluated in the literature, little is known about their proclivity to change language or reduce its richness when adopted by a large user base. In this paper, we take a first step towards quantifying this risk by measuring the semantic and vocabulary change enacted by the use of rephrasing tools on a multi-domain corpus of human-generated text.
☆ ReflecTool: Towards Reflection-Aware Tool-Augmented Clinical Agents
Large Language Models (LLMs) have shown promising potential in the medical domain, assisting with tasks like clinical note generation and patient communication. However, current LLMs are limited to text-based communication, hindering their ability to interact with diverse forms of information in clinical environments. Despite clinical agents succeeding in diverse signal interaction, they are oriented to a single clinical scenario and hence fail for broader applications. To evaluate clinical agents holistically, we propose ClinicalAgent Bench~(CAB), a comprehensive medical agent benchmark consisting of 18 tasks across five key realistic clinical dimensions. Building on this, we introduce ReflecTool, a novel framework that excels at utilizing domain-specific tools within two stages. The first optimization stage progressively enlarges a long-term memory by saving successful solving processes and tool-wise experience of agents in a tiny pre-defined training set. In the following inference stage, ReflecTool can search for supportive successful demonstrations from already built long-term memory to guide the tool selection strategy, and a verifier improves the tool usage according to the tool-wise experience with two verification methods--iterative refinement and candidate selection. Extensive experiments on ClinicalAgent Benchmark demonstrate that ReflecTool surpasses the pure LLMs with more than 10 points and the well-established agent-based methods with 3 points, highlighting its adaptability and effectiveness in solving complex clinical tasks.
comment: 20 pages
☆ Markov Chain of Thought for Efficient Mathematical Reasoning
Chain of Thought (CoT) of multi-step benefits from the logical structure of the reasoning steps and task-specific actions, significantly enhancing the mathematical reasoning capabilities of large language models. As the prevalence of long CoT, the number of reasoning steps exceeds manageable token limits and leads to higher computational demands. Inspired by the fundamental logic of human cognition, ``derive, then reduce'', we conceptualize the standard multi-step CoT as a novel Markov Chain of Thought (MCoT). In this study, we consider the mathematical reasoning task, defining each reasoning step as text accompanied by a Python code snippet. To facilitate a longer reasoning path, self-correction is enabled through interactions with the code interpreter. Our MCoT aims to compress previous reasoning steps into a simplified question, enabling efficient next-step inference without relying on a lengthy KV cache. In our experiments, we curate the \texttt{MCoTInstruct} dataset, and the empirical results indicate that MCoT not only significantly enhances efficiency but also maintains comparable accuracy. While much remains to be explored, this work paves the way for exploring the long CoT reasoning abilities of LLMs.
comment: Work in progress
☆ LMLPA: Language Model Linguistic Personality Assessment
Large Language Models (LLMs) are increasingly used in everyday life and research. One of the most common use cases is conversational interactions, enabled by the language generation capabilities of LLMs. Just as between two humans, a conversation between an LLM-powered entity and a human depends on the personality of the conversants. However, measuring the personality of a given LLM is currently a challenge. This paper introduces the Language Model Linguistic Personality Assessment (LMLPA), a system designed to evaluate the linguistic personalities of LLMs. Our system helps to understand LLMs' language generation capabilities by quantitatively assessing the distinct personality traits reflected in their linguistic outputs. Unlike traditional human-centric psychometrics, the LMLPA adapts a personality assessment questionnaire, specifically the Big Five Inventory, to align with the operational capabilities of LLMs, and also incorporates the findings from previous language-based personality measurement literature. To mitigate sensitivity to the order of options, our questionnaire is designed to be open-ended, resulting in textual answers. Thus, the AI rater is needed to transform ambiguous personality information from text responses into clear numerical indicators of personality traits. Utilising Principal Component Analysis and reliability validations, our findings demonstrate that LLMs possess distinct personality traits that can be effectively quantified by the LMLPA. This research contributes to Human-Computer Interaction and Human-Centered AI, providing a robust framework for future studies to refine AI personality assessments and expand their applications in multiple areas, including education and manufacturing.
☆ Graphusion: A RAG Framework for Knowledge Graph Construction with a Global Perspective
Knowledge Graphs (KGs) are crucial in the field of artificial intelligence and are widely used in downstream tasks, such as question-answering (QA). The construction of KGs typically requires significant effort from domain experts. Large Language Models (LLMs) have recently been used for Knowledge Graph Construction (KGC). However, most existing approaches focus on a local perspective, extracting knowledge triplets from individual sentences or documents, missing a fusion process to combine the knowledge in a global KG. This work introduces Graphusion, a zero-shot KGC framework from free text. It contains three steps: in Step 1, we extract a list of seed entities using topic modeling to guide the final KG includes the most relevant entities; in Step 2, we conduct candidate triplet extraction using LLMs; in Step 3, we design the novel fusion module that provides a global view of the extracted knowledge, incorporating entity merging, conflict resolution, and novel triplet discovery. Results show that Graphusion achieves scores of 2.92 and 2.37 out of 3 for entity extraction and relation recognition, respectively. Moreover, we showcase how Graphusion could be applied to the Natural Language Processing (NLP) domain and validate it in an educational scenario. Specifically, we introduce TutorQA, a new expert-verified benchmark for QA, comprising six tasks and a total of 1,200 QA pairs. Using the Graphusion-constructed KG, we achieve a significant improvement on the benchmark, for example, a 9.2% accuracy improvement on sub-graph completion.
comment: arXiv admin note: substantial text overlap with arXiv:2407.10794
☆ Cross-model Control: Improving Multiple Large Language Models in One-time Training NeurIPS 2024
The number of large language models (LLMs) with varying parameter scales and vocabularies is increasing. While they deliver powerful performance, they also face a set of common optimization needs to meet specific requirements or standards, such as instruction following or avoiding the output of sensitive information from the real world. However, how to reuse the fine-tuning outcomes of one model to other models to reduce training costs remains a challenge. To bridge this gap, we introduce Cross-model Control (CMC), a method that improves multiple LLMs in one-time training with a portable tiny language model. Specifically, we have observed that the logit shift before and after fine-tuning is remarkably similar across different models. Based on this insight, we incorporate a tiny language model with a minimal number of parameters. By training alongside a frozen template LLM, the tiny model gains the capability to alter the logits output by the LLMs. To make this tiny language model applicable to models with different vocabularies, we propose a novel token mapping strategy named PM-MinED. We have conducted extensive experiments on instruction tuning and unlearning tasks, demonstrating the effectiveness of CMC. Our code is available at https://github.com/wujwyi/CMC.
comment: Accepted by NeurIPS 2024
☆ MM-Eval: A Multilingual Meta-Evaluation Benchmark for LLM-as-a-Judge and Reward Models
Large language models (LLMs) are commonly used as evaluators in tasks (e.g., reward modeling, LLM-as-a-judge), where they act as proxies for human preferences or judgments. This leads to the need for meta-evaluation: evaluating the credibility of LLMs as evaluators. However, existing benchmarks primarily focus on English, offering limited insight into LLMs' effectiveness as evaluators in non-English contexts. To address this, we introduce MM-Eval, a multilingual meta-evaluation benchmark that covers 18 languages across six categories. MM-Eval evaluates various dimensions, including language-specific challenges like linguistics and language hallucinations. Evaluation results show that both proprietary and open-source language models have considerable room for improvement. Further analysis reveals a tendency for these models to assign middle-ground scores to low-resource languages. We publicly release our benchmark and code.
comment: work in progress
☆ Differentially Private Learning Needs Better Model Initialization and Self-Distillation
Differentially private SGD (DPSGD) enables privacy-preserving training of language models, but often reduces utility, diversity, and linguistic quality. We introduce DPRefine, a three-phase method that initializes a model using data synthesis from a small pre-trained LM with rigorous filtering, applies DP finetuning on private data, and performs self-distillation to refine outputs. This approach significantly outperforms vanilla DPSGD, with AlpacaEval preferring DPRefine's generations in 78.4% of cases across all datasets. Our analysis reveals that DPRefine reduces linguistic errors in generated text by 84.0%, mitigating grammar and spelling errors, commonly associated with DPSGD. It also reduces inconsistencies of non-private models, such as hallucinated details and misattributed quotes. We find that small models like GPT-2 can be effective for initialization and distillation, highlighting their potential in enabling scalable and efficient deployment of privacy-preserving language.
comment: 18 pages
☆ ESpeW: Robust Copyright Protection for LLM-based EaaS via Embedding-Specific Watermark
Embeddings as a Service (EaaS) is emerging as a crucial role in AI applications. Unfortunately, EaaS is vulnerable to model extraction attacks, highlighting the urgent need for copyright protection.Although some preliminary works propose applying embedding watermarks to protect EaaS, recent research reveals that these watermarks can be easily removed. Hence, it is crucial to inject robust watermarks resistant to watermark removal attacks.Existing watermarking methods typically inject a target embedding into embeddings through linear interpolation when the text contains triggers. However, this mechanism results in each watermarked embedding having the same component, which makes the watermark easy to identify and eliminate.Motivated by this, in this paper, we propose a novel embedding-specific watermarking (ESpeW) mechanism to offer robust copyright protection for EaaS. Our approach involves injecting unique, yet readily identifiable watermarks into each embedding. Watermarks inserted by ESpeW are designed to maintain a significant distance from one another and to avoid sharing common components, thus making it significantly more challenging to remove the watermarks.Extensive experiments on four popular datasets demonstrate that ESpeW can even watermark successfully against a highly aggressive removal strategy without sacrificing the quality of embeddings.
☆ ProtoLens: Advancing Prototype Learning for Fine-Grained Interpretability in Text Classification
Deep neural networks have achieved remarkable performance in various text-based tasks but often lack interpretability, making them less suitable for applications where transparency is critical. To address this, we propose ProtoLens, a novel prototype-based model that provides fine-grained, sub-sentence level interpretability for text classification. ProtoLens uses a Prototype-aware Span Extraction module to identify relevant text spans associated with learned prototypes and a Prototype Alignment mechanism to ensure prototypes are semantically meaningful throughout training. By aligning the prototype embeddings with human-understandable examples, ProtoLens provides interpretable predictions while maintaining competitive accuracy. Extensive experiments demonstrate that ProtoLens outperforms both prototype-based and non-interpretable baselines on multiple text classification benchmarks. Code and data are available at \url{https://anonymous.4open.science/r/ProtoLens-CE0B/}.
☆ Responsible Multilingual Large Language Models: A Survey of Development, Applications, and Societal Impact
Multilingual Large Language Models (MLLMs) represent a pivotal advancement in democratizing artificial intelligence across linguistic boundaries. While theoretical foundations are well-established, practical implementation guidelines remain scattered. This work bridges this gap by providing a comprehensive end-to-end framework for developing and deploying MLLMs in production environments. We make three distinctive contributions: First, we present an actionable pipeline from data pre-processing through deployment, integrating insights from academic research and industrial applications. Second, using Llama2 as a case study, we provide detailed optimization strategies for enhancing multilingual capabilities, including curriculum learning approaches for balancing high-resource and low-resource languages, tokenization strategies, and effective sampling methods. Third, we offer an interdisciplinary analysis that considers technical, linguistic, and cultural perspectives in MLLM development. Our findings reveal critical challenges in supporting linguistic diversity, with 88.38% of world languages categorized as low-resource, affecting over a billion speakers. We examine practical solutions through real-world applications in customer service, search engines, and machine translation. By synthesizing theoretical frameworks with production-ready implementation strategies, this survey provides essential guidance for practitioners and researchers working to develop more inclusive and effective multilingual AI systems.
☆ Navigate Complex Physical Worlds via Geometrically Constrained LLM
This study investigates the potential of Large Language Models (LLMs) for reconstructing and constructing the physical world solely based on textual knowledge. It explores the impact of model performance on spatial understanding abilities. To enhance the comprehension of geometric and spatial relationships in the complex physical world, the study introduces a set of geometric conventions and develops a workflow based on multi-layer graphs and multi-agent system frameworks. It examines how LLMs achieve multi-step and multi-objective geometric inference in a spatial environment using multi-layer graphs under unified geometric conventions. Additionally, the study employs a genetic algorithm, inspired by large-scale model knowledge, to solve geometric constraint problems. In summary, this work innovatively explores the feasibility of using text-based LLMs as physical world builders and designs a workflow to enhance their capabilities.
☆ MobileSafetyBench: Evaluating Safety of Autonomous Agents in Mobile Device Control
Autonomous agents powered by large language models (LLMs) show promising potential in assistive tasks across various domains, including mobile device control. As these agents interact directly with personal information and device settings, ensuring their safe and reliable behavior is crucial to prevent undesirable outcomes. However, no benchmark exists for standardized evaluation of the safety of mobile device-control agents. In this work, we introduce MobileSafetyBench, a benchmark designed to evaluate the safety of device-control agents within a realistic mobile environment based on Android emulators. We develop a diverse set of tasks involving interactions with various mobile applications, including messaging and banking applications. To clearly evaluate safety apart from general capabilities, we design separate tasks measuring safety and tasks evaluating helpfulness. The safety tasks challenge agents with managing potential risks prevalent in daily life and include tests to evaluate robustness against indirect prompt injections. Our experiments demonstrate that while baseline agents, based on state-of-the-art LLMs, perform well in executing helpful tasks, they show poor performance in safety tasks. To mitigate these safety concerns, we propose a prompting method that encourages agents to prioritize safety considerations. While this method shows promise in promoting safer behaviors, there is still considerable room for improvement to fully earn user trust. This highlights the urgent need for continued research to develop more robust safety mechanisms in mobile environments. We open-source our benchmark at: https://mobilesafetybench.github.io/.
☆ Large Language Models Still Exhibit Bias in Long Text
Existing fairness benchmarks for large language models (LLMs) primarily focus on simple tasks, such as multiple-choice questions, overlooking biases that may arise in more complex scenarios like long-text generation. To address this gap, we introduce the Long Text Fairness Test (LTF-TEST), a framework that evaluates biases in LLMs through essay-style prompts. LTF-TEST covers 14 topics and 10 demographic axes, including gender and race, resulting in 11,948 samples. By assessing both model responses and the reasoning behind them, LTF-TEST uncovers subtle biases that are difficult to detect in simple responses. In our evaluation of five recent LLMs, including GPT-4o and LLaMa3, we identify two key patterns of bias. First, these models frequently favor certain demographic groups in their responses. Second, they show excessive sensitivity toward traditionally disadvantaged groups, often providing overly protective responses while neglecting others. To mitigate these biases, we propose FT-REGARD, a finetuning approach that pairs biased prompts with neutral responses. FT-REGARD reduces gender bias by 34.6% and improves performance by 1.4 percentage points on the BBQ benchmark, offering a promising approach to addressing biases in long-text generation tasks.
comment: 22 page, 38 figures, Neurips (SoLaR Workshop)
☆ Mechanisms of Symbol Processing for In-Context Learning in Transformer Networks
Large Language Models (LLMs) have demonstrated impressive abilities in symbol processing through in-context learning (ICL). This success flies in the face of decades of predictions that artificial neural networks cannot master abstract symbol manipulation. We seek to understand the mechanisms that can enable robust symbol processing in transformer networks, illuminating both the unanticipated success, and the significant limitations, of transformers in symbol processing. Borrowing insights from symbolic AI on the power of Production System architectures, we develop a high-level language, PSL, that allows us to write symbolic programs to do complex, abstract symbol processing, and create compilers that precisely implement PSL programs in transformer networks which are, by construction, 100% mechanistically interpretable. We demonstrate that PSL is Turing Universal, so the work can inform the understanding of transformer ICL in general. The type of transformer architecture that we compile from PSL programs suggests a number of paths for enhancing transformers' capabilities at symbol processing. (Note: The first section of the paper gives an extended synopsis of the entire paper.)
comment: 101 pages (including 30 pages of Appendices), 18 figures
☆ BadFair: Backdoored Fairness Attacks with Group-conditioned Triggers EMNLP 2024
Attacking fairness is crucial because compromised models can introduce biased outcomes, undermining trust and amplifying inequalities in sensitive applications like hiring, healthcare, and law enforcement. This highlights the urgent need to understand how fairness mechanisms can be exploited and to develop defenses that ensure both fairness and robustness. We introduce BadFair, a novel backdoored fairness attack methodology. BadFair stealthily crafts a model that operates with accuracy and fairness under regular conditions but, when activated by certain triggers, discriminates and produces incorrect results for specific groups. This type of attack is particularly stealthy and dangerous, as it circumvents existing fairness detection methods, maintaining an appearance of fairness in normal use. Our findings reveal that BadFair achieves a more than 85% attack success rate in attacks aimed at target groups on average while only incurring a minimal accuracy loss. Moreover, it consistently exhibits a significant discrimination score, distinguishing between pre-defined target and non-target attacked groups across various datasets and models.
comment: Accepted by EMNLP 2024
☆ VoiceTextBlender: Augmenting Large Language Models with Speech Capabilities via Single-Stage Joint Speech-Text Supervised Fine-Tuning
Recent studies have augmented large language models (LLMs) with speech capabilities, leading to the development of speech language models (SpeechLMs). Earlier SpeechLMs focused on single-turn speech-based question answering (QA), where user input comprised a speech context and a text question. More recent studies have extended this to multi-turn conversations, though they often require complex, multi-stage supervised fine-tuning (SFT) with diverse data. Another critical challenge with SpeechLMs is catastrophic forgetting-where models optimized for speech tasks suffer significant degradation in text-only performance. To mitigate these issues, we propose a novel single-stage joint speech-text SFT approach on the low-rank adaptation (LoRA) of the LLM backbone. Our joint SFT combines text-only SFT data with three types of speech-related data: speech recognition and translation, speech-based QA, and mixed-modal SFT. Compared to previous SpeechLMs with 7B or 13B parameters, our 3B model demonstrates superior performance across various speech benchmarks while preserving the original capabilities on text-only tasks. Furthermore, our model shows emergent abilities of effectively handling previously unseen prompts and tasks, including multi-turn, mixed-modal inputs.
☆ Which Client is Reliable?: A Reliable and Personalized Prompt-based Federated Learning for Medical Image Question Answering
Conventional medical artificial intelligence (AI) models face barriers in clinical application and ethical issues owing to their inability to handle the privacy-sensitive characteristics of medical data. We present a novel personalized federated learning (pFL) method for medical visual question answering (VQA) models, addressing privacy reliability challenges in the medical domain. Our method introduces learnable prompts into a Transformer architecture to efficiently train it on diverse medical datasets without massive computational costs. Then we introduce a reliable client VQA model that incorporates Dempster-Shafer evidence theory to quantify uncertainty in predictions, enhancing the model's reliability. Furthermore, we propose a novel inter-client communication mechanism that uses maximum likelihood estimation to balance accuracy and uncertainty, fostering efficient integration of insights across clients.
☆ Is artificial intelligence still intelligence? LLMs generalize to novel adjective-noun pairs, but don't mimic the full human distribution
Inferences from adjective-noun combinations like "Is artificial intelligence still intelligence?" provide a good test bed for LLMs' understanding of meaning and compositional generalization capability, since there are many combinations which are novel to both humans and LLMs but nevertheless elicit convergent human judgments. We study a range of LLMs and find that the largest models we tested are able to draw human-like inferences when the inference is determined by context and can generalize to unseen adjective-noun combinations. We also propose three methods to evaluate LLMs on these inferences out of context, where there is a distribution of human-like answers rather than a single correct answer. We find that LLMs show a human-like distribution on at most 75\% of our dataset, which is promising but still leaves room for improvement.
comment: 9 pages (23 pages with appendix). Accepted to GenBench 2024
♻ ☆ MADial-Bench: Towards Real-world Evaluation of Memory-Augmented Dialogue Generation NAACL 2025
Long-term memory is important for chatbots and dialogue systems (DS) to create consistent and human-like conversations, evidenced by numerous developed memory-augmented DS (MADS). To evaluate the effectiveness of such MADS, existing commonly used evaluation metrics, like retrieval accuracy and perplexity (PPL), mainly focus on query-oriented factualness and language quality assessment. However, these metrics often lack practical value. Moreover, the evaluation dimensions are insufficient for human-like assessment in DS. Regarding memory-recalling paradigms, current evaluation schemes only consider passive memory retrieval while ignoring diverse memory recall with rich triggering factors, e.g., emotions and surroundings, which can be essential in emotional support scenarios. To bridge the gap, we construct a novel Memory-Augmented Dialogue Benchmark (MADail-Bench) covering various memory-recalling paradigms based on cognitive science and psychology theories. The benchmark assesses two tasks separately: memory retrieval and memory recognition with the incorporation of both passive and proactive memory recall data. We introduce new scoring criteria to the evaluation, including memory injection, emotion support (ES) proficiency, and intimacy, to comprehensively assess generated responses. Results from cutting-edge embedding models and large language models on this benchmark indicate the potential for further advancement. Extensive testing further reveals correlations between memory injection, ES proficiency, and intimacy.
comment: Submitted to NAACL 2025
♻ ☆ Does Generative AI speak Nigerian-Pidgin?: Issues about Representativeness and Bias for Multilingualism in LLMs
Nigeria is a multilingual country with 500+ languages. Naija is a Nigerian-Pidgin spoken by approx. 120M speakers in Nigeria and it is a mixed language (e.g., English, Portuguese, Yoruba, Hausa and Igbo). Although it has mainly been a spoken language until recently, there are now various platforms publishing exclusively in Naija such as Naija Wikipedia. However, it is hard to distinguish by non-native from a larger pidgin languages spoken across West Africa known as West African Pidgin English (WAPE) -- which is more simplied and understandable by wider audience in Ghana, Nigeria, and Cameroon. BBC news platform publishes exclusively in WAPE to cater for several countries in West Africa. In our paper, we show through statistical analyses and Machine Translation experiments that these two creole varieties do not represent each other (i.e., there are linguistic differences in word order and vocabulary) and Generative AI operates only based on WAPE. In other words, Naija is under-represented in Generative AI, and it is hard to teach LLMs with few examples.
comment: under review
♻ ☆ Conditional Language Policy: A General Framework for Steerable Multi-Objective Finetuning EMNLP 2024
Reward-based finetuning is crucial for aligning language policies with intended behaviors (e.g., creativity and safety). A key challenge is to develop steerable language models that trade-off multiple (conflicting) objectives in a flexible and efficient manner. This paper presents Conditional Language Policy (CLP), a general framework for finetuning language models on multiple objectives. Building on techniques from multi-task training and parameter-efficient finetuning, CLP learn steerable models that effectively trade-off conflicting objectives at inference time. Notably, this does not require training or maintaining multiple models to achieve different trade-offs between the objectives. Through extensive experiments and ablations on two summarization datasets, we show that CLP learns steerable language models that outperform and Pareto-dominate the existing approaches for multi-objective finetuning.
comment: 40 pages. Findings of EMNLP 2024
♻ ☆ STAR: SocioTechnical Approach to Red Teaming Language Models
This research introduces STAR, a sociotechnical framework that improves on current best practices for red teaming safety of large language models. STAR makes two key contributions: it enhances steerability by generating parameterised instructions for human red teamers, leading to improved coverage of the risk surface. Parameterised instructions also provide more detailed insights into model failures at no increased cost. Second, STAR improves signal quality by matching demographics to assess harms for specific groups, resulting in more sensitive annotations. STAR further employs a novel step of arbitration to leverage diverse viewpoints and improve label reliability, treating disagreement not as noise but as a valuable contribution to signal quality.
comment: 8 pages, 5 figures, 5 pages appendix. * denotes equal contribution
♻ ☆ Proof of Thought : Neurosymbolic Program Synthesis allows Robust and Interpretable Reasoning NeurIPS 2024
Large Language Models (LLMs) have revolutionized natural language processing, yet they struggle with inconsistent reasoning, particularly in novel domains and complex logical sequences. This research introduces Proof of Thought, a framework that enhances the reliability and transparency of LLM outputs. Our approach bridges LLM-generated ideas with formal logic verification, employing a custom interpreter to convert LLM outputs into First Order Logic constructs for theorem prover scrutiny. Central to our method is an intermediary JSON-based Domain-Specific Language, which by design balances precise logical structures with intuitive human concepts. This hybrid representation enables both rigorous validation and accessible human comprehension of LLM reasoning processes. Key contributions include a robust type system with sort management for enhanced logical integrity, explicit representation of rules for clear distinction between factual and inferential knowledge, and a flexible architecture that allows for easy extension to various domain-specific applications. We demonstrate Proof of Thought's effectiveness through benchmarking on StrategyQA and a novel multimodal reasoning task, showing improved performance in open-ended scenarios. By providing verifiable and interpretable results, our technique addresses critical needs for AI system accountability and sets a foundation for human-in-the-loop oversight in high-stakes domains.
comment: 38th Conference on Neural Information Processing Systems (NeurIPS 2024) System 2 Reasoning At Scale Workshop
♻ ☆ AlleNoise: large-scale text classification benchmark dataset with real-world label noise
Label noise remains a challenge for training robust classification models. Most methods for mitigating label noise have been benchmarked using primarily datasets with synthetic noise. While the need for datasets with realistic noise distribution has partially been addressed by web-scraped benchmarks such as WebVision and Clothing1M, those benchmarks are restricted to the computer vision domain. With the growing importance of Transformer-based models, it is crucial to establish text classification benchmarks for learning with noisy labels. In this paper, we present AlleNoise, a new curated text classification benchmark dataset with real-world instance-dependent label noise, containing over 500,000 examples across approximately 5,600 classes, complemented with a meaningful, hierarchical taxonomy of categories. The noise distribution comes from actual users of a major e-commerce marketplace, so it realistically reflects the semantics of human mistakes. In addition to the noisy labels, we provide human-verified clean labels, which help to get a deeper insight into the noise distribution, unlike web-scraped datasets typically used in the field. We demonstrate that a representative selection of established methods for learning with noisy labels is inadequate to handle such real-world noise. In addition, we show evidence that these algorithms do not alleviate excessive memorization. As such, with AlleNoise, we set the bar high for the development of label noise methods that can handle real-world label noise in text classification tasks. The code and dataset are available for download at https://github.com/allegro/AlleNoise.
♻ ☆ Annotator-Centric Active Learning for Subjective NLP Tasks EMNLP2024
Active Learning (AL) addresses the high costs of collecting human annotations by strategically annotating the most informative samples. However, for subjective NLP tasks, incorporating a wide range of perspectives in the annotation process is crucial to capture the variability in human judgments. We introduce Annotator-Centric Active Learning (ACAL), which incorporates an annotator selection strategy following data sampling. Our objective is two-fold: 1) to efficiently approximate the full diversity of human judgments, and 2) to assess model performance using annotator-centric metrics, which value minority and majority perspectives equally. We experiment with multiple annotator selection strategies across seven subjective NLP tasks, employing both traditional and novel, human-centered evaluation metrics. Our findings indicate that ACAL improves data efficiency and excels in annotator-centric performance evaluations. However, its success depends on the availability of a sufficiently large and diverse pool of annotators to sample from.
comment: Accepted at EMNLP2024
♻ ☆ Do Large Language Models Truly Grasp Mathematics? An Empirical Exploration
Despite their proficiency in math tasks, the mechanisms underlying LLMs' mathematical reasoning abilities remain a subject of debate. Recent studies suggest that chain-of-thought (CoT) prompts can bolster mathematical reasoning by encouraging LLMs to employ human-like logical reasoning (System 2), enabling them to excel on the Cognitive Reflection Test (CRT). To assess whether LLMs genuinely possess System 2-like logical reasoning, we introduced targeted modifications to CRT problems. Our findings reveal that, despite the use of CoT prompts, mainstream LLMs, including the latest o1-preview model, continue to exhibit a significant error rate. Further analysis indicates that they predominantly rely on System 1-like intuitive reasoning and pattern matching derived from training data, rather than demonstrating mastery of mathematical thinking. This discovery challenges the prevailing notion that LLMs possess genuine logical reasoning abilities and that CoT can enhance them. Consequently, this work may temper overly optimistic projections regarding LLMs' advancement toward artificial general intelligence.
♻ ☆ Linear Adversarial Concept Erasure ICML 2022
Modern neural models trained on textual data rely on pre-trained representations that emerge without direct supervision. As these representations are increasingly being used in real-world applications, the inability to \emph{control} their content becomes an increasingly important problem. We formulate the problem of identifying and erasing a linear subspace that corresponds to a given concept, in order to prevent linear predictors from recovering the concept. We model this problem as a constrained, linear maximin game, and show that existing solutions are generally not optimal for this task. We derive a closed-form solution for certain objectives, and propose a convex relaxation, \method, that works well for others. When evaluated in the context of binary gender removal, the method recovers a low-dimensional subspace whose removal mitigates bias by intrinsic and extrinsic evaluation. We show that the method is highly expressive, effectively mitigating bias in deep nonlinear classifiers while maintaining tractability and interpretability.
comment: Accepted in ICML 2022; a revised version
♻ ☆ Fast and Slow Generating: An Empirical Study on Large and Small Language Models Collaborative Decoding
Large Language Models (LLMs) exhibit impressive capabilities across various applications but encounter substantial challenges such as high inference latency, considerable training costs, and the generation of hallucinations. Collaborative decoding between large and small language models (SLMs) presents a promising strategy to mitigate these issues through methods including speculative decoding, contrastive decoding, and emulator or proxy fine-tuning. However, the specifics of such collaborations, particularly from a unified perspective, remain largely unexplored. Inspired by dual-process cognitive theory, we propose a unified framework in this paper, termed Fast and Slow Generating (FS-GEN). Within this framework, LLMs (sometimes along with SLMs) are categorized as System 2 (slow and deliberate), while independent SLMs are designated as System 1 (fast and intuitive). We provide a comprehensive analysis of these collaborative methodologies, elucidating their common properties and shedding light on the differential knowledge capabilities of System 2 versus System 1 through the FS-GEN framework. Our findings indicate that only a small proportion of collaborative interactions (approximately less than 20\% in most instances) are necessary across various methods. These interactions between System 1 and System 2 conform to a scaling law related to the parameter ratios, enabling predictable collaboration. Furthermore, we explore the specific conditions under which collaboration proves most effective, particularly from an uncertainty perspective, offering novel insights that may guide future optimization efforts. Our research underscores that the fundamental distinction between System 1 and System 2 lies in the uncertainty of next token predictions, where interventions by System 2 are crucial to support System 1. Code for Reproduction: https://github.com/TsinghuaC3I/FS-GEN
comment: update figures and results on Pythia Series
♻ ☆ LocoMotion: Learning Motion-Focused Video-Language Representations ACCV 2024
This paper strives for motion-focused video-language representations. Existing methods to learn video-language representations use spatial-focused data, where identifying the objects and scene is often enough to distinguish the relevant caption. We instead propose LocoMotion to learn from motion-focused captions that describe the movement and temporal progression of local object motions. We achieve this by adding synthetic motions to videos and using the parameters of these motions to generate corresponding captions. Furthermore, we propose verb-variation paraphrasing to increase the caption variety and learn the link between primitive motions and high-level verbs. With this, we are able to learn a motion-focused video-language representation. Experiments demonstrate our approach is effective for a variety of downstream tasks, particularly when limited data is available for fine-tuning. Code is available: https://hazeldoughty.github.io/Papers/LocoMotion/
comment: ACCV 2024 Oral
♻ ☆ Reconfidencing LLMs from the Grouping Loss Perspective EMNLP 2024
Large Language Models (LLMs), including ChatGPT and LLaMA, are susceptible to generating hallucinated answers in a confident tone. While efforts to elicit and calibrate confidence scores have proven useful, recent findings show that controlling uncertainty must go beyond calibration: predicted scores may deviate significantly from the actual posterior probabilities due to the impact of grouping loss. In this work, we construct a new evaluation dataset derived from a knowledge base to assess confidence scores given to answers of Mistral and LLaMA. Experiments show that they tend to be overconfident. Further, we show that they are more overconfident on some answers than others, \emph{eg} depending on the nationality of the person in the query. In uncertainty-quantification theory, this is grouping loss. To address this, we propose a solution to reconfidence LLMs, canceling not only calibration but also grouping loss. The LLMs, after the reconfidencing process, indicate improved confidence alignment with the accuracy of their responses.
comment: EMNLP 2024 Findings
♻ ☆ TravelPlanner: A Benchmark for Real-World Planning with Language Agents ICML 2024
Planning has been part of the core pursuit for artificial intelligence since its conception, but earlier AI agents mostly focused on constrained settings because many of the cognitive substrates necessary for human-level planning have been lacking. Recently, language agents powered by large language models (LLMs) have shown interesting capabilities such as tool use and reasoning. Are these language agents capable of planning in more complex settings that are out of the reach of prior AI agents? To advance this investigation, we propose TravelPlanner, a new planning benchmark that focuses on travel planning, a common real-world planning scenario. It provides a rich sandbox environment, various tools for accessing nearly four million data records, and 1,225 meticulously curated planning intents and reference plans. Comprehensive evaluations show that the current language agents are not yet capable of handling such complex planning tasks-even GPT-4 only achieves a success rate of 0.6%. Language agents struggle to stay on task, use the right tools to collect information, or keep track of multiple constraints. However, we note that the mere possibility for language agents to tackle such a complex problem is in itself non-trivial progress. TravelPlanner provides a challenging yet meaningful testbed for future language agents.
comment: ICML 2024 (Spotlight)
♻ ☆ Trends in Integration of Knowledge and Large Language Models: A Survey and Taxonomy of Methods, Benchmarks, and Applications
Large language models (LLMs) exhibit superior performance on various natural language tasks, but they are susceptible to issues stemming from outdated data and domain-specific limitations. In order to address these challenges, researchers have pursued two primary strategies, knowledge editing and retrieval augmentation, to enhance LLMs by incorporating external information from different aspects. Nevertheless, there is still a notable absence of a comprehensive survey. In this paper, we propose a review to discuss the trends in integration of knowledge and large language models, including taxonomy of methods, benchmarks, and applications. In addition, we conduct an in-depth analysis of different methods and point out potential research directions in the future. We hope this survey offers the community quick access and a comprehensive overview of this research area, with the intention of inspiring future research endeavors.
comment: Work in progress; 22 pages. This work has been submitted to the IEEE for possible publication
♻ ☆ Task Prompt Vectors: Effective Initialization through Multi-Task Soft-Prompt Transfer
Prompt tuning is an efficient solution for training large language models (LLMs). However, current soft-prompt-based methods often sacrifice multi-task modularity, requiring the training process to be fully or partially repeated for each newly added task. While recent work on task vectors applied arithmetic operations on full model weights to achieve the desired multi-task performance, a similar approach for soft-prompts is still missing. To this end, we introduce Task Prompt Vectors, created by element-wise difference between weights of tuned soft-prompts and their random initialization. Experimental results on 12 NLU datasets show that task prompt vectors can be used in low-resource settings to effectively initialize prompt tuning on similar tasks. In addition, we show that task prompt vectors are independent of the random initialization of prompt tuning on 2 different language model architectures. This allows prompt arithmetics with the pre-trained vectors from different tasks. In this way, we provide a competitive alternative to state-of-the-art baselines by arithmetic addition of task prompt vectors from multiple tasks.
♻ ☆ Let Me Teach You: Pedagogical Foundations of Feedback for Language Models EMNLP 2024
Natural Language Feedback (NLF) is an increasingly popular mechanism for aligning Large Language Models (LLMs) to human preferences. Despite the diversity of the information it can convey, NLF methods are often hand-designed and arbitrary, with little systematic grounding. At the same time, research in learning sciences has long established several effective feedback models. In this opinion piece, we compile ideas from pedagogy to introduce FELT, a feedback framework for LLMs that outlines various characteristics of the feedback space, and a feedback content taxonomy based on these variables, providing a general mapping of the feedback space. In addition to streamlining NLF designs, FELT also brings out new, unexplored directions for research in NLF. We make our taxonomy available to the community, providing guides and examples for mapping our categorizations to future research.
comment: EMNLP 2024; 9 pages, 3 figures
♻ ☆ CPE-Pro: A Structure-Sensitive Deep Learning Method for Protein Representation and Origin Evaluation
Protein structures are important for understanding their functions and interactions. Currently, many protein structure prediction methods are enriching the structure database. Discriminating the origin of structures is crucial for distinguishing between experimentally resolved and computationally predicted structures, evaluating the reliability of prediction methods, and guiding downstream biological studies. Building on works in structure prediction, We developed a structure-sensitive supervised deep learning model, Crystal vs Predicted Evaluator for Protein Structure (CPE-Pro), to represent and discriminate the origin of protein structures. CPE-Pro learns the structural information of proteins and captures inter-structural differences to achieve accurate traceability on four data classes, and is expected to be extended to more. Simultaneously, we utilized Foldseek to encode protein structures into "structure-sequences" and trained a protein Structural Sequence Language Model, SSLM. Preliminary experiments demonstrated that, compared to large-scale protein language models pre-trained on vast amounts of amino acid sequences, the "structure-sequence" enables the language model to learn more informative protein features, enhancing and optimizing structural representations. We have provided the code, model weights, and all related materials on https://github.com/GouWenrui/CPE-Pro-main.git.
♻ ☆ Do LLMs Have Distinct and Consistent Personality? TRAIT: Personality Testset designed for LLMs with Psychometrics
Recent advancements in Large Language Models (LLMs) have led to their adaptation in various domains as conversational agents. We wonder: can personality tests be applied to these agents to analyze their behavior, similar to humans? We introduce TRAIT, a new benchmark consisting of 8K multi-choice questions designed to assess the personality of LLMs. TRAIT is built on two psychometrically validated small human questionnaires, Big Five Inventory (BFI) and Short Dark Triad (SD-3), enhanced with the ATOMIC-10X knowledge graph to a variety of real-world scenarios. TRAIT also outperforms existing personality tests for LLMs in terms of reliability and validity, achieving the highest scores across four key metrics: Content Validity, Internal Validity, Refusal Rate, and Reliability. Using TRAIT, we reveal two notable insights into personalities of LLMs: 1) LLMs exhibit distinct and consistent personality, which is highly influenced by their training data (e.g., data used for alignment tuning), and 2) current prompting techniques have limited effectiveness in eliciting certain traits, such as high psychopathy or low conscientiousness, suggesting the need for further research in this direction.
comment: Preprint; Under review
♻ ☆ Attribute or Abstain: Large Language Models as Long Document Assistants EMNLP 2024
LLMs can help humans working with long documents, but are known to hallucinate. Attribution can increase trust in LLM responses: The LLM provides evidence that supports its response, which enhances verifiability. Existing approaches to attribution have only been evaluated in RAG settings, where the initial retrieval confounds LLM performance. This is crucially different from the long document setting, where retrieval is not needed, but could help. Thus, a long document specific evaluation of attribution is missing. To fill this gap, we present LAB, a benchmark of 6 diverse long document tasks with attribution, and experiments with different approaches to attribution on 5 LLMs of different sizes. We find that citation, i.e. response generation and evidence extraction in one step, performs best for large and fine-tuned models, while additional retrieval can help for small, prompted models. We investigate whether the "Lost in the Middle'' phenomenon exists for attribution, but do not find this. We also find that evidence quality can predict response quality on datasets with simple responses, but not so for complex responses, as models struggle with providing evidence for complex claims.
comment: Accepted at EMNLP 2024. Code and data: https://github.com/UKPLab/arxiv2024-attribute-or-abstain
♻ ☆ I've Got 99 Problems But FLOPS Ain't One
Hyperscalers dominate the landscape of large network deployments, yet they rarely share data or insights about the challenges they face. In light of this supremacy, what problems can we find to solve in this space? We take an unconventional approach to find relevant research directions, starting from public plans to build a $100 billion datacenter for machine learning applications. Leveraging the language models scaling laws, we discover what workloads such a datacenter might carry and explore the challenges one may encounter in doing so, with a focus on networking research. We conclude that building the datacenter and training such models is technically possible, but this requires novel wide-area transports for inter-DC communication, a multipath transport and novel datacenter topologies for intra-datacenter communication, high speed scale-up networks and transports, outlining a rich research agenda for the networking community.
♻ ☆ Interpreting Context Look-ups in Transformers: Investigating Attention-MLP Interactions EMNLP 2024
Understanding the inner workings of large language models (LLMs) is crucial for advancing their theoretical foundations and real-world applications. While the attention mechanism and multi-layer perceptrons (MLPs) have been studied independently, their interactions remain largely unexplored. This study investigates how attention heads and next-token neurons interact in LLMs to predict new words. We propose a methodology to identify next-token neurons, find prompts that highly activate them, and determine the upstream attention heads responsible. We then generate and evaluate explanations for the activity of these attention heads in an automated manner. Our findings reveal that some attention heads recognize specific contexts relevant to predicting a token and activate a downstream token-predicting neuron accordingly. This mechanism provides a deeper understanding of how attention heads work with MLP neurons to perform next-token prediction. Our approach offers a foundation for further research into the intricate workings of LLMs and their impact on text generation and understanding.
comment: Accepted to EMNLP 2024 Main Conference
♻ ☆ Few-Shot Adversarial Prompt Learning on Vision-Language Models NeurIPS 2024
The vulnerability of deep neural networks to imperceptible adversarial perturbations has attracted widespread attention. Inspired by the success of vision-language foundation models, previous efforts achieved zero-shot adversarial robustness by aligning adversarial visual features with text supervision. However, in practice, they are still unsatisfactory due to several issues, including heavy adaptation cost, suboptimal text supervision, and uncontrolled natural generalization capacity. In this paper, to address these issues, we propose a few-shot adversarial prompt framework where adapting input sequences with limited data makes significant adversarial robustness improvement. Specifically, we achieve this by providing adversarially correlated text supervision that is end-to-end learned from adversarial examples. We also propose a novel training objective that enhances the consistency of multi-modal features while encourages differentiated uni-modal features between natural and adversarial examples. The proposed framework gives access to learn adversarial text supervision, which provides superior cross-modal adversarial alignment and matches state-of-the-art zero-shot adversarial robustness with only 1% training data. Code is available at: https://github.com/lionel-w2/FAP.
comment: NeurIPS 2024
♻ ☆ Do Large Language Models Have an English Accent? Evaluating and Improving the Naturalness of Multilingual LLMs
Current Large Language Models (LLMs) are predominantly designed with English as the primary language, and even the few that are multilingual tend to exhibit strong English-centric biases. Much like speakers who might produce awkward expressions when learning a second language, LLMs often generate unnatural outputs in non-English languages, reflecting English-centric patterns in both vocabulary and grammar. Despite the importance of this issue, the naturalness of multilingual LLM outputs has received limited attention. In this paper, we address this gap by introducing novel automatic corpus-level metrics to assess the lexical and syntactic naturalness of LLM outputs in a multilingual context. Using our new metrics, we evaluate state-of-the-art LLMs on a curated benchmark in French and Chinese, revealing a tendency towards English-influenced patterns. To mitigate this issue, we also propose a simple and effective alignment method to improve the naturalness of an LLM in a target language and domain, achieving consistent improvements in naturalness without compromising the performance on general-purpose benchmarks. Our work highlights the importance of developing multilingual metrics, resources and methods for the new wave of multilingual LLMs.
♻ ☆ RaTEScore: A Metric for Radiology Report Generation EMNLP 2024
This paper introduces a novel, entity-aware metric, termed as Radiological Report (Text) Evaluation (RaTEScore), to assess the quality of medical reports generated by AI models. RaTEScore emphasizes crucial medical entities such as diagnostic outcomes and anatomical details, and is robust against complex medical synonyms and sensitive to negation expressions. Technically, we developed a comprehensive medical NER dataset, RaTE-NER, and trained an NER model specifically for this purpose. This model enables the decomposition of complex radiological reports into constituent medical entities. The metric itself is derived by comparing the similarity of entity embeddings, obtained from a language model, based on their types and relevance to clinical significance. Our evaluations demonstrate that RaTEScore aligns more closely with human preference than existing metrics, validated both on established public benchmarks and our newly proposed RaTE-Eval benchmark.
comment: EMNLP 2024
♻ ☆ Can Language Models Induce Grammatical Knowledge from Indirect Evidence? EMNLP 2024
What kinds of and how much data is necessary for language models to induce grammatical knowledge to judge sentence acceptability? Recent language models still have much room for improvement in their data efficiency compared to humans. This paper investigates whether language models efficiently use indirect data (indirect evidence), from which they infer sentence acceptability. In contrast, humans use indirect evidence efficiently, which is considered one of the inductive biases contributing to efficient language acquisition. To explore this question, we introduce the Wug InDirect Evidence Test (WIDET), a dataset consisting of training instances inserted into the pre-training data and evaluation instances. We inject synthetic instances with newly coined wug words into pretraining data and explore the model's behavior on evaluation data that assesses grammatical acceptability regarding those words. We prepare the injected instances by varying their levels of indirectness and quantity. Our experiments surprisingly show that language models do not induce grammatical knowledge even after repeated exposure to instances with the same structure but differing only in lexical items from evaluation instances in certain language phenomena. Our findings suggest a potential direction for future research: developing models that use latent indirect evidence to induce grammatical knowledge.
comment: This paper is accepted at EMNLP 2024 Main
♻ ☆ A Review of Prominent Paradigms for LLM-Based Agents: Tool Use (Including RAG), Planning, and Feedback Learning
Tool use, planning, and feedback learning are currently three prominent paradigms for developing Large Language Model (LLM)-based agents across various tasks. Although numerous frameworks have been devised for each paradigm, their intricate workflows and inconsistent taxonomy create challenges in understanding and reviewing the frameworks across different paradigms. This survey introduces a unified taxonomy to systematically review and discuss these frameworks. Specifically, 1) the taxonomy defines environments/tasks, common LLM-profiled roles or LMPRs (policy models, evaluators, and dynamic models), and universally applicable workflows found in prior work, and 2) it enables a comparison of key perspectives on the implementations of LMPRs and workflow designs across different agent paradigms and frameworks. 3) Finally, we identify three limitations in existing workflow designs and systematically discuss the future work.
comment: Under Review
♻ ☆ GRAMMAR: Grounded and Modular Methodology for Assessment of Closed-Domain Retrieval-Augmented Language Model
Retrieval-Augmented Generation (RAG) systems are widely used across various industries for querying closed-domain and in-house knowledge bases. However, evaluating these systems presents significant challenges due to the private nature of closed-domain data and a scarcity of queries with verifiable ground truths. Moreover, there is a lack of analytical methods to diagnose problematic modules and identify types of failure, such as those caused by knowledge deficits or issues with robustness. To address these challenges, we introduce GRAMMAR (GRounded And Modular Methodology for Assessment of RAG), an evaluation framework comprising a grounded data generation process and an evaluation protocol that effectively pinpoints defective modules. Our validation experiments reveal that GRAMMAR provides a reliable approach for identifying vulnerable modules and supports hypothesis testing for textual form vulnerabilities. An open-source tool accompanying this framework is available in our GitHub repository (see https://github.com/xinzhel/grammar), allowing for easy reproduction of our results and enabling reliable and modular evaluation in closed-domain settings.
comment: Under Review
♻ ☆ 1-bit AI Infra: Part 1.1, Fast and Lossless BitNet b1.58 Inference on CPUs
Recent advances in 1-bit Large Language Models (LLMs), such as BitNet and BitNet b1.58, present a promising approach to enhancing the efficiency of LLMs in terms of speed and energy consumption. These developments also enable local LLM deployment across a broad range of devices. In this work, we introduce bitnet.cpp, a tailored software stack designed to unlock the full potential of 1-bit LLMs. Specifically, we develop a set of kernels to support fast and lossless inference of ternary BitNet b1.58 LLMs on CPUs. Extensive experiments demonstrate that bitnet.cpp achieves significant speedups, ranging from 2.37x to 6.17x on x86 CPUs and from 1.37x to 5.07x on ARM CPUs, across various model sizes. The code is available at https://github.com/microsoft/BitNet.
♻ ☆ From Keywords to Structured Summaries: Streamlining Scholarly Information Access ISWC 2024
This paper highlights the growing importance of information retrieval (IR) engines in the scientific community, addressing the inefficiency of traditional keyword-based search engines due to the rising volume of publications. The proposed solution involves structured records, underpinning advanced information technology (IT) tools, including visualization dashboards, to revolutionize how researchers access and filter articles, replacing the traditional text-heavy approach. This vision is exemplified through a proof of concept centered on the "reproductive number estimate of infectious diseases" research theme, using a fine-tuned large language model (LLM) to automate the creation of structured records to populate a backend database that now goes beyond keywords. The result is a next-generation information access system as an IR method accessible at https://orkg.org/usecases/r0-estimates.
comment: 8 pages, 3 figures | Accepted for publication as a poster paper at the International Semantic Web Conference (ISWC 2024)
♻ ☆ Learning to Plan for Retrieval-Augmented Large Language Models from Knowledge Graphs EMNLP2024
Improving the performance of large language models (LLMs) in complex question-answering (QA) scenarios has always been a research focal point. Recent studies have attempted to enhance LLMs' performance by combining step-wise planning with external retrieval. While effective for advanced models like GPT-3.5, smaller LLMs face challenges in decomposing complex questions, necessitating supervised fine-tuning. Previous work has relied on manual annotation and knowledge distillation from teacher LLMs, which are time-consuming and not accurate enough. In this paper, we introduce a novel framework for enhancing LLMs' planning capabilities by using planning data derived from knowledge graphs (KGs). LLMs fine-tuned with this data have improved planning capabilities, better equipping them to handle complex QA tasks that involve retrieval. Evaluations on multiple datasets, including our newly proposed benchmark, highlight the effectiveness of our framework and the benefits of KG-derived planning data.
comment: EMNLP2024 Findings
♻ ☆ Uncertainty Estimation and Quantification for LLMs: A Simple Supervised Approach
In this paper, we study the problem of uncertainty estimation and calibration for LLMs. We begin by formulating the uncertainty estimation problem, a relevant yet underexplored area in existing literature. We then propose a supervised approach that leverages labeled datasets to estimate the uncertainty in LLMs' responses. Based on the formulation, we illustrate the difference between the uncertainty estimation for LLMs and that for standard ML models and explain why the hidden neurons of the LLMs may contain uncertainty information. Our designed approach demonstrates the benefits of utilizing hidden activations to enhance uncertainty estimation across various tasks and shows robust transferability in out-of-distribution settings. We distinguish the uncertainty estimation task from the uncertainty calibration task and show that better uncertainty estimation leads to better calibration performance. Furthermore, our method is easy to implement and adaptable to different levels of model accessibility including black box, grey box, and white box.
comment: 29 pages, 14 figures
♻ ☆ Regularizing Hidden States Enables Learning Generalizable Reward Model for LLMs NeurIPS 2024
Reward models trained on human preference data have been proven to effectively align Large Language Models (LLMs) with human intent within the framework of reinforcement learning from human feedback (RLHF). However, current reward models have limited generalization capabilities to unseen prompts and responses, which can lead to an unexpected phenomenon known as reward over-optimization, resulting in a decline in actual performance due to excessive optimization of rewards. While previous research has advocated for constraining policy optimization, our study introduces a novel approach to enhance the reward model's generalization ability against distribution shifts by regularizing the hidden states. Specifically, we retain the base model's language model head and incorporate a suite of text-generation losses to preserve the hidden states' text-generation capabilities, while concurrently learning a reward head behind the same hidden states. Our experimental results demonstrate that the introduced regularization technique markedly improves the accuracy of learned reward models across a variety of out-of-distribution (OOD) tasks and effectively alleviates the over-optimization issue in RLHF, offering a more reliable and robust preference learning paradigm.
comment: NeurIPS 2024
♻ ☆ Mind's Eye of LLMs: Visualization-of-Thought Elicits Spatial Reasoning in Large Language Models NeurIPS 2024
Large language models (LLMs) have exhibited impressive performance in language comprehension and various reasoning tasks. However, their abilities in spatial reasoning, a crucial aspect of human cognition, remain relatively unexplored. Human possess a remarkable ability to create mental images of unseen objects and actions through a process known as the Mind's Eye, enabling the imagination of the unseen world. Inspired by this cognitive capacity, we propose Visualization-of-Thought (VoT) prompting. VoT aims to elicit spatial reasoning of LLMs by visualizing their reasoning traces, thereby guiding subsequent reasoning steps. We employed VoT for multi-hop spatial reasoning tasks, including natural language navigation, visual navigation, and visual tiling in 2D grid worlds. Experimental results demonstrated that VoT significantly enhances the spatial reasoning abilities of LLMs. Notably, VoT outperformed existing multimodal large language models (MLLMs) in these tasks. While VoT works surprisingly well on LLMs, the ability to generate mental images to facilitate spatial reasoning resembles the mind's eye process, suggesting its potential viability in MLLMs. Please find the dataset and codes at https://microsoft.github.io/visualization-of-thought
comment: 38th Conference on Neural Information Processing Systems (NeurIPS 2024)
♻ ☆ GPT-SW3: An Autoregressive Language Model for the Nordic Languages
This paper details the process of developing the first native large generative language model for the Nordic languages, GPT-SW3. We cover all parts of the development process, from data collection and processing, training configuration and instruction finetuning, to evaluation and considerations for release strategies. We hope that this paper can serve as a guide and reference for other researchers that undertake the development of large generative models for smaller languages.
♻ ☆ Non-myopic Generation of Language Model for Reasoning and Planning
Large Language Models have demonstrated remarkable abilities in reasoning and planning by breaking down complex problems into sequential steps. Despite their success in various domains like mathematical problem-solving and coding, LLMs face challenges in ensuring reliable and optimal planning due to their inherent myopic nature of autoregressive decoding. This paper revisits LLM reasoning from an optimal-control perspective, proposing a novel method, Predictive-Decoding, that leverages Model Predictive Control to enhance planning accuracy. By re-weighting LLM distributions based on foresight trajectories, Predictive-Decoding aims to mitigate early errors and promote non-myopic planning. Our experiments show significant improvements in a wide range of tasks for math, coding, and agents. Furthermore, Predictive-Decoding demonstrates computational efficiency, outperforming search baselines with reduced computational resources. This study provides insights into optimizing LLM planning capabilities.
♻ ☆ Generative AI Security: Challenges and Countermeasures
Generative AI's expanding footprint across numerous industries has led to both excitement and increased scrutiny. This paper delves into the unique security challenges posed by Generative AI, and outlines potential research directions for managing these risks.
♻ ☆ OpenMU: Your Swiss Army Knife for Music Understanding
We present OpenMU-Bench, a large-scale benchmark suite for addressing the data scarcity issue in training multimodal language models to understand music. To construct OpenMU-Bench, we leveraged existing datasets and bootstrapped new annotations. OpenMU-Bench also broadens the scope of music understanding by including lyrics understanding and music tool usage. Using OpenMU-Bench, we trained our music understanding model, OpenMU, with extensive ablations, demonstrating that OpenMU outperforms baseline models such as MU-Llama. Both OpenMU and OpenMU-Bench are open-sourced to facilitate future research in music understanding and to enhance creative music production efficiency.
comment: Resources: https://github.com/mzhaojp22/openmu
♻ ☆ Reinforcement Learning with Dynamic Multi-Reward Weighting for Multi-Style Controllable Generation
Textual style expresses a diverse set of information, including interpersonal dynamics (e.g., formality) and the author's emotions or attitudes (e.g., disgust). An open question is how language models can be explicitly controlled so that they weave together target styles when generating text: for example, to produce text that is both negative and non-toxic. One approach to such controlled generation is multi-objective reinforcement learning (RL), but how best to combine multiple objectives in a reward function is an open question. In this paper, we investigate various formulations of multi-style rewards, including calibrated outputs from discriminators and dynamic weighting by discriminator gradient magnitudes. We find that our proposed dynamic weighting outperforms static weighting approaches with respect to style control while maintaining linguistic quality, and we explore its effectiveness in 2- and 3-style control.
♻ ☆ LLMScan: Causal Scan for LLM Misbehavior Detection
Despite the success of Large Language Models (LLMs) across various fields, their potential to generate untruthful, biased and harmful responses poses significant risks, particularly in critical applications. This highlights the urgent need for systematic methods to detect and prevent such misbehavior. While existing approaches target specific issues such as harmful responses, this work introduces LLMScan, an innovative LLM monitoring technique based on causality analysis, offering a comprehensive solution. LLMScan systematically monitors the inner workings of an LLM through the lens of causal inference, operating on the premise that the LLM's `brain' behaves differently when misbehaving. By analyzing the causal contributions of the LLM's input tokens and transformer layers, LLMScan effectively detects misbehavior. Extensive experiments across various tasks and models reveal clear distinctions in the causal distributions between normal behavior and misbehavior, enabling the development of accurate, lightweight detectors for a variety of misbehavior detection tasks.
♻ ☆ BrainTransformers: SNN-LLM
This study introduces BrainTransformers, an innovative Large Language Model (LLM) implemented using Spiking Neural Networks (SNN). Our key contributions include: (1) designing SNN-compatible Transformer components such as SNNMatmul, SNNSoftmax, and SNNSiLU; (2) implementing an SNN approximation of the SiLU activation function; and (3) developing a Synapsis module to simulate synaptic plasticity. Our 3-billion parameter model, BrainTransformers-3B-Chat, demonstrates competitive performance across various benchmarks, including MMLU (63.2), BBH (54.1), ARC-C (54.3), and GSM8K (76.3), while potentially offering improved energy efficiency and biological plausibility. The model employs a three-stage training approach, including SNN-specific neuronal synaptic plasticity training. This research opens new avenues for brain-like AI systems in natural language processing and neuromorphic computing. Future work will focus on hardware optimization, developing specialized SNN fine-tuning tools, and exploring practical applications in energy-efficient computing environments.
♻ ☆ TSDS: Data Selection for Task-Specific Model Finetuning
Finetuning foundation models for specific tasks is an emerging paradigm in modern machine learning. The efficacy of task-specific finetuning largely depends on the selection of appropriate training data. We present TSDS (Task-Specific Data Selection), a framework to select data for task-specific model finetuning, guided by a small but representative set of examples from the target task. To do so, we formulate data selection for task-specific finetuning as an optimization problem with a distribution alignment loss based on optimal transport to capture the discrepancy between the selected data and the target distribution. In addition, we add a regularizer to encourage the diversity of the selected data and incorporate kernel density estimation into the regularizer to reduce the negative effects of near-duplicates among the candidate data. We connect our optimization problem to nearest neighbor search and design efficient algorithms to compute the optimal solution based on approximate nearest neighbor search techniques. We evaluate our method on data selection for both continued pretraining and instruction tuning of language models. We show that instruction tuning using data selected by our method with a 1% selection ratio often outperforms using the full dataset and beats the baseline selection methods by 1.5 points in F1 score on average.
comment: 31 pages, 1 figure
♻ ☆ Selective Vision is the Challenge for Visual Reasoning: A Benchmark for Visual Argument Understanding EMNLP 2024
Visual arguments, often used in advertising or social causes, rely on images to persuade viewers to do or believe something. Understanding these arguments requires selective vision: only specific visual stimuli within an image are relevant to the argument, and relevance can only be understood within the context of a broader argumentative structure. While visual arguments are readily appreciated by human audiences, we ask: are today's AI capable of similar understanding? We present VisArgs, a dataset of 1,611 images annotated with 5,112 visual premises (with regions), 5,574 commonsense premises, and reasoning trees connecting them into structured arguments. We propose three tasks for evaluating visual argument understanding: premise localization, premise identification, and conclusion deduction. Experiments show that 1) machines struggle to capture visual cues: GPT-4-O achieved 78.5% accuracy, while humans reached 98.0%. Models also performed 19.5% worse when distinguishing between irrelevant objects within the image compared to external objects. 2) Providing relevant visual premises improved model performance significantly.
comment: 12 pages, 6 figures. Accepted as main paper in EMNLP 2024
♻ ☆ Do LLMs Have Political Correctness? Analyzing Ethical Biases and Jailbreak Vulnerabilities in AI Systems
Although large language models (LLMs) demonstrate impressive proficiency in various tasks, they present potential safety risks, such as `jailbreaks', where malicious inputs can coerce LLMs into generating harmful content. To address these issues, many LLM developers have implemented various safety measures to align these models. This alignment involves several techniques, including data filtering during pre-training, supervised fine-tuning, reinforcement learning from human feedback, and red-teaming exercises. These methods often introduce deliberate and intentional biases similar to Political Correctness (PC) to ensure the ethical behavior of LLMs. In this paper, we delve into the intentional biases injected into LLMs for safety purposes and examine methods to circumvent these safety alignment techniques. Notably, these intentional biases result in a jailbreaking success rate in GPT-4o models that differs by 20% between non-binary and cisgender keywords and by 16% between white and black keywords, even when the other parts of the prompts are identical. We introduce the concept of PCJailbreak, highlighting the inherent risks posed by these safety-induced biases. Additionally, we propose an efficient defense method PCDefense, which prevents jailbreak attempts by injecting defense prompts prior to generation. PCDefense stands as an appealing alternative to Guard Models, such as Llama-Guard, that require additional inference cost after text generation. Our findings emphasize the urgent need for LLM developers to adopt a more responsible approach when designing and implementing safety measures.
♻ ☆ Susu Box or Piggy Bank: Assessing Cultural Commonsense Knowledge between Ghana and the U.S EMNLP 2024
Recent work has highlighted the culturally-contingent nature of commonsense knowledge. We introduce AMAMMER${\epsilon}$, a test set of 525 multiple-choice questions designed to evaluate the commonsense knowledge of English LLMs, relative to the cultural contexts of Ghana and the United States. To create AMAMMER${\epsilon}$, we select a set of multiple-choice questions (MCQs) from existing commonsense datasets and rewrite them in a multi-stage process involving surveys of Ghanaian and U.S. participants. In three rounds of surveys, participants from both pools are solicited to (1) write correct and incorrect answer choices, (2) rate individual answer choices on a 5-point Likert scale, and (3) select the best answer choice from the newly-constructed MCQ items, in a final validation step. By engaging participants at multiple stages, our procedure ensures that participant perspectives are incorporated both in the creation and validation of test items, resulting in high levels of agreement within each pool. We evaluate several off-the-shelf English LLMs on AMAMMER${\epsilon}$. Uniformly, models prefer answers choices that align with the preferences of U.S. annotators over Ghanaian annotators. Additionally, when test items specify a cultural context (Ghana or the U.S.), models exhibit some ability to adapt, but performance is consistently better in U.S. contexts than Ghanaian. As large resources are devoted to the advancement of English LLMs, our findings underscore the need for culturally adaptable models and evaluations to meet the needs of diverse English-speaking populations around the world.
comment: Accepted to EMNLP 2024
♻ ☆ A Bi-consolidating Model for Joint Relational Triple Extraction
Current methods to extract relational triples directly make a prediction based on a possible entity pair in a raw sentence without depending on entity recognition. The task suffers from a serious semantic overlapping problem, in which several relation triples may share one or two entities in a sentence. In this paper, based on a two-dimensional sentence representation, a bi-consolidating model is proposed to address this problem by simultaneously reinforcing the local and global semantic features relevant to a relation triple. This model consists of a local consolidation component and a global consolidation component. The first component uses a pixel difference convolution to enhance semantic information of a possible triple representation from adjacent regions and mitigate noise in neighbouring neighbours. The second component strengthens the triple representation based a channel attention and a spatial attention, which has the advantage to learn remote semantic dependencies in a sentence. They are helpful to improve the performance of both entity identification and relation type classification in relation triple extraction. After evaluated on several publish datasets, the bi-consolidating model achieves competitive performance. Analytical experiments demonstrate the effectiveness of our model for relational triple extraction and give motivation for other natural language processing tasks.
♻ ☆ When "Competency" in Reasoning Opens the Door to Vulnerability: Jailbreaking LLMs via Novel Complex Ciphers
Recent advancements in the safety of Large Language Models (LLMs) have primarily focused on mitigating attacks crafted in natural language or in common encryption techniques like Base64. However, new models which often possess better reasoning capabilities, open the door to new attack vectors that were previously non-existent in older models. This seems counter-intuitive at first glance, but these advanced models can decipher more complex cryptic queries that previous models could not, making them susceptible to attacks using such prompts. To exploit this vulnerability, we propose Attacks using Custom Encryptions (ACE), a novel method to jailbreak LLMs by leveraging custom encryption schemes. We evaluate the effectiveness of ACE on four state-of-the-art LLMs, achieving Attack Success Rates (ASR) of up to 66% on close-source models and 88% on open-source models. Building upon this, we introduce Layered Attacks using Custom Encryptions (LACE), which employs multiple layers of encryption through our custom ciphers to further enhance the ASR. Our findings demonstrate that LACE significantly enhances the ability to jailbreak LLMs, increasing the ASR of GPT-4o from 40% to 78%, a 38% improvement. Our results highlight that the advanced capabilities of LLMs introduce unforeseen vulnerabilities to complex attacks. Specifically complex and layered ciphers increase the chance of jailbreaking.
comment: 14 pages, 7 figures
Computer Vision and Pattern Recognition 124
☆ DynamicCity: Large-Scale LiDAR Generation from Dynamic Scenes
LiDAR scene generation has been developing rapidly recently. However, existing methods primarily focus on generating static and single-frame scenes, overlooking the inherently dynamic nature of real-world driving environments. In this work, we introduce DynamicCity, a novel 4D LiDAR generation framework capable of generating large-scale, high-quality LiDAR scenes that capture the temporal evolution of dynamic environments. DynamicCity mainly consists of two key models. 1) A VAE model for learning HexPlane as the compact 4D representation. Instead of using naive averaging operations, DynamicCity employs a novel Projection Module to effectively compress 4D LiDAR features into six 2D feature maps for HexPlane construction, which significantly enhances HexPlane fitting quality (up to 12.56 mIoU gain). Furthermore, we utilize an Expansion & Squeeze Strategy to reconstruct 3D feature volumes in parallel, which improves both network training efficiency and reconstruction accuracy than naively querying each 3D point (up to 7.05 mIoU gain, 2.06x training speedup, and 70.84% memory reduction). 2) A DiT-based diffusion model for HexPlane generation. To make HexPlane feasible for DiT generation, a Padded Rollout Operation is proposed to reorganize all six feature planes of the HexPlane as a squared 2D feature map. In particular, various conditions could be introduced in the diffusion or sampling process, supporting versatile 4D generation applications, such as trajectory- and command-driven generation, inpainting, and layout-conditioned generation. Extensive experiments on the CarlaSC and Waymo datasets demonstrate that DynamicCity significantly outperforms existing state-of-the-art 4D LiDAR generation methods across multiple metrics. The code will be released to facilitate future research.
comment: Preprint; 29 pages, 15 figures, 7 tables; Project Page at https://dynamic-city.github.io/
☆ FIPER: Generalizable Factorized Fields for Joint Image Compression and Super-Resolution
In this work, we propose a unified representation for Super-Resolution (SR) and Image Compression, termed **Factorized Fields**, motivated by the shared principles between these two tasks. Both SISR and Image Compression require recovering and preserving fine image details--whether by enhancing resolution or reconstructing compressed data. Unlike previous methods that mainly focus on network architecture, our proposed approach utilizes a basis-coefficient decomposition to explicitly capture multi-scale visual features and structural components in images, addressing the core challenges of both tasks. We first derive our SR model, which includes a Coefficient Backbone and Basis Swin Transformer for generalizable Factorized Fields. Then, to further unify these two tasks, we leverage the strong information-recovery capabilities of the trained SR modules as priors in the compression pipeline, improving both compression efficiency and detail reconstruction. Additionally, we introduce a merged-basis compression branch that consolidates shared structures, further optimizing the compression process. Extensive experiments show that our unified representation delivers state-of-the-art performance, achieving an average relative improvement of 204.4% in PSNR over the baseline in Super-Resolution (SR) and 9.35% BD-rate reduction in Image Compression compared to the previous SOTA.
comment: Project page: https://jayisaking.github.io/FIPER/
☆ FreeVS: Generative View Synthesis on Free Driving Trajectory
Existing reconstruction-based novel view synthesis methods for driving scenes focus on synthesizing camera views along the recorded trajectory of the ego vehicle. Their image rendering performance will severely degrade on viewpoints falling out of the recorded trajectory, where camera rays are untrained. We propose FreeVS, a novel fully generative approach that can synthesize camera views on free new trajectories in real driving scenes. To control the generation results to be 3D consistent with the real scenes and accurate in viewpoint pose, we propose the pseudo-image representation of view priors to control the generation process. Viewpoint transformation simulation is applied on pseudo-images to simulate camera movement in each direction. Once trained, FreeVS can be applied to any validation sequences without reconstruction process and synthesis views on novel trajectories. Moreover, we propose two new challenging benchmarks tailored to driving scenes, which are novel camera synthesis and novel trajectory synthesis, emphasizing the freedom of viewpoints. Given that no ground truth images are available on novel trajectories, we also propose to evaluate the consistency of images synthesized on novel trajectories with 3D perception models. Experiments on the Waymo Open Dataset show that FreeVS has a strong image synthesis performance on both the recorded trajectories and novel trajectories. Project Page: https://freevs24.github.io/
comment: Project Page: https://freevs24.github.io/
☆ UnCLe: Unsupervised Continual Learning of Depth Completion
We propose UnCLe, a standardized benchmark for Unsupervised Continual Learning of a multimodal depth estimation task: Depth completion aims to infer a dense depth map from a pair of synchronized RGB image and sparse depth map. We benchmark depth completion models under the practical scenario of unsupervised learning over continuous streams of data. Existing methods are typically trained on a static, or stationary, dataset. However, when adapting to novel non-stationary distributions, they "catastrophically forget" previously learned information. UnCLe simulates these non-stationary distributions by adapting depth completion models to sequences of datasets containing diverse scenes captured from distinct domains using different visual and range sensors. We adopt representative methods from continual learning paradigms and translate them to enable unsupervised continual learning of depth completion. We benchmark these models for indoor and outdoor and investigate the degree of catastrophic forgetting through standard quantitative metrics. Furthermore, we introduce model inversion quality as an additional measure of forgetting. We find that unsupervised continual learning of depth completion is an open problem, and we invite researchers to leverage UnCLe as a development platform.
comment: Preprint
☆ WorldSimBench: Towards Video Generation Models as World Simulators
Recent advancements in predictive models have demonstrated exceptional capabilities in predicting the future state of objects and scenes. However, the lack of categorization based on inherent characteristics continues to hinder the progress of predictive model development. Additionally, existing benchmarks are unable to effectively evaluate higher-capability, highly embodied predictive models from an embodied perspective. In this work, we classify the functionalities of predictive models into a hierarchy and take the first step in evaluating World Simulators by proposing a dual evaluation framework called WorldSimBench. WorldSimBench includes Explicit Perceptual Evaluation and Implicit Manipulative Evaluation, encompassing human preference assessments from the visual perspective and action-level evaluations in embodied tasks, covering three representative embodied scenarios: Open-Ended Embodied Environment, Autonomous, Driving, and Robot Manipulation. In the Explicit Perceptual Evaluation, we introduce the HF-Embodied Dataset, a video assessment dataset based on fine-grained human feedback, which we use to train a Human Preference Evaluator that aligns with human perception and explicitly assesses the visual fidelity of World Simulators. In the Implicit Manipulative Evaluation, we assess the video-action consistency of World Simulators by evaluating whether the generated situation-aware video can be accurately translated into the correct control signals in dynamic environments. Our comprehensive evaluation offers key insights that can drive further innovation in video generation models, positioning World Simulators as a pivotal advancement toward embodied artificial intelligence.
☆ TP-Eval: Tap Multimodal LLMs' Potential in Evaluation by Customizing Prompts
Recently, multimodal large language models (MLLMs) have received much attention for their impressive capabilities. The evaluation of MLLMs is becoming critical to analyzing attributes of MLLMs and providing valuable insights. However, current benchmarks overlook the problem of prompt sensitivity - minor prompt variations may lead to significant performance fluctuations. Thus, inappropriate prompts may obscure the models' capabilities, underestimating the models' performance. Moreover, different models have different preferences for different prompts, and thus, using the same prompt for all models will cause evaluation bias. This paper analyzes this deficiency in existing benchmarks and further introduces a new evaluation framework named TP-Eval, which introduces a prompt customization method to reduce evaluation biases and tap models' potential. TP-Eval will rewrite the original prompts to different customized prompts for different models. In particular, we propose some well-designed modules for prompt customization tailored to the scenario of MLLM evaluation. Extensive experiments demonstrate the effectiveness of our approach to uncovering models' capabilities, and TP-Eval should benefit the community in developing more comprehensive and convincing MLLM evaluation benchmarks.
☆ SPIRE: Synergistic Planning, Imitation, and Reinforcement Learning for Long-Horizon Manipulation
Robot learning has proven to be a general and effective technique for programming manipulators. Imitation learning is able to teach robots solely from human demonstrations but is bottlenecked by the capabilities of the demonstrations. Reinforcement learning uses exploration to discover better behaviors; however, the space of possible improvements can be too large to start from scratch. And for both techniques, the learning difficulty increases proportional to the length of the manipulation task. Accounting for this, we propose SPIRE, a system that first uses Task and Motion Planning (TAMP) to decompose tasks into smaller learning subproblems and second combines imitation and reinforcement learning to maximize their strengths. We develop novel strategies to train learning agents when deployed in the context of a planning system. We evaluate SPIRE on a suite of long-horizon and contact-rich robot manipulation problems. We find that SPIRE outperforms prior approaches that integrate imitation learning, reinforcement learning, and planning by 35% to 50% in average task performance, is 6 times more data efficient in the number of human demonstrations needed to train proficient agents, and learns to complete tasks nearly twice as efficiently. View https://sites.google.com/view/spire-corl-2024 for more details.
comment: Conference on Robot Learning (CoRL) 2024
☆ CLEAR: Character Unlearning in Textual and Visual Modalities
Machine Unlearning (MU) is critical for enhancing privacy and security in deep learning models, particularly in large multimodal language models (MLLMs), by removing specific private or hazardous information. While MU has made significant progress in textual and visual modalities, multimodal unlearning (MMU) remains significantly underexplored, partially due to the absence of a suitable open-source benchmark. To address this, we introduce CLEAR, a new benchmark designed to evaluate MMU methods. CLEAR contains 200 fictitious individuals and 3,700 images linked with corresponding question-answer pairs, enabling a thorough evaluation across modalities. We assess 10 MU methods, adapting them for MMU, and highlight new challenges specific to multimodal forgetting. We also demonstrate that simple $\ell_1$ regularization on LoRA weights significantly mitigates catastrophic forgetting, preserving model performance on retained data. The dataset is available at https://huggingface.co/datasets/therem/CLEAR
☆ In-Pixel Foreground and Contrast Enhancement Circuits with Customizable Mapping
This paper presents an innovative in-pixel contrast enhancement circuit that performs image processing directly within the pixel circuit. The circuit can be tuned for different modes of operation. In foreground enhancement mode, it suppresses low-intensity background pixels to nearly zero, isolating the foreground for better object visibility. In contrast enhancement mode, it improves overall image contrast. The contrast enhancement function is customizable both during the design phase and in real-time, allowing the circuit to adapt to specific applications and varying lighting conditions. A model of the designed pixel circuit is developed and applied to a full pixel array, demonstrating significant improvements in image quality. Simulations performed in HSPICE show a nearly 6x increase in Michelson Contrast Ratio (CR) in the foreground enhancement mode. The simulation results indicate its potential for real-time, adaptive contrast enhancement across various imaging environments.
☆ Real time anomalies detection on video
Nowadays, many places use security cameras. Unfortunately, when an incident occurs, these technologies are used to show past events. So it can be considered as a deterrence tool than a detection tool. In this article, we will propose a deep learning approach trying to solve this problematic. This approach uses convolutional models (CNN) to extract relevant characteristics linked to the video images, theses characteristics will form times series to be analyzed by LSTM / GRU models.
☆ Scalable Ranked Preference Optimization for Text-to-Image Generation
Direct Preference Optimization (DPO) has emerged as a powerful approach to align text-to-image (T2I) models with human feedback. Unfortunately, successful application of DPO to T2I models requires a huge amount of resources to collect and label large-scale datasets, e.g., millions of generated paired images annotated with human preferences. In addition, these human preference datasets can get outdated quickly as the rapid improvements of T2I models lead to higher quality images. In this work, we investigate a scalable approach for collecting large-scale and fully synthetic datasets for DPO training. Specifically, the preferences for paired images are generated using a pre-trained reward function, eliminating the need for involving humans in the annotation process, greatly improving the dataset collection efficiency. Moreover, we demonstrate that such datasets allow averaging predictions across multiple models and collecting ranked preferences as opposed to pairwise preferences. Furthermore, we introduce RankDPO to enhance DPO-based methods using the ranking feedback. Applying RankDPO on SDXL and SD3-Medium models with our synthetically generated preference dataset ``Syn-Pic'' improves both prompt-following (on benchmarks like T2I-Compbench, GenEval, and DPG-Bench) and visual quality (through user studies). This pipeline presents a practical and scalable solution to develop better preference datasets to enhance the performance of text-to-image models.
comment: Project Page: https://snap-research.github.io/RankDPO/
☆ Characterization of the multiplicity of solutions for camera pose given two vertically-aligned landmarks and accelerometer
We consider the problem of recovering the position and orientation of a camera equipped with an accelerometer from sensor images of two labeled landmarks whose positions in a coordinate system aligned in a known way with gravity are known. This a variant on the much studied P$n$P problem of recovering camera position and orientation from $n$ points without any gravitational data. It is proved that in three types of singular cases there are infinitely many solutions, in another type of case there is one, and in a final type of case there are two. A precise characterization of each type of case. In particular, there is always a unique solution in the practically interesting case where the two landmarks are at the same altitude and the camera is at a different altitude. This case is studied by numerical simulation and an implementation on a consumer cellphone. It is also proved that if the two landmarks are unlabeled, then apart from the same singular cases, there are still always one or two solutions.
comment: 32 pages, 8 figures
☆ A Pipeline for Segmenting and Structuring RGB-D Data for Robotics Applications
We introduce a novel pipeline for segmenting and structuring color and depth (RGB-D) data. Existing processing pipelines for RGB-D data have focused on extracting geometric information alone. This approach precludes the development of more advanced robotic navigation and manipulation algorithms, which benefit from a semantic understanding of their environment. Our pipeline can segment RGB-D data into accurate semantic masks. These masks are then used to fuse raw captured point clouds into semantically separated point clouds. We store this information using the Universal Scene Description (USD) file format, a format suitable for easy querying by downstream robotics algorithms, human-friendly visualization, and robotics simulation.
☆ Robust Two-View Geometry Estimation with Implicit Differentiation IROS 2024
We present a novel two-view geometry estimation framework which is based on a differentiable robust loss function fitting. We propose to treat the robust fundamental matrix estimation as an implicit layer, which allows us to avoid backpropagation through time and significantly improves the numerical stability. To take full advantage of the information from the feature matching stage we incorporate learnable weights that depend on the matching confidences. In this way our solution brings together feature extraction, matching and two-view geometry estimation in a unified end-to-end trainable pipeline. We evaluate our approach on the camera pose estimation task in both outdoor and indoor scenarios. The experiments on several datasets show that the proposed method outperforms both classic and learning-based state-of-the-art methods by a large margin. The project webpage is available at: https://github.com/VladPyatov/ihls
comment: IROS 2024 Accepted
☆ A Wavelet Diffusion GAN for Image Super-Resolution
In recent years, diffusion models have emerged as a superior alternative to generative adversarial networks (GANs) for high-fidelity image generation, with wide applications in text-to-image generation, image-to-image translation, and super-resolution. However, their real-time feasibility is hindered by slow training and inference speeds. This study addresses this challenge by proposing a wavelet-based conditional Diffusion GAN scheme for Single-Image Super-Resolution (SISR). Our approach utilizes the diffusion GAN paradigm to reduce the timesteps required by the reverse diffusion process and the Discrete Wavelet Transform (DWT) to achieve dimensionality reduction, decreasing training and inference times significantly. The results of an experimental validation on the CelebA-HQ dataset confirm the effectiveness of our proposed scheme. Our approach outperforms other state-of-the-art methodologies successfully ensuring high-fidelity output while overcoming inherent drawbacks associated with diffusion models in time-sensitive applications.
comment: The paper has been accepted at Italian Workshop on Neural Networks (WIRN) 2024
☆ Medical Imaging Complexity and its Effects on GAN Performance ACCV
The proliferation of machine learning models in diverse clinical applications has led to a growing need for high-fidelity, medical image training data. Such data is often scarce due to cost constraints and privacy concerns. Alleviating this burden, medical image synthesis via generative adversarial networks (GANs) emerged as a powerful method for synthetically generating photo-realistic images based on existing sets of real medical images. However, the exact image set size required to efficiently train such a GAN is unclear. In this work, we experimentally establish benchmarks that measure the relationship between a sample dataset size and the fidelity of the generated images, given the dataset's distribution of image complexities. We analyze statistical metrics based on delentropy, an image complexity measure rooted in Shannon's entropy in information theory. For our pipeline, we conduct experiments with two state-of-the-art GANs, StyleGAN 3 and SPADE-GAN, trained on multiple medical imaging datasets with variable sample sizes. Across both GANs, general performance improved with increasing training set size but suffered with increasing complexity.
comment: Accepted to ACCV, Workshop on Generative AI for Synthetic Medical Data
☆ VR-Splatting: Foveated Radiance Field Rendering via 3D Gaussian Splatting and Neural Points
Recent advances in novel view synthesis (NVS), particularly neural radiance fields (NeRF) and Gaussian splatting (3DGS), have demonstrated impressive results in photorealistic scene rendering. These techniques hold great potential for applications in virtual tourism and teleportation, where immersive realism is crucial. However, the high-performance demands of virtual reality (VR) systems present challenges in directly utilizing even such fast-to-render scene representations like 3DGS due to latency and computational constraints. In this paper, we propose foveated rendering as a promising solution to these obstacles. We analyze state-of-the-art NVS methods with respect to their rendering performance and compatibility with the human visual system. Our approach introduces a novel foveated rendering approach for Virtual Reality, that leverages the sharp, detailed output of neural point rendering for the foveal region, fused with a smooth rendering of 3DGS for the peripheral vision. Our evaluation confirms that perceived sharpness and detail-richness are increased by our approach compared to a standard VR-ready 3DGS configuration. Our system meets the necessary performance requirements for real-time VR interactions, ultimately enhancing the user's immersive experience. Project page: https://lfranke.github.io/vr_splatting
☆ Gaze-Assisted Medical Image Segmentation NeurIPS'24
The annotation of patient organs is a crucial part of various diagnostic and treatment procedures, such as radiotherapy planning. Manual annotation is extremely time-consuming, while its automation using modern image analysis techniques has not yet reached levels sufficient for clinical adoption. This paper investigates the idea of semi-supervised medical image segmentation using human gaze as interactive input for segmentation correction. In particular, we fine-tuned the Segment Anything Model in Medical Images (MedSAM), a public solution that uses various prompt types as additional input for semi-automated segmentation correction. We used human gaze data from reading abdominal images as a prompt for fine-tuning MedSAM. The model was validated on a public WORD database, which consists of 120 CT scans of 16 abdominal organs. The results of the gaze-assisted MedSAM were shown to be superior to the results of the state-of-the-art segmentation models. In particular, the average Dice coefficient for 16 abdominal organs was 85.8%, 86.7%, 81.7%, and 90.5% for nnUNetV2, ResUNet, original MedSAM, and our gaze-assisted MedSAM model, respectively.
comment: 16 pages, 4 figures, Accepted to AIM-FM Workshop @ NeurIPS'24
☆ Addressing Asynchronicity in Clinical Multimodal Fusion via Individualized Chest X-ray Generation NeurIPS-24
Integrating multi-modal clinical data, such as electronic health records (EHR) and chest X-ray images (CXR), is particularly beneficial for clinical prediction tasks. However, in a temporal setting, multi-modal data are often inherently asynchronous. EHR can be continuously collected but CXR is generally taken with a much longer interval due to its high cost and radiation dose. When clinical prediction is needed, the last available CXR image might have been outdated, leading to suboptimal predictions. To address this challenge, we propose DDL-CXR, a method that dynamically generates an up-to-date latent representation of the individualized CXR images. Our approach leverages latent diffusion models for patient-specific generation strategically conditioned on a previous CXR image and EHR time series, providing information regarding anatomical structures and disease progressions, respectively. In this way, the interaction across modalities could be better captured by the latent CXR generation process, ultimately improving the prediction performance. Experiments using MIMIC datasets show that the proposed model could effectively address asynchronicity in multimodal fusion and consistently outperform existing methods.
comment: Accepted by NeurIPS-24
☆ R-CoT: Reverse Chain-of-Thought Problem Generation for Geometric Reasoning in Large Multimodal Models
Existing Large Multimodal Models (LMMs) struggle with mathematical geometric reasoning due to a lack of high-quality image-text paired data. Current geometric data generation approaches, which apply preset templates to generate geometric data or use Large Language Models (LLMs) to rephrase questions and answers (Q&A), unavoidably limit data accuracy and diversity. To synthesize higher-quality data, we propose a two-stage Reverse Chain-of-Thought (R-CoT) geometry problem generation pipeline. First, we introduce GeoChain to produce high-fidelity geometric images and corresponding descriptions highlighting relations among geometric elements. We then design a Reverse A&Q method that reasons step-by-step based on the descriptions and generates questions in reverse from the reasoning results. Experiments demonstrate that the proposed method brings significant and consistent improvements on multiple LMM baselines, achieving new performance records in the 2B, 7B, and 8B settings. Notably, R-CoT-8B significantly outperforms previous state-of-the-art open-source mathematical models by 16.6% on MathVista and 9.2% on GeoQA, while also surpassing the closed-source model GPT-4o by an average of 13% across both datasets. The code is available at https://github.com/dle666/R-CoT.
☆ A utility-based spatial analysis of residential street-level conditions; A case study of Rotterdam
Residential location choices are traditionally modelled using factors related to accessibility and socioeconomic environments, neglecting the importance of local street-level conditions. Arguably, this neglect is due to data practices. Today, however, street-level images -- which are highly effective at encoding street-level conditions -- are widely available. Additionally, recent advances in discrete choice models incorporating computer vision capabilities offer opportunities to integrate street-level conditions into residential location choice analysis. This study leverages these developments to investigate the spatial distribution of utility derived from street-level conditions in residential location choices on a city-wide scale. In our case study of Rotterdam, the Netherlands, we find that the utility derived from street-level conditions varies significantly on a highly localised scale, with conditions rapidly changing even within neighbourhoods. Our results also reveal that the high real-estate prices in the city centre cannot be attributed to attractive street-level conditions. Furthermore, whereas the city centre is characterised by relatively unattractive residential street-level conditions, neighbourhoods in the southern part of the city -- often perceived as problematic -- exhibit surprisingly appealing street-level environments. The methodological contribution of this paper is that it advances the discrete choice models incorporating computer vision capabilities by introducing a semantic regularisation layer to the model. Thereby, it adds explainability and eliminates the need for a separate pipeline to extract information from images, streamlining the analysis. As such, this paper's findings and methodological advancements pave the way for further studies to explore integrating street-level conditions in urban planning.
☆ CASCRNet: An Atrous Spatial Pyramid Pooling and Shared Channel Residual based Network for Capsule Endoscopy
This manuscript summarizes work on the Capsule Vision Challenge 2024 by MISAHUB. To address the multi-class disease classification task, which is challenging due to the complexity and imbalance in the Capsule Vision challenge dataset, this paper proposes CASCRNet (Capsule endoscopy-Aspp-SCR-Network), a parameter-efficient and novel model that uses Shared Channel Residual (SCR) blocks and Atrous Spatial Pyramid Pooling (ASPP) blocks. Further, the performance of the proposed model is compared with other well-known approaches. The experimental results yield that proposed model provides better disease classification results. The proposed model was successful in classifying diseases with an F1 Score of 78.5% and a Mean AUC of 98.3%, which is promising given its compact architecture.
comment: 8 pages, 4 figures
☆ Blendify -- Python rendering framework for Blender
With the rapid growth of the volume of research fields like computer vision and computer graphics, researchers require effective and user-friendly rendering tools to visualize results. While advanced tools like Blender offer powerful capabilities, they also require a significant effort to master. This technical report introduces Blendify, a lightweight Python-based framework that seamlessly integrates with Blender, providing a high-level API for scene creation and rendering. Blendify reduces the complexity of working with Blender's native API by automating object creation, handling the colors and material linking, and implementing features such as shadow-catcher objects while maintaining support for high-quality ray-tracing rendering output. With a focus on usability Blendify enables efficient and flexible rendering workflow for rendering in common computer vision and computer graphics use cases. The code is available at https://github.com/ptrvilya/blendify
comment: Project page: https://virtualhumans.mpi-inf.mpg.de/blendify/
☆ ROCKET-1: Master Open-World Interaction with Visual-Temporal Context Prompting
Vision-language models (VLMs) have excelled in multimodal tasks, but adapting them to embodied decision-making in open-world environments presents challenges. A key issue is the difficulty in smoothly connecting individual entities in low-level observations with abstract concepts required for planning. A common approach to address this problem is through the use of hierarchical agents, where VLMs serve as high-level reasoners that break down tasks into executable sub-tasks, typically specified using language and imagined observations. However, language often fails to effectively convey spatial information, while generating future images with sufficient accuracy remains challenging. To address these limitations, we propose visual-temporal context prompting, a novel communication protocol between VLMs and policy models. This protocol leverages object segmentation from both past and present observations to guide policy-environment interactions. Using this approach, we train ROCKET-1, a low-level policy that predicts actions based on concatenated visual observations and segmentation masks, with real-time object tracking provided by SAM-2. Our method unlocks the full potential of VLMs visual-language reasoning abilities, enabling them to solve complex creative tasks, especially those heavily reliant on spatial understanding. Experiments in Minecraft demonstrate that our approach allows agents to accomplish previously unattainable tasks, highlighting the effectiveness of visual-temporal context prompting in embodied decision-making. Codes and demos will be available on the project page: https://craftjarvis.github.io/ROCKET-1.
☆ TAGE: Trustworthy Attribute Group Editing for Stable Few-shot Image Generation
Generative Adversarial Networks (GANs) have emerged as a prominent research focus for image editing tasks, leveraging the powerful image generation capabilities of the GAN framework to produce remarkable results.However, prevailing approaches are contingent upon extensive training datasets and explicit supervision, presenting a significant challenge in manipulating the diverse attributes of new image classes with limited sample availability. To surmount this hurdle, we introduce TAGE, an innovative image generation network comprising three integral modules: the Codebook Learning Module (CLM), the Code Prediction Module (CPM) and the Prompt-driven Semantic Module (PSM). The CPM module delves into the semantic dimensions of category-agnostic attributes, encapsulating them within a discrete codebook. This module is predicated on the concept that images are assemblages of attributes, and thus, by editing these category-independent attributes, it is theoretically possible to generate images from unseen categories. Subsequently, the CPM module facilitates naturalistic image editing by predicting indices of category-independent attribute vectors within the codebook. Additionally, the PSM module generates semantic cues that are seamlessly integrated into the Transformer architecture of the CPM, enhancing the model's comprehension of the targeted attributes for editing. With these semantic cues, the model can generate images that accentuate desired attributes more prominently while maintaining the integrity of the original category, even with a limited number of samples. We have conducted extensive experiments utilizing the Animal Faces, Flowers, and VGGFaces datasets. The results of these experiments demonstrate that our proposed method not only achieves superior performance but also exhibits a high degree of stability when compared to other few-shot image generation techniques.
comment: Accepted by International Conference on Signal Processing Systems Conference
☆ Few-shot NeRF by Adaptive Rendering Loss Regularization ECCV2024
Novel view synthesis with sparse inputs poses great challenges to Neural Radiance Field (NeRF). Recent works demonstrate that the frequency regularization of Positional Encoding (PE) can achieve promising results for few-shot NeRF. In this work, we reveal that there exists an inconsistency between the frequency regularization of PE and rendering loss. This prevents few-shot NeRF from synthesizing higher-quality novel views. To mitigate this inconsistency, we propose Adaptive Rendering loss regularization for few-shot NeRF, dubbed AR-NeRF. Specifically, we present a two-phase rendering supervision and an adaptive rendering loss weight learning strategy to align the frequency relationship between PE and 2D-pixel supervision. In this way, AR-NeRF can learn global structures better in the early training phase and adaptively learn local details throughout the training process. Extensive experiments show that our AR-NeRF achieves state-of-the-art performance on different datasets, including object-level and complex scenes.
comment: Accepted by ECCV2024
☆ Exploiting Text-Image Latent Spaces for the Description of Visual Concepts ICPR
Concept Activation Vectors (CAVs) offer insights into neural network decision-making by linking human friendly concepts to the model's internal feature extraction process. However, when a new set of CAVs is discovered, they must still be translated into a human understandable description. For image-based neural networks, this is typically done by visualizing the most relevant images of a CAV, while the determination of the concept is left to humans. In this work, we introduce an approach to aid the interpretation of newly discovered concept sets by suggesting textual descriptions for each CAV. This is done by mapping the most relevant images representing a CAV into a text-image embedding where a joint description of these relevant images can be computed. We propose utilizing the most relevant receptive fields instead of full images encoded. We demonstrate the capabilities of this approach in multiple experiments with and without given CAV labels, showing that the proposed approach provides accurate descriptions for the CAVs and reduces the challenge of concept interpretation.
comment: 19 pages, 7 figures, to be published in ICPR
☆ Att2CPC: Attention-Guided Lossy Attribute Compression of Point Clouds
With the great progress of 3D sensing and acquisition technology, the volume of point cloud data has grown dramatically, which urges the development of efficient point cloud compression methods. In this paper, we focus on the task of learned lossy point cloud attribute compression (PCAC). We propose an efficient attention-based method for lossy compression of point cloud attributes leveraging on an autoencoder architecture. Specifically, at the encoding side, we conduct multiple downsampling to best exploit the local attribute patterns, in which effective External Cross Attention (ECA) is devised to hierarchically aggregate features by intergrating attributes and geometry contexts. At the decoding side, the attributes of the point cloud are progressively reconstructed based on the multi-scale representation and the zero-padding upsampling tactic. To the best of our knowledge, this is the first approach to introduce attention mechanism to point-based lossy PCAC task. We verify the compression efficiency of our model on various sequences, including human body frames, sparse objects, and large-scale point cloud scenes. Experiments show that our method achieves an average improvement of 1.15 dB and 2.13 dB in BD-PSNR of Y channel and YUV channel, respectively, when comparing with the state-of-the-art point-based method Deep-PCAC. Codes of this paper are available at https://github.com/I2-Multimedia-Lab/Att2CPC.
☆ DREB-Net: Dual-stream Restoration Embedding Blur-feature Fusion Network for High-mobility UAV Object Detection
Object detection algorithms are pivotal components of unmanned aerial vehicle (UAV) imaging systems, extensively employed in complex fields. However, images captured by high-mobility UAVs often suffer from motion blur cases, which significantly impedes the performance of advanced object detection algorithms. To address these challenges, we propose an innovative object detection algorithm specifically designed for blurry images, named DREB-Net (Dual-stream Restoration Embedding Blur-feature Fusion Network). First, DREB-Net addresses the particularities of blurry image object detection problem by incorporating a Blurry image Restoration Auxiliary Branch (BRAB) during the training phase. Second, it fuses the extracted shallow features via Multi-level Attention-Guided Feature Fusion (MAGFF) module, to extract richer features. Here, the MAGFF module comprises local attention modules and global attention modules, which assign different weights to the branches. Then, during the inference phase, the deep feature extraction of the BRAB can be removed to reduce computational complexity and improve detection speed. In loss function, a combined loss of MSE and SSIM is added to the BRAB to restore blurry images. Finally, DREB-Net introduces Fast Fourier Transform in the early stages of feature extraction, via a Learnable Frequency domain Amplitude Modulation Module (LFAMM), to adjust feature amplitude and enhance feature processing capability. Experimental results indicate that DREB-Net can still effectively perform object detection tasks under motion blur in captured images, showcasing excellent performance and broad application prospects. Our source code will be available at https://github.com/EEIC-Lab/DREB-Net.git.
☆ Deep Learning for Active Region Classification: A Systematic Study from Convolutional Neural Networks to Vision Transformers
A solar active region can significantly disrupt the Sun Earth space environment, often leading to severe space weather events such as solar flares and coronal mass ejections. As a consequence, the automatic classification of active region groups is the crucial starting point for accurately and promptly predicting solar activity. This study presents our results concerned with the application of deep learning techniques to the classification of active region cutouts based on the Mount Wilson classification scheme. Specifically, we have explored the latest advancements in image classification architectures, from Convolutional Neural Networks to Vision Transformers, and reported on their performances for the active region classification task, showing that the crucial point for their effectiveness consists in a robust training process based on the latest advances in the field.
☆ Learning Lossless Compression for High Bit-Depth Volumetric Medical Image
Recent advances in learning-based methods have markedly enhanced the capabilities of image compression. However, these methods struggle with high bit-depth volumetric medical images, facing issues such as degraded performance, increased memory demand, and reduced processing speed. To address these challenges, this paper presents the Bit-Division based Lossless Volumetric Image Compression (BD-LVIC) framework, which is tailored for high bit-depth medical volume compression. The BD-LVIC framework skillfully divides the high bit-depth volume into two lower bit-depth segments: the Most Significant Bit-Volume (MSBV) and the Least Significant Bit-Volume (LSBV). The MSBV concentrates on the most significant bits of the volumetric medical image, capturing vital structural details in a compact manner. This reduction in complexity greatly improves compression efficiency using traditional codecs. Conversely, the LSBV deals with the least significant bits, which encapsulate intricate texture details. To compress this detailed information effectively, we introduce an effective learning-based compression model equipped with a Transformer-Based Feature Alignment Module, which exploits both intra-slice and inter-slice redundancies to accurately align features. Subsequently, a Parallel Autoregressive Coding Module merges these features to precisely estimate the probability distribution of the least significant bit-planes. Our extensive testing demonstrates that the BD-LVIC framework not only sets new performance benchmarks across various datasets but also maintains a competitive coding speed, highlighting its significant potential and practical utility in the realm of volumetric medical image compression.
comment: 13 pages
☆ PGDiffSeg: Prior-Guided Denoising Diffusion Model with Parameter-Shared Attention for Breast Cancer Segmentation
Early detection through imaging and accurate diagnosis is crucial in mitigating the high mortality rate associated with breast cancer. However, locating tumors from low-resolution and high-noise medical images is extremely challenging. Therefore, this paper proposes a novel PGDiffSeg (Prior-Guided Diffusion Denoising Model with Parameter-Shared Attention) that applies diffusion denoising methods to breast cancer medical image segmentation, accurately recovering the affected areas from Gaussian noise. Firstly, we design a parallel pipeline for noise processing and semantic information processing and propose a parameter-shared attention module (PSA) in multi-layer that seamlessly integrates these two pipelines. This integration empowers PGDiffSeg to incorporate semantic details at multiple levels during the denoising process, producing highly accurate segmentation maps. Secondly, we introduce a guided strategy that leverages prior knowledge to simulate the decision-making process of medical professionals, thereby enhancing the model's ability to locate tumor positions precisely. Finally, we provide the first-ever discussion on the interpretability of the generative diffusion model in the context of breast cancer segmentation. Extensive experiments have demonstrated the superiority of our model over the current state-of-the-art approaches, confirming its effectiveness as a flexible diffusion denoising method suitable for medical image research. Our code will be publicly available later.
☆ EntityCLIP: Entity-Centric Image-Text Matching via Multimodal Attentive Contrastive Learning
Recent advancements in image-text matching have been notable, yet prevailing models predominantly cater to broad queries and struggle with accommodating fine-grained query intention. In this paper, we work towards the \textbf{E}ntity-centric \textbf{I}mage-\textbf{T}ext \textbf{M}atching (EITM), a task that the text and image involve specific entity-related information. The challenge of this task mainly lies in the larger semantic gap in entity association modeling, comparing with the general image-text matching problem.To narrow the huge semantic gap between the entity-centric text and the images, we take the fundamental CLIP as the backbone and devise a multimodal attentive contrastive learning framework to tam CLIP to adapt EITM problem, developing a model named EntityCLIP. The key of our multimodal attentive contrastive learning is to generate interpretive explanation text using Large Language Models (LLMs) as the bridge clues. In specific, we proceed by extracting explanatory text from off-the-shelf LLMs. This explanation text, coupled with the image and text, is then input into our specially crafted Multimodal Attentive Experts (MMAE) module, which effectively integrates explanation texts to narrow the gap of the entity-related text and image in a shared semantic space. Building on the enriched features derived from MMAE, we further design an effective Gated Integrative Image-text Matching (GI-ITM) strategy. The GI-ITM employs an adaptive gating mechanism to aggregate MMAE's features, subsequently applying image-text matching constraints to steer the alignment between the text and the image. Extensive experiments are conducted on three social media news benchmarks including N24News, VisualNews, and GoodNews, the results shows that our method surpasses the competition methods with a clear margin.
☆ An Intelligent Agentic System for Complex Image Restoration Problems
Real-world image restoration (IR) is inherently complex and often requires combining multiple specialized models to address diverse degradations. Inspired by human problem-solving, we propose AgenticIR, an agentic system that mimics the human approach to image processing by following five key stages: Perception, Scheduling, Execution, Reflection, and Rescheduling. AgenticIR leverages large language models (LLMs) and vision-language models (VLMs) that interact via text generation to dynamically operate a toolbox of IR models. We fine-tune VLMs for image quality analysis and employ LLMs for reasoning, guiding the system step by step. To compensate for LLMs' lack of specific IR knowledge and experience, we introduce a self-exploration method, allowing the LLM to observe and summarize restoration results into referenceable documents. Experiments demonstrate AgenticIR's potential in handling complex IR tasks, representing a promising path toward achieving general intelligence in visual processing.
☆ GenUDC: High Quality 3D Mesh Generation with Unsigned Dual Contouring Representation
Generating high-quality meshes with complex structures and realistic surfaces is the primary goal of 3D generative models. Existing methods typically employ sequence data or deformable tetrahedral grids for mesh generation. However, sequence-based methods have difficulty producing complex structures with many faces due to memory limits. The deformable tetrahedral grid-based method MeshDiffusion fails to recover realistic surfaces due to the inherent ambiguity in deformable grids. We propose the GenUDC framework to address these challenges by leveraging the Unsigned Dual Contouring (UDC) as the mesh representation. UDC discretizes a mesh in a regular grid and divides it into the face and vertex parts, recovering both complex structures and fine details. As a result, the one-to-one mapping between UDC and mesh resolves the ambiguity problem. In addition, GenUDC adopts a two-stage, coarse-to-fine generative process for 3D mesh generation. It first generates the face part as a rough shape and then the vertex part to craft a detailed shape. Extensive evaluations demonstrate the superiority of UDC as a mesh representation and the favorable performance of GenUDC in mesh generation. The code and trained models are available at https://github.com/TrepangCat/GenUDC.
comment: ACMMM 2024, code:https://github.com/TrepangCat/GenUDC
☆ TranSPORTmer: A Holistic Approach to Trajectory Understanding in Multi-Agent Sports ACCV 2024
Understanding trajectories in multi-agent scenarios requires addressing various tasks, including predicting future movements, imputing missing observations, inferring the status of unseen agents, and classifying different global states. Traditional data-driven approaches often handle these tasks separately with specialized models. We introduce TranSPORTmer, a unified transformer-based framework capable of addressing all these tasks, showcasing its application to the intricate dynamics of multi-agent sports scenarios like soccer and basketball. Using Set Attention Blocks, TranSPORTmer effectively captures temporal dynamics and social interactions in an equivariant manner. The model's tasks are guided by an input mask that conceals missing or yet-to-be-predicted observations. Additionally, we introduce a CLS extra agent to classify states along soccer trajectories, including passes, possessions, uncontrolled states, and out-of-play intervals, contributing to an enhancement in modeling trajectories. Evaluations on soccer and basketball datasets show that TranSPORTmer outperforms state-of-the-art task-specific models in player forecasting, player forecasting-imputation, ball inference, and ball imputation. https://youtu.be/8VtSRm8oGoE
comment: Accepted to ACCV 2024
☆ ADEM-VL: Adaptive and Embedded Fusion for Efficient Vision-Language Tuning
Recent advancements in multimodal fusion have witnessed the remarkable success of vision-language (VL) models, which excel in various multimodal applications such as image captioning and visual question answering. However, building VL models requires substantial hardware resources, where efficiency is restricted by two key factors: the extended input sequence of the language model with vision features demands more computational operations, and a large number of additional learnable parameters increase memory complexity. These challenges significantly restrict the broader applicability of such models. To bridge this gap, we propose ADEM-VL, an efficient vision-language method that tunes VL models based on pretrained large language models (LLMs) by adopting a parameter-free cross-attention mechanism for similarity measurements in multimodal fusion. This approach only requires embedding vision features into the language space, significantly reducing the number of trainable parameters and accelerating both training and inference speeds. To enhance representation learning in fusion module, we introduce an efficient multiscale feature generation scheme that requires only a single forward pass through the vision encoder. Moreover, we propose an adaptive fusion scheme that dynamically discards less relevant visual information for each text token based on its attention score. This ensures that the fusion process prioritizes the most pertinent visual features. With experiments on various tasks including visual question answering, image captioning, and instruction-following, we demonstrate that our framework outperforms existing approaches. Specifically, our method surpasses existing methods by an average accuracy of 0.77% on ScienceQA dataset, with reduced training and inference latency, demonstrating the superiority of our framework. The code is available at https://github.com/Hao840/ADEM-VL.
☆ Quasi-Medial Distance Field (Q-MDF): A Robust Method for Approximating and Discretizing Neural Medial Axis
The medial axis, a lower-dimensional shape descriptor, plays an important role in the field of digital geometry processing. Despite its importance, robust computation of the medial axis transform from diverse inputs, especially point clouds with defects, remains a significant challenge. In this paper, we tackle the challenge by proposing a new implicit method that diverges from mainstream explicit medial axis computation techniques. Our key technical insight is the difference between the signed distance field (SDF) and the medial field (MF) of a solid shape is the unsigned distance field (UDF) of the shape's medial axis. This allows for formulating medial axis computation as an implicit reconstruction problem. Utilizing a modified double covering method, we extract the medial axis as the zero level-set of the UDF. Extensive experiments show that our method has enhanced accuracy and robustness in learning compact medial axis transform from thorny meshes and point clouds compared to existing methods.
☆ Scaling Robot Policy Learning via Zero-Shot Labeling with Foundation Models
A central challenge towards developing robots that can relate human language to their perception and actions is the scarcity of natural language annotations in diverse robot datasets. Moreover, robot policies that follow natural language instructions are typically trained on either templated language or expensive human-labeled instructions, hindering their scalability. To this end, we introduce NILS: Natural language Instruction Labeling for Scalability. NILS automatically labels uncurated, long-horizon robot data at scale in a zero-shot manner without any human intervention. NILS combines pretrained vision-language foundation models in order to detect objects in a scene, detect object-centric changes, segment tasks from large datasets of unlabelled interaction data and ultimately label behavior datasets. Evaluations on BridgeV2, Fractal, and a kitchen play dataset show that NILS can autonomously annotate diverse robot demonstrations of unlabeled and unstructured datasets while alleviating several shortcomings of crowdsourced human annotations, such as low data quality and diversity. We use NILS to label over 115k trajectories obtained from over 430 hours of robot data. We open-source our auto-labeling code and generated annotations on our website: http://robottasklabeling.github.io.
comment: Project Website at https://robottasklabeling.github.io/
☆ AdaDiffSR: Adaptive Region-aware Dynamic Acceleration Diffusion Model for Real-World Image Super-Resolution ECCV2024
Diffusion models (DMs) have shown promising results on single-image super-resolution and other image-to-image translation tasks. Benefiting from more computational resources and longer inference times, they are able to yield more realistic images. Existing DMs-based super-resolution methods try to achieve an overall average recovery over all regions via iterative refinement, ignoring the consideration that different input image regions require different timesteps to reconstruct. In this work, we notice that previous DMs-based super-resolution methods suffer from wasting computational resources to reconstruct invisible details. To further improve the utilization of computational resources, we propose AdaDiffSR, a DMs-based SR pipeline with dynamic timesteps sampling strategy (DTSS). Specifically, by introducing the multi-metrics latent entropy module (MMLE), we can achieve dynamic perception of the latent spatial information gain during the denoising process, thereby guiding the dynamic selection of the timesteps. In addition, we adopt a progressive feature injection module (PFJ), which dynamically injects the original image features into the denoising process based on the current information gain, so as to generate images with both fidelity and realism. Experiments show that our AdaDiffSR achieves comparable performance over current state-of-the-art DMs-based SR methods while consuming less computational resources and inference time on both synthetic and real-world datasets.
comment: 18 pages, 6 figures, ECCV2024 accepted
☆ VISAGE: Video Synthesis using Action Graphs for Surgery MICCAI 2024
Surgical data science (SDS) is a field that analyzes patient data before, during, and after surgery to improve surgical outcomes and skills. However, surgical data is scarce, heterogeneous, and complex, which limits the applicability of existing machine learning methods. In this work, we introduce the novel task of future video generation in laparoscopic surgery. This task can augment and enrich the existing surgical data and enable various applications, such as simulation, analysis, and robot-aided surgery. Ultimately, it involves not only understanding the current state of the operation but also accurately predicting the dynamic and often unpredictable nature of surgical procedures. Our proposed method, VISAGE (VIdeo Synthesis using Action Graphs for Surgery), leverages the power of action scene graphs to capture the sequential nature of laparoscopic procedures and utilizes diffusion models to synthesize temporally coherent video sequences. VISAGE predicts the future frames given only a single initial frame, and the action graph triplets. By incorporating domain-specific knowledge through the action graph, VISAGE ensures the generated videos adhere to the expected visual and motion patterns observed in real laparoscopic procedures. The results of our experiments demonstrate high-fidelity video generation for laparoscopy procedures, which enables various applications in SDS.
comment: Accepted at MICCAI 2024 Embodied AI and Robotics for HealTHcare (EARTH) Workshop
☆ Efficient Neural Implicit Representation for 3D Human Reconstruction
High-fidelity digital human representations are increasingly in demand in the digital world, particularly for interactive telepresence, AR/VR, 3D graphics, and the rapidly evolving metaverse. Even though they work well in small spaces, conventional methods for reconstructing 3D human motion frequently require the use of expensive hardware and have high processing costs. This study presents HumanAvatar, an innovative approach that efficiently reconstructs precise human avatars from monocular video sources. At the core of our methodology, we integrate the pre-trained HuMoR, a model celebrated for its proficiency in human motion estimation. This is adeptly fused with the cutting-edge neural radiance field technology, Instant-NGP, and the state-of-the-art articulated model, Fast-SNARF, to enhance the reconstruction fidelity and speed. By combining these two technologies, a system is created that can render quickly and effectively while also providing estimation of human pose parameters that are unmatched in accuracy. We have enhanced our system with an advanced posture-sensitive space reduction technique, which optimally balances rendering quality with computational efficiency. In our detailed experimental analysis using both artificial and real-world monocular videos, we establish the advanced performance of our approach. HumanAvatar consistently equals or surpasses contemporary leading-edge reconstruction techniques in quality. Furthermore, it achieves these complex reconstructions in minutes, a fraction of the time typically required by existing methods. Our models achieve a training speed that is 110X faster than that of State-of-The-Art (SoTA) NeRF-based models. Our technique performs noticeably better than SoTA dynamic human NeRF methods if given an identical runtime limit. HumanAvatar can provide effective visuals after only 30 seconds of training.
☆ Emotion Recognition with Facial Attention and Objective Activation Functions
In this paper, we study the effect of introducing channel and spatial attention mechanisms, namely SEN-Net, ECA-Net, and CBAM, to existing CNN vision-based models such as VGGNet, ResNet, and ResNetV2 to perform the Facial Emotion Recognition task. We show that not only attention can significantly improve the performance of these models but also that combining them with a different activation function can further help increase the performance of these models.
☆ New Insight in Cervical Cancer Diagnosis Using Convolution Neural Network Architecture
The Pap smear is a screening method for early cervical cancer diagnosis. The selection of the right optimizer in the convolutional neural network (CNN) model is key to the success of the CNN in image classification, including the classification of cervical cancer Pap smear images. In this study, stochastic gradient descent (SGD), RMSprop, Adam, AdaGrad, AdaDelta, Adamax, and Nadam optimizers were used to classify cervical cancer Pap smear images from the SipakMed dataset. Resnet-18, Resnet-34, and VGG-16 are the CNN architectures used in this study, and each architecture uses a transfer-learning model. Based on the test results, we conclude that the transfer learning model performs better on all CNNs and optimization techniques and that in the transfer learning model, the optimization has little influence on the training of the model. Adamax, with accuracy values of 72.8% and 66.8%, had the best accuracy for the VGG-16 and Resnet-18 architectures, respectively. Resnet-34 had 54.0%. This is 0.034% lower than Nadam. Overall, Adamax is a suitable optimizer for CNN in cervical cancer classification on Resnet-18, Resnet-34, and VGG-16 architectures. This study provides new insights into the configuration of CNN models for Pap smear image analysis.
☆ YOLO-Vehicle-Pro: A Cloud-Edge Collaborative Framework for Object Detection in Autonomous Driving under Adverse Weather Conditions
With the rapid advancement of autonomous driving technology, efficient and accurate object detection capabilities have become crucial factors in ensuring the safety and reliability of autonomous driving systems. However, in low-visibility environments such as hazy conditions, the performance of traditional object detection algorithms often degrades significantly, failing to meet the demands of autonomous driving. To address this challenge, this paper proposes two innovative deep learning models: YOLO-Vehicle and YOLO-Vehicle-Pro. YOLO-Vehicle is an object detection model tailored specifically for autonomous driving scenarios, employing multimodal fusion techniques to combine image and textual information for object detection. YOLO-Vehicle-Pro builds upon this foundation by introducing an improved image dehazing algorithm, enhancing detection performance in low-visibility environments. In addition to model innovation, this paper also designs and implements a cloud-edge collaborative object detection system, deploying models on edge devices and offloading partial computational tasks to the cloud in complex situations. Experimental results demonstrate that on the KITTI dataset, the YOLO-Vehicle-v1s model achieved 92.1% accuracy while maintaining a detection speed of 226 FPS and an inference time of 12ms, meeting the real-time requirements of autonomous driving. When processing hazy images, the YOLO-Vehicle-Pro model achieved a high accuracy of 82.3% mAP@50 on the Foggy Cityscapes dataset while maintaining a detection speed of 43 FPS.
☆ YOLOv11: An Overview of the Key Architectural Enhancements
This study presents an architectural analysis of YOLOv11, the latest iteration in the YOLO (You Only Look Once) series of object detection models. We examine the models architectural innovations, including the introduction of the C3k2 (Cross Stage Partial with kernel size 2) block, SPPF (Spatial Pyramid Pooling - Fast), and C2PSA (Convolutional block with Parallel Spatial Attention) components, which contribute in improving the models performance in several ways such as enhanced feature extraction. The paper explores YOLOv11's expanded capabilities across various computer vision tasks, including object detection, instance segmentation, pose estimation, and oriented object detection (OBB). We review the model's performance improvements in terms of mean Average Precision (mAP) and computational efficiency compared to its predecessors, with a focus on the trade-off between parameter count and accuracy. Additionally, the study discusses YOLOv11's versatility across different model sizes, from nano to extra-large, catering to diverse application needs from edge devices to high-performance computing environments. Our research provides insights into YOLOv11's position within the broader landscape of object detection and its potential impact on real-time computer vision applications.
☆ Continual Learning on a Data Diet
Continual Learning (CL) methods usually learn from all available data. However, this is not the case in human cognition which efficiently focuses on key experiences while disregarding the redundant information. Similarly, not all data points in a dataset have equal potential; some can be more informative than others. This disparity may significantly impact the performance, as both the quality and quantity of samples directly influence the model's generalizability and efficiency. Drawing inspiration from this, we explore the potential of learning from important samples and present an empirical study for evaluating coreset selection techniques in the context of CL to stimulate research in this unexplored area. We train different continual learners on increasing amounts of selected samples and investigate the learning-forgetting dynamics by shedding light on the underlying mechanisms driving their improved stability-plasticity balance. We present several significant observations: learning from selectively chosen samples (i) enhances incremental accuracy, (ii) improves knowledge retention of previous tasks, and (iii) refines learned representations. This analysis contributes to a deeper understanding of selective learning strategies in CL scenarios.
comment: 18 pages, 6 figures
☆ Longitudinal Causal Image Synthesis
Clinical decision-making relies heavily on causal reasoning and longitudinal analysis. For example, for a patient with Alzheimer's disease (AD), how will the brain grey matter atrophy in a year if intervened on the A-beta level in cerebrospinal fluid? The answer is fundamental to diagnosis and follow-up treatment. However, this kind of inquiry involves counterfactual medical images which can not be acquired by instrumental or correlation-based image synthesis models. Yet, such queries require counterfactual medical images, not obtainable through standard image synthesis models. Hence, a causal longitudinal image synthesis (CLIS) method, enabling the synthesis of such images, is highly valuable. However, building a CLIS model confronts three primary yet unmet challenges: mismatched dimensionality between high-dimensional images and low-dimensional tabular variables, inconsistent collection intervals of follow-up data, and inadequate causal modeling capability of existing causal graph methods for image data. In this paper, we established a tabular-visual causal graph (TVCG) for CLIS overcoming these challenges through a novel integration of generative imaging, continuous-time modeling, and structural causal models combined with a neural network. We train our CLIS based on the ADNI dataset and evaluate it on two other AD datasets, which illustrate the outstanding yet controllable quality of the synthesized images and the contributions of synthesized MRI to the characterization of AD progression, substantiating the reliability and utility in clinics.
☆ Deep Generative Models for 3D Medical Image Synthesis
Deep generative modeling has emerged as a powerful tool for synthesizing realistic medical images, driving advances in medical image analysis, disease diagnosis, and treatment planning. This chapter explores various deep generative models for 3D medical image synthesis, with a focus on Variational Autoencoders (VAEs), Generative Adversarial Networks (GANs), and Denoising Diffusion Models (DDMs). We discuss the fundamental principles, recent advances, as well as strengths and weaknesses of these models and examine their applications in clinically relevant problems, including unconditional and conditional generation tasks like image-to-image translation and image reconstruction. We additionally review commonly used evaluation metrics for assessing image fidelity, diversity, utility, and privacy and provide an overview of current challenges in the field.
☆ Surgical Scene Segmentation by Transformer With Asymmetric Feature Enhancement
Surgical scene segmentation is a fundamental task for robotic-assisted laparoscopic surgery understanding. It often contains various anatomical structures and surgical instruments, where similar local textures and fine-grained structures make the segmentation a difficult task. Vision-specific transformer method is a promising way for surgical scene understanding. However, there are still two main challenges. Firstly, the absence of inner-patch information fusion leads to poor segmentation performance. Secondly, the specific characteristics of anatomy and instruments are not specifically modeled. To tackle the above challenges, we propose a novel Transformer-based framework with an Asymmetric Feature Enhancement module (TAFE), which enhances local information and then actively fuses the improved feature pyramid into the embeddings from transformer encoders by a multi-scale interaction attention strategy. The proposed method outperforms the SOTA methods in several different surgical segmentation tasks and additionally proves its ability of fine-grained structure recognition. Code is available at https://github.com/cyuan-sjtu/ViT-asym.
☆ MIA-DPO: Multi-Image Augmented Direct Preference Optimization For Large Vision-Language Models
Visual preference alignment involves training Large Vision-Language Models (LVLMs) to predict human preferences between visual inputs. This is typically achieved by using labeled datasets of chosen/rejected pairs and employing optimization algorithms like direct preference optimization (DPO). Existing visual alignment methods, primarily designed for single-image scenarios, struggle to effectively handle the complexity of multi-image tasks due to the scarcity of diverse training data and the high cost of annotating chosen/rejected pairs. We present Multi-Image Augmented Direct Preference Optimization (MIA-DPO), a visual preference alignment approach that effectively handles multi-image inputs. MIA-DPO mitigates the scarcity of diverse multi-image training data by extending single-image data with unrelated images arranged in grid collages or pic-in-pic formats, significantly reducing the costs associated with multi-image data annotations. Our observation reveals that attention values of LVLMs vary considerably across different images. We use attention values to identify and filter out rejected responses the model may have mistakenly focused on. Our attention-aware selection for constructing the chosen/rejected pairs without relying on (i) human annotation, (ii) extra data, and (iii) external models or APIs. MIA-DPO is compatible with various architectures and outperforms existing methods on five multi-image benchmarks, achieving an average performance boost of 3.0% on LLaVA-v1.5 and 4.3% on the recent InternLM-XC2.5. Moreover, MIA-DPO has a minimal effect on the model's ability to understand single images.
comment: Project URL: https://github.com/Liuziyu77/MIA-DPO
☆ Bridging the Gaps: Utilizing Unlabeled Face Recognition Datasets to Boost Semi-Supervised Facial Expression Recognition
In recent years, Facial Expression Recognition (FER) has gained increasing attention. Most current work focuses on supervised learning, which requires a large amount of labeled and diverse images, while FER suffers from the scarcity of large, diverse datasets and annotation difficulty. To address these problems, we focus on utilizing large unlabeled Face Recognition (FR) datasets to boost semi-supervised FER. Specifically, we first perform face reconstruction pre-training on large-scale facial images without annotations to learn features of facial geometry and expression regions, followed by two-stage fine-tuning on FER datasets with limited labels. In addition, to further alleviate the scarcity of labeled and diverse images, we propose a Mixup-based data augmentation strategy tailored for facial images, and the loss weights of real and virtual images are determined according to the intersection-over-union (IoU) of the faces in the two images. Experiments on RAF-DB, AffectNet, and FERPlus show that our method outperforms existing semi-supervised FER methods and achieves new state-of-the-art performance. Remarkably, with only 5%, 25% training sets,our method achieves 64.02% on AffectNet,and 88.23% on RAF-DB, which is comparable to fully supervised state-of-the-art methods. Codes will be made publicly available at https://github.com/zhelishisongjie/SSFER.
☆ ImDy: Human Inverse Dynamics from Imitated Observations
Inverse dynamics (ID), which aims at reproducing the driven torques from human kinematic observations, has been a critical tool for gait analysis. However, it is hindered from wider application to general motion due to its limited scalability. Conventional optimization-based ID requires expensive laboratory setups, restricting its availability. To alleviate this problem, we propose to exploit the recently progressive human motion imitation algorithms to learn human inverse dynamics in a data-driven manner. The key insight is that the human ID knowledge is implicitly possessed by motion imitators, though not directly applicable. In light of this, we devise an efficient data collection pipeline with state-of-the-art motion imitation algorithms and physics simulators, resulting in a large-scale human inverse dynamics benchmark as Imitated Dynamics (ImDy). ImDy contains over 150 hours of motion with joint torque and full-body ground reaction force data. With ImDy, we train a data-driven human inverse dynamics solver ImDyS(olver) in a fully supervised manner, which conducts ID and ground reaction force estimation simultaneously. Experiments on ImDy and real-world data demonstrate the impressive competency of ImDyS in human inverse dynamics and ground reaction force estimation. Moreover, the potential of ImDy(-S) as a fundamental motion analysis tool is exhibited with downstream applications. The project page is https://foruck.github.io/ImDy/.
comment: Yong-Lu Li and Cewu Lu are the corresponding authors
☆ Towards Effective Data-Free Knowledge Distillation via Diverse Diffusion Augmentation
Data-free knowledge distillation (DFKD) has emerged as a pivotal technique in the domain of model compression, substantially reducing the dependency on the original training data. Nonetheless, conventional DFKD methods that employ synthesized training data are prone to the limitations of inadequate diversity and discrepancies in distribution between the synthesized and original datasets. To address these challenges, this paper introduces an innovative approach to DFKD through diverse diffusion augmentation (DDA). Specifically, we revise the paradigm of common data synthesis in DFKD to a composite process through leveraging diffusion models subsequent to data synthesis for self-supervised augmentation, which generates a spectrum of data samples with similar distributions while retaining controlled variations. Furthermore, to mitigate excessive deviation in the embedding space, we introduce an image filtering technique grounded in cosine similarity to maintain fidelity during the knowledge distillation process. Comprehensive experiments conducted on CIFAR-10, CIFAR-100, and Tiny-ImageNet datasets showcase the superior performance of our method across various teacher-student network configurations, outperforming the contemporary state-of-the-art DFKD methods. Code will be available at:https://github.com/SLGSP/DDA.
☆ PlantCamo: Plant Camouflage Detection
Camouflaged Object Detection (COD) aims to detect objects with camouflaged properties. Although previous studies have focused on natural (animals and insects) and unnatural (artistic and synthetic) camouflage detection, plant camouflage has been neglected. However, plant camouflage plays a vital role in natural camouflage. Therefore, this paper introduces a new challenging problem of Plant Camouflage Detection (PCD). To address this problem, we introduce the PlantCamo dataset, which comprises 1,250 images with camouflaged plants representing 58 object categories in various natural scenes. To investigate the current status of plant camouflage detection, we conduct a large-scale benchmark study using 20+ cutting-edge COD models on the proposed dataset. Due to the unique characteristics of plant camouflage, including holes and irregular borders, we developed a new framework, named PCNet, dedicated to PCD. Our PCNet surpasses performance thanks to its multi-scale global feature enhancement and refinement. Finally, we discuss the potential applications and insights, hoping this work fills the gap in fine-grained COD research and facilitates further intelligent ecology research. All resources will be available on https://github.com/yjybuaa/PlantCamo.
☆ How to Continually Adapt Text-to-Image Diffusion Models for Flexible Customization? NeurIPS2024
Custom diffusion models (CDMs) have attracted widespread attention due to their astonishing generative ability for personalized concepts. However, most existing CDMs unreasonably assume that personalized concepts are fixed and cannot change over time. Moreover, they heavily suffer from catastrophic forgetting and concept neglect on old personalized concepts when continually learning a series of new concepts. To address these challenges, we propose a novel Concept-Incremental text-to-image Diffusion Model (CIDM), which can resolve catastrophic forgetting and concept neglect to learn new customization tasks in a concept-incremental manner. Specifically, to surmount the catastrophic forgetting of old concepts, we develop a concept consolidation loss and an elastic weight aggregation module. They can explore task-specific and task-shared knowledge during training, and aggregate all low-rank weights of old concepts based on their contributions during inference. Moreover, in order to address concept neglect, we devise a context-controllable synthesis strategy that leverages expressive region features and noise estimation to control the contexts of generated images according to user conditions. Experiments validate that our CIDM surpasses existing custom diffusion models. The source codes are available at https://github.com/JiahuaDong/CIFC.
comment: Accepted to NeurIPS2024
☆ Double Banking on Knowledge: Customized Modulation and Prototypes for Multi-Modality Semi-supervised Medical Image Segmentation
Multi-modality (MM) semi-supervised learning (SSL) based medical image segmentation has recently gained increasing attention for its ability to utilize MM data and reduce reliance on labeled images. However, current methods face several challenges: (1) Complex network designs hinder scalability to scenarios with more than two modalities. (2) Focusing solely on modality-invariant representation while neglecting modality-specific features, leads to incomplete MM learning. (3) Leveraging unlabeled data with generative methods can be unreliable for SSL. To address these problems, we propose Double Bank Dual Consistency (DBDC), a novel MM-SSL approach for medical image segmentation. To address challenge (1), we propose a modality all-in-one segmentation network that accommodates data from any number of modalities, removing the limitation on modality count. To address challenge (2), we design two learnable plug-in banks, Modality-Level Modulation bank (MLMB) and Modality-Level Prototype (MLPB) bank, to capture both modality-invariant and modality-specific knowledge. These banks are updated using our proposed Modality Prototype Contrastive Learning (MPCL). Additionally, we design Modality Adaptive Weighting (MAW) to dynamically adjust learning weights for each modality, ensuring balanced MM learning as different modalities learn at different rates. Finally, to address challenge (3), we introduce a Dual Consistency (DC) strategy that enforces consistency at both the image and feature levels without relying on generative methods. We evaluate our method on a 2-to-4 modality segmentation task using three open-source datasets, and extensive experiments show that our method outperforms state-of-the-art approaches.
☆ BlurryScope: a cost-effective and compact scanning microscope for automated HER2 scoring using deep learning on blurry image data
We developed a rapid scanning optical microscope, termed "BlurryScope", that leverages continuous image acquisition and deep learning to provide a cost-effective and compact solution for automated inspection and analysis of tissue sections. BlurryScope integrates specialized hardware with a neural network-based model to quickly process motion-blurred histological images and perform automated pathology classification. This device offers comparable speed to commercial digital pathology scanners, but at a significantly lower price point and smaller size/weight, making it ideal for fast triaging in small clinics, as well as for resource-limited settings. To demonstrate the proof-of-concept of BlurryScope, we implemented automated classification of human epidermal growth factor receptor 2 (HER2) scores on immunohistochemically (IHC) stained breast tissue sections, achieving concordant results with those obtained from a high-end digital scanning microscope. We evaluated this approach by scanning HER2-stained tissue microarrays (TMAs) at a continuous speed of 5 mm/s, which introduces bidirectional motion blur artifacts. These compromised images were then used to train our network models. Using a test set of 284 unique patient cores, we achieved blind testing accuracies of 79.3% and 89.7% for 4-class (0, 1+, 2+, 3+) and 2-class (0/1+ , 2+/3+) HER2 score classification, respectively. BlurryScope automates the entire workflow, from image scanning to stitching and cropping of regions of interest, as well as HER2 score classification. We believe BlurryScope has the potential to enhance the current pathology infrastructure in resource-scarce environments, save diagnostician time and bolster cancer identification and classification across various clinical environments.
comment: 18 Pages, 6 Figures
☆ Unsupervised Low-dose CT Reconstruction with One-way Conditional Normalizing Flows
Deep-learning methods have shown promising performance for low-dose computed tomography (LDCT) reconstruction. However, supervised methods face the problem of lacking labeled data in clinical scenarios, and the CNN-based unsupervised denoising methods would cause excessive smoothing in the reconstructed image. Recently, the normalizing flows (NFs) based methods have shown advantages in producing detail-rich images and avoiding over-smoothing, however, there are still issues: (1) Although the alternating optimization in the data and latent space can well utilize the regularization and generation capabilities of NFs, the current two-way transformation strategy of noisy images and latent variables would cause detail loss and secondary artifacts; and (2) Training NFs on high-resolution CT images is hard due to huge computation. Though using conditional normalizing flows (CNFs) to learn conditional probability can reduce the computational burden, current methods require labeled data for conditionalization, and the unsupervised CNFs-based LDCT reconstruction remains a problem. To tackle these problems, we propose a novel CNFs-based unsupervised LDCT iterative reconstruction algorithm. It employs strict one-way transformation when performing alternating optimization in the dual spaces, thus effectively avoiding the problems of detail loss and secondary artifacts. By proposing a novel unsupervised conditionalization strategy, we train CNFs on high-resolution CT images, thus achieving fast and high-quality unsupervised reconstruction. Experiments on different datasets suggest that the performance of the proposed algorithm could surpass some state-of-the-art unsupervised and even supervised methods.
☆ OVT-B: A New Large-Scale Benchmark for Open-Vocabulary Multi-Object Tracking NeurIPS 2024
Open-vocabulary object perception has become an important topic in artificial intelligence, which aims to identify objects with novel classes that have not been seen during training. Under this setting, open-vocabulary object detection (OVD) in a single image has been studied in many literature. However, open-vocabulary object tracking (OVT) from a video has been studied less, and one reason is the shortage of benchmarks. In this work, we have built a new large-scale benchmark for open-vocabulary multi-object tracking namely OVT-B. OVT-B contains 1,048 categories of objects and 1,973 videos with 637,608 bounding box annotations, which is much larger than the sole open-vocabulary tracking dataset, i.e., OVTAO-val dataset (200+ categories, 900+ videos). The proposed OVT-B can be used as a new benchmark to pave the way for OVT research. We also develop a simple yet effective baseline method for OVT. It integrates the motion features for object tracking, which is an important feature for MOT but is ignored in previous OVT methods. Experimental results have verified the usefulness of the proposed benchmark and the effectiveness of our method. We have released the benchmark to the public at https://github.com/Coo1Sea/OVT-B-Dataset.
comment: 15 pages, 6 figures, accepted at NeurIPS 2024 Dataset and Benchmark Track
☆ Diffusion Priors for Variational Likelihood Estimation and Image Denoising NeurIPS2024
Real-world noise removal is crucial in low-level computer vision. Due to the remarkable generation capabilities of diffusion models, recent attention has shifted towards leveraging diffusion priors for image restoration tasks. However, existing diffusion priors-based methods either consider simple noise types or rely on approximate posterior estimation, limiting their effectiveness in addressing structured and signal-dependent noise commonly found in real-world images. In this paper, we build upon diffusion priors and propose adaptive likelihood estimation and MAP inference during the reverse diffusion process to tackle real-world noise. We introduce an independent, non-identically distributed likelihood combined with the noise precision (inverse variance) prior and dynamically infer the precision posterior using variational Bayes during the generation process. Meanwhile, we rectify the estimated noise variance through local Gaussian convolution. The final denoised image is obtained by propagating intermediate MAP solutions that balance the updated likelihood and diffusion prior. Additionally, we explore the local diffusion prior inherent in low-resolution diffusion models, enabling direct handling of high-resolution noisy images. Extensive experiments and analyses on diverse real-world datasets demonstrate the effectiveness of our method. Code is available at https://github.com/HUST-Tan/DiffusionVI.
comment: Accepted by NeurIPS2024 as Spotlight
☆ PathMoCo: A Novel Framework to Improve Feature Embedding in Self-supervised Contrastive Learning for Histopathological Images
Self-supervised contrastive learning has become a cornerstone in various areas, particularly histopathological image analysis. Image augmentation plays a crucial role in self-supervised contrastive learning, as it generates variations in image samples. However, traditional image augmentation techniques often overlook the unique characteristics of histopathological images. In this paper, we propose a new histopathology-specific image augmentation method called stain reconstruction augmentation (SRA). We integrate our SRA with MoCo v3, a leading model in self-supervised contrastive learning, along with our additional contrastive loss terms, and call the new model PathMoCo. We demonstrate that our PathMoCo always outperforms the standard MoCo v3 across various downstream tasks and achieves comparable or superior performance to other foundation models pre-trained on significantly larger histopathology datasets.
☆ HCDN: A Change Detection Network for Construction Housekeeping Using Feature Fusion and Large Vision Models
Workplace safety has received increasing attention as millions of workers worldwide suffer from work-related accidents. Despite poor housekeeping is a significant contributor to construction accidents, there remains a significant lack of technological research focused on improving housekeeping practices in construction sites. Recognizing and locating poor housekeeping in a dynamic construction site is an important task that can be improved through computer vision approaches. Despite advances in AI and computer vision, existing methods for detecting poor housekeeping conditions face many challenges, including limited explanations, lack of locating of poor housekeeping, and lack of annotated datasets. On the other hand, change detection which aims to detect the changed environmental conditions (e.g., changing from good to poor housekeeping) and 'where' the change has occurred (e.g., location of objects causing poor housekeeping), has not been explored to the problem of housekeeping management. To address these challenges, we propose the Housekeeping Change Detection Network (HCDN), an advanced change detection neural network that integrates a feature fusion module and a large vision model, achieving state-of-the-art performance. Additionally, we introduce the approach to establish a novel change detection dataset (named Housekeeping-CCD) focused on housekeeping in construction sites, along with a housekeeping segmentation dataset. Our contributions include significant performance improvements compared to existing methods, providing an effective tool for enhancing construction housekeeping and safety. To promote further development, we share our source code and trained models for global researchers: https://github.com/NUS-DBE/Housekeeping-CD.
☆ PLGS: Robust Panoptic Lifting with 3D Gaussian Splatting
Previous methods utilize the Neural Radiance Field (NeRF) for panoptic lifting, while their training and rendering speed are unsatisfactory. In contrast, 3D Gaussian Splatting (3DGS) has emerged as a prominent technique due to its rapid training and rendering speed. However, unlike NeRF, the conventional 3DGS may not satisfy the basic smoothness assumption as it does not rely on any parameterized structures to render (e.g., MLPs). Consequently, the conventional 3DGS is, in nature, more susceptible to noisy 2D mask supervision. In this paper, we propose a new method called PLGS that enables 3DGS to generate consistent panoptic segmentation masks from noisy 2D segmentation masks while maintaining superior efficiency compared to NeRF-based methods. Specifically, we build a panoptic-aware structured 3D Gaussian model to introduce smoothness and design effective noise reduction strategies. For the semantic field, instead of initialization with structure from motion, we construct reliable semantic anchor points to initialize the 3D Gaussians. We then use these anchor points as smooth regularization during training. Additionally, we present a self-training approach using pseudo labels generated by merging the rendered masks with the noisy masks to enhance the robustness of PLGS. For the instance field, we project the 2D instance masks into 3D space and match them with oriented bounding boxes to generate cross-view consistent instance masks for supervision. Experiments on various benchmarks demonstrate that our method outperforms previous state-of-the-art methods in terms of both segmentation quality and speed.
☆ Bilateral Hippocampi Segmentation in Low Field MRIs Using Mutual Feature Learning via Dual-Views
Accurate hippocampus segmentation in brain MRI is critical for studying cognitive and memory functions and diagnosing neurodevelopmental disorders. While high-field MRIs provide detailed imaging, low-field MRIs are more accessible and cost-effective, which eliminates the need for sedation in children, though they often suffer from lower image quality. In this paper, we present a novel deep-learning approach for the automatic segmentation of bilateral hippocampi in low-field MRIs. Extending recent advancements in infant brain segmentation to underserved communities through the use of low-field MRIs ensures broader access to essential diagnostic tools, thereby supporting better healthcare outcomes for all children. Inspired by our previous work, Co-BioNet, the proposed model employs a dual-view structure to enable mutual feature learning via high-frequency masking, enhancing segmentation accuracy by leveraging complementary information from different perspectives. Extensive experiments demonstrate that our method provides reliable segmentation outcomes for hippocampal analysis in low-resource settings. The code is publicly available at: https://github.com/himashi92/LoFiHippSeg.
☆ Enhancing Multimodal Medical Image Classification using Cross-Graph Modal Contrastive Learning
The classification of medical images is a pivotal aspect of disease diagnosis, often enhanced by deep learning techniques. However, traditional approaches typically focus on unimodal medical image data, neglecting the integration of diverse non-image patient data. This paper proposes a novel Cross-Graph Modal Contrastive Learning (CGMCL) framework for multimodal medical image classification. The model effectively integrates both image and non-image data by constructing cross-modality graphs and leveraging contrastive learning to align multimodal features in a shared latent space. An inter-modality feature scaling module further optimizes the representation learning process by reducing the gap between heterogeneous modalities. The proposed approach is evaluated on two datasets: a Parkinson's disease (PD) dataset and a public melanoma dataset. Results demonstrate that CGMCL outperforms conventional unimodal methods in accuracy, interpretability, and early disease prediction. Additionally, the method shows superior performance in multi-class melanoma classification. The CGMCL framework provides valuable insights into medical image classification while offering improved disease interpretability and predictive capabilities.
☆ Unsupervised Domain Adaptation for Action Recognition via Self-Ensembling and Conditional Embedding Alignment
Recent advancements in deep learning-based wearable human action recognition (wHAR) have improved the capture and classification of complex motions, but adoption remains limited due to the lack of expert annotations and domain discrepancies from user variations. Limited annotations hinder the model's ability to generalize to out-of-distribution samples. While data augmentation can improve generalizability, unsupervised augmentation techniques must be applied carefully to avoid introducing noise. Unsupervised domain adaptation (UDA) addresses domain discrepancies by aligning conditional distributions with labeled target samples, but vanilla pseudo-labeling can lead to error propagation. To address these challenges, we propose $\mu$DAR, a novel joint optimization architecture comprised of three functions: (i) consistency regularizer between augmented samples to improve model classification generalizability, (ii) temporal ensemble for robust pseudo-label generation and (iii) conditional distribution alignment to improve domain generalizability. The temporal ensemble works by aggregating predictions from past epochs to smooth out noisy pseudo-label predictions, which are then used in the conditional distribution alignment module to minimize kernel-based class-wise conditional maximum mean discrepancy ($k$CMMD) between the source and target feature space to learn a domain invariant embedding. The consistency-regularized augmentations ensure that multiple augmentations of the same sample share the same labels; this results in (a) strong generalization with limited source domain samples and (b) consistent pseudo-label generation in target samples. The novel integration of these three modules in $\mu$DAR results in a range of $\approx$ 4-12% average macro-F1 score improvement over six state-of-the-art UDA methods in four benchmark wHAR datasets
comment: This work has been accepted to the Proceedings of the IEEE International Conference on Data Mining, 2024
☆ GenDP: 3D Semantic Fields for Category-Level Generalizable Diffusion Policy
Diffusion-based policies have shown remarkable capability in executing complex robotic manipulation tasks but lack explicit characterization of geometry and semantics, which often limits their ability to generalize to unseen objects and layouts. To enhance the generalization capabilities of Diffusion Policy, we introduce a novel framework that incorporates explicit spatial and semantic information via 3D semantic fields. We generate 3D descriptor fields from multi-view RGBD observations with large foundational vision models, then compare these descriptor fields against reference descriptors to obtain semantic fields. The proposed method explicitly considers geometry and semantics, enabling strong generalization capabilities in tasks requiring category-level generalization, resolving geometric ambiguities, and attention to subtle geometric details. We evaluate our method across eight tasks involving articulated objects and instances with varying shapes and textures from multiple object categories. Our method demonstrates its effectiveness by increasing Diffusion Policy's average success rate on unseen instances from 20% to 93%. Additionally, we provide a detailed analysis and visualization to interpret the sources of performance gain and explain how our method can generalize to novel instances.
comment: Accepted to Conference on Robot Learning (CoRL 2024). Project Page: https://robopil.github.io/GenDP/
☆ Which Client is Reliable?: A Reliable and Personalized Prompt-based Federated Learning for Medical Image Question Answering
Conventional medical artificial intelligence (AI) models face barriers in clinical application and ethical issues owing to their inability to handle the privacy-sensitive characteristics of medical data. We present a novel personalized federated learning (pFL) method for medical visual question answering (VQA) models, addressing privacy reliability challenges in the medical domain. Our method introduces learnable prompts into a Transformer architecture to efficiently train it on diverse medical datasets without massive computational costs. Then we introduce a reliable client VQA model that incorporates Dempster-Shafer evidence theory to quantify uncertainty in predictions, enhancing the model's reliability. Furthermore, we propose a novel inter-client communication mechanism that uses maximum likelihood estimation to balance accuracy and uncertainty, fostering efficient integration of insights across clients.
♻ ☆ Pruning By Explaining Revisited: Optimizing Attribution Methods to Prune CNNs and Transformers ECCV 2024
To solve ever more complex problems, Deep Neural Networks are scaled to billions of parameters, leading to huge computational costs. An effective approach to reduce computational requirements and increase efficiency is to prune unnecessary components of these often over-parameterized networks. Previous work has shown that attribution methods from the field of eXplainable AI serve as effective means to extract and prune the least relevant network components in a few-shot fashion. We extend the current state by proposing to explicitly optimize hyperparameters of attribution methods for the task of pruning, and further include transformer-based networks in our analysis. Our approach yields higher model compression rates of large transformer- and convolutional architectures (VGG, ResNet, ViT) compared to previous works, while still attaining high performance on ImageNet classification tasks. Here, our experiments indicate that transformers have a higher degree of over-parameterization compared to convolutional neural networks. Code is available at https://github.com/erfanhatefi/Pruning-by-eXplaining-in-PyTorch.
comment: Accepted as a workshop paper at ECCV 2024, 26 pages (11 pages manuscript, 3 pages references, 12 pages appendix)
♻ ☆ VILA-U: a Unified Foundation Model Integrating Visual Understanding and Generation
VILA-U is a Unified foundation model that integrates Video, Image, Language understanding and generation. Traditional visual language models (VLMs) use separate modules for understanding and generating visual content, which can lead to misalignment and increased complexity. In contrast, VILA-U employs a single autoregressive next-token prediction framework for both tasks, eliminating the need for additional components like diffusion models. This approach not only simplifies the model but also achieves near state-of-the-art performance in visual language understanding and generation. The success of VILA-U is attributed to two main factors: the unified vision tower that aligns discrete visual tokens with textual inputs during pretraining, which enhances visual perception, and autoregressive image generation can achieve similar quality as diffusion models with high-quality dataset. This allows VILA-U to perform comparably to more complex models using a fully token-based autoregressive framework.
comment: Code: https://github.com/mit-han-lab/vila-u. The first two authors contributed equally to this work
♻ ☆ JointMotion: Joint Self-Supervision for Joint Motion Prediction
We present JointMotion, a self-supervised pre-training method for joint motion prediction in self-driving vehicles. Our method jointly optimizes a scene-level objective connecting motion and environments, and an instance-level objective to refine learned representations. Scene-level representations are learned via non-contrastive similarity learning of past motion sequences and environment context. At the instance level, we use masked autoencoding to refine multimodal polyline representations. We complement this with an adaptive pre-training decoder that enables JointMotion to generalize across different environment representations, fusion mechanisms, and dataset characteristics. Notably, our method reduces the joint final displacement error of Wayformer, HPTR, and Scene Transformer models by 3\%, 8\%, and 12\%, respectively; and enables transfer learning between the Waymo Open Motion and the Argoverse 2 Motion Forecasting datasets. Code: https://github.com/kit-mrt/future-motion
comment: CoRL'24 camera-ready
♻ ☆ Telling Stories for Common Sense Zero-Shot Action Recognition ACCV 2024
Video understanding has long suffered from reliance on large labeled datasets, motivating research into zero-shot learning. Recent progress in language modeling presents opportunities to advance zero-shot video analysis, but constructing an effective semantic space relating action classes remains challenging. We address this by introducing a novel dataset, Stories, which contains rich textual descriptions for diverse action classes extracted from WikiHow articles. For each class, we extract multi-sentence narratives detailing the necessary steps, scenes, objects, and verbs that characterize the action. This contextual data enables modeling of nuanced relationships between actions, paving the way for zero-shot transfer. We also propose an approach that harnesses Stories to improve feature generation for training zero-shot classification. Without any target dataset fine-tuning, our method achieves new state-of-the-art on multiple benchmarks, improving top-1 accuracy by up to 6.1%. We believe Stories provides a valuable resource that can catalyze progress in zero-shot action recognition. The textual narratives forge connections between seen and unseen classes, overcoming the bottleneck of labeled data that has long impeded advancements in this exciting domain. The data can be found here: https://github.com/kini5gowda/Stories .
comment: Accepted in ACCV 2024!
♻ ☆ Generalizable Prompt Tuning for Vision-Language Models
Prompt tuning for vision-language models such as CLIP involves optimizing the text prompts used to generate image-text pairs for specific downstream tasks. While hand-crafted or template-based prompts are generally applicable to a wider range of unseen classes, they tend to perform poorly in downstream tasks (i.e., seen classes). Learnable soft prompts, on the other hand, often perform well in downstream tasks but lack generalizability. Additionally, prior research has predominantly concentrated on the textual modality, with very few studies attempting to explore the prompt's generalization potential from the visual modality. Keeping these limitations in mind, we investigate how to prompt tuning to obtain both a competitive downstream performance and generalization. The study shows that by treating soft and hand-crafted prompts as dual views of the textual modality, and maximizing their mutual information, we can better ensemble task-specific and general semantic information. Moreover, to generate more expressive prompts, the study introduces a class-wise augmentation from the visual modality, resulting in significant robustness to a wider range of unseen classes. Extensive evaluations on several benchmarks report that the proposed approach achieves competitive results in terms of both task-specific performance and general abilities.
comment: in progress
♻ ☆ Exploring the Adversarial Robustness of CLIP for AI-generated Image Detection
In recent years, many forensic detectors have been proposed to detect AI-generated images and prevent their use for malicious purposes. Convolutional neural networks (CNNs) have long been the dominant architecture in this field and have been the subject of intense study. However, recently proposed Transformer-based detectors have been shown to match or even outperform CNN-based detectors, especially in terms of generalization. In this paper, we study the adversarial robustness of AI-generated image detectors, focusing on Contrastive Language-Image Pretraining (CLIP)-based methods that rely on Visual Transformer (ViT) backbones and comparing their performance with CNN-based methods. We study the robustness to different adversarial attacks under a variety of conditions and analyze both numerical results and frequency-domain patterns. CLIP-based detectors are found to be vulnerable to white-box attacks just like CNN-based detectors. However, attacks do not easily transfer between CNN-based and CLIP-based methods. This is also confirmed by the different distribution of the adversarial noise patterns in the frequency domain. Overall, this analysis provides new insights into the properties of forensic detectors that can help to develop more effective strategies.
♻ ☆ Leveraging Hallucinations to Reduce Manual Prompt Dependency in Promptable Segmentation NeurIPS 2024
Promptable segmentation typically requires instance-specific manual prompts to guide the segmentation of each desired object. To minimize such a need, task-generic promptable segmentation has been introduced, which employs a single task-generic prompt to segment various images of different objects in the same task. Current methods use Multimodal Large Language Models (MLLMs) to reason detailed instance-specific prompts from a task-generic prompt for improving segmentation accuracy. The effectiveness of this segmentation heavily depends on the precision of these derived prompts. However, MLLMs often suffer hallucinations during reasoning, resulting in inaccurate prompting. While existing methods focus on eliminating hallucinations to improve a model, we argue that MLLM hallucinations can reveal valuable contextual insights when leveraged correctly, as they represent pre-trained large-scale knowledge beyond individual images. In this paper, we utilize hallucinations to mine task-related information from images and verify its accuracy for enhancing precision of the generated prompts. Specifically, we introduce an iterative Prompt-Mask Cycle generation framework (ProMaC) with a prompt generator and a mask generator.The prompt generator uses a multi-scale chain of thought prompting, initially exploring hallucinations for extracting extended contextual knowledge on a test image.These hallucinations are then reduced to formulate precise instance-specific prompts, directing the mask generator to produce masks that are consistent with task semantics by mask semantic alignment. The generated masks iteratively induce the prompt generator to focus more on task-relevant image areas and reduce irrelevant hallucinations, resulting jointly in better prompts and masks. Experiments on 5 benchmarks demonstrate the effectiveness of ProMaC. Code given in https://lwpyh.github.io/ProMaC/.
comment: NeurIPS 2024
♻ ☆ LocoMotion: Learning Motion-Focused Video-Language Representations ACCV 2024
This paper strives for motion-focused video-language representations. Existing methods to learn video-language representations use spatial-focused data, where identifying the objects and scene is often enough to distinguish the relevant caption. We instead propose LocoMotion to learn from motion-focused captions that describe the movement and temporal progression of local object motions. We achieve this by adding synthetic motions to videos and using the parameters of these motions to generate corresponding captions. Furthermore, we propose verb-variation paraphrasing to increase the caption variety and learn the link between primitive motions and high-level verbs. With this, we are able to learn a motion-focused video-language representation. Experiments demonstrate our approach is effective for a variety of downstream tasks, particularly when limited data is available for fine-tuning. Code is available: https://hazeldoughty.github.io/Papers/LocoMotion/
comment: ACCV 2024 Oral
♻ ☆ Accessible, At-Home Detection of Parkinson's Disease via Multi-task Video Analysis
Limited accessibility to neurological care leads to underdiagnosed Parkinson's Disease (PD), preventing early intervention. Existing AI-based PD detection methods primarily focus on unimodal analysis of motor or speech tasks, overlooking the multifaceted nature of the disease. To address this, we introduce a large-scale, multi-task video dataset consisting of 1102 sessions (each containing videos of finger tapping, facial expression, and speech tasks captured via webcam) from 845 participants (272 with PD). We propose a novel Uncertainty-calibrated Fusion Network (UFNet) that leverages this multimodal data to enhance diagnostic accuracy. UFNet employs independent task-specific networks, trained with Monte Carlo Dropout for uncertainty quantification, followed by self-attended fusion of features, with attention weights dynamically adjusted based on task-specific uncertainties. To ensure patient-centered evaluation, the participants were randomly split into three sets: 60% for training, 20% for model selection, and 20% for final performance evaluation. UFNet significantly outperformed single-task models in terms of accuracy, area under the ROC curve (AUROC), and sensitivity while maintaining non-inferior specificity. Withholding uncertain predictions further boosted the performance, achieving 88.0+-0.3%$ accuracy, 93.0+-0.2% AUROC, 79.3+-0.9% sensitivity, and 92.6+-0.3% specificity, at the expense of not being able to predict for 2.3+-0.3% data (+- denotes 95% confidence interval). Further analysis suggests that the trained model does not exhibit any detectable bias across sex and ethnic subgroups and is most effective for individuals aged between 50 and 80. Requiring only a webcam and microphone, our approach facilitates accessible home-based PD screening, especially in regions with limited healthcare resources.
♻ ☆ SCA: Highly Efficient Semantic-Consistent Unrestricted Adversarial Attack
Deep neural network based systems deployed in sensitive environments are vulnerable to adversarial attacks. Unrestricted adversarial attacks typically manipulate the semantic content of an image (e.g., color or texture) to create adversarial examples that are both effective and photorealistic. Recent works have utilized the diffusion inversion process to map images into a latent space, where high-level semantics are manipulated by introducing perturbations. However, they often results in substantial semantic distortions in the denoised output and suffers from low efficiency. In this study, we propose a novel framework called Semantic-Consistent Unrestricted Adversarial Attacks (SCA), which employs an inversion method to extract edit-friendly noise maps and utilizes Multimodal Large Language Model (MLLM) to provide semantic guidance throughout the process. Under the condition of rich semantic information provided by MLLM, we perform the DDPM denoising process of each step using a series of edit-friendly noise maps, and leverage DPM Solver++ to accelerate this process, enabling efficient sampling with semantic consistency. Compared to existing methods, our framework enables the efficient generation of adversarial examples that exhibit minimal discernible semantic changes. Consequently, we for the first time introduce Semantic-Consistent Adversarial Examples (SCAE). Extensive experiments and visualizations have demonstrated the high efficiency of SCA, particularly in being on average 12 times faster than the state-of-the-art attacks. Our research can further draw attention to the security of multimedia information.
♻ ☆ PnLCalib: Sports Field Registration via Points and Lines Optimization
Camera calibration in broadcast sports videos presents numerous challenges for accurate sports field registration due to multiple camera angles, varying camera parameters, and frequent occlusions of the field. Traditional search-based methods depend on initial camera pose estimates, which can struggle in non-standard positions and dynamic environments. In response, we propose an optimization-based calibration pipeline that leverages a 3D soccer field model and a predefined set of keypoints to overcome these limitations. Our method also introduces a novel refinement module that improves initial calibration by using detected field lines in a non-linear optimization process. This approach outperforms existing techniques in both multi-view and single-view 3D camera calibration tasks, while maintaining competitive performance in homography estimation. Extensive experimentation on real-world soccer datasets, including SoccerNet-Calibration, WorldCup 2014, and TS-WorldCup, highlights the robustness and accuracy of our method across diverse broadcast scenarios. Our approach offers significant improvements in camera calibration precision and reliability.
comment: Extended version of "No Bells, Just Whistles: Sports Field Registration Leveraging Geometric Properties"
♻ ☆ PixLore: A Dataset-driven Approach to Rich Image Captioning
In the domain of vision-language integration, generating detailed image captions poses a significant challenge due to the lack of curated and rich datasets. This study introduces PixLore, a novel method that leverages Querying Transformers through the fine-tuning of the BLIP-2 model using the LoRa method on a standard commercial GPU. The followed approach, which involves training on a carefully assembled dataset from state-of-the-art Computer Vision models combined and augmented by ChatGPT, addresses the question of whether intricate image understanding can be achieved with an ensemble of smaller-scale models, referred to as Knowledge Stitching. Comparative evaluations against major models such as GPT-4 and Google Bard demonstrate that PixLore-2.7B, despite having considerably fewer parameters, is rated higher than the existing State-of-the-Art models in over half of the assessments. Precisely, PixLore outperform Bard and BLIP-2, which score approximately 35.18% and 27.98% lower than PixLore in the task of image captioning. This research not only presents a groundbreaking approach but also highlights the importance of well-curated datasets in enhancing the performance of smaller models.
comment: Paper in preprint pending of publication
♻ ☆ Denoising Diffusion Models for Inpainting of Healthy Brain Tissue MICCAI
This paper is a contribution to the "BraTS 2023 Local Synthesis of Healthy Brain Tissue via Inpainting Challenge". The task of this challenge is to transform tumor tissue into healthy tissue in brain magnetic resonance (MR) images. This idea originates from the problem that MR images can be evaluated using automatic processing tools, however, many of these tools are optimized for the analysis of healthy tissue. By solving the given inpainting task, we enable the automatic analysis of images featuring lesions, and further downstream tasks. Our approach builds on denoising diffusion probabilistic models. We use a 2D model that is trained using slices in which healthy tissue was cropped out and is learned to be inpainted again. This allows us to use the ground truth healthy tissue during training. In the sampling stage, we replace the slices containing diseased tissue in the original 3D volume with the slices containing the healthy tissue inpainting. With our approach, we achieve comparable results to the competing methods. On the validation set our model achieves a mean SSIM of 0.7804, a PSNR of 20.3525 and a MSE of 0.0113. In future we plan to extend our 2D model to a 3D model, allowing to inpaint the region of interest as a whole without losing context information of neighboring slices.
comment: 12 pages, 5 figures, MICCAI challenge submission
♻ ☆ A Multimodal Fusion Network For Student Emotion Recognition Based on Transformer and Tensor Product
This paper introduces a new multi-modal model based on the Transformer architecture and tensor product fusion strategy, combining BERT's text vectors and ViT's image vectors to classify students' psychological conditions, with an accuracy of 93.65%. The purpose of the study is to accurately analyze the mental health status of students from various data sources. This paper discusses modal fusion methods, including early, late and intermediate fusion, to overcome the challenges of integrating multi-modal information. Ablation studies compare the performance of different models and fusion techniques, showing that the proposed model outperforms existing methods such as CLIP and ViLBERT in terms of accuracy and inference speed. Conclusions indicate that while this model has significant advantages in emotion recognition, its potential to incorporate other data modalities provides areas for future research.
♻ ☆ Towards Croppable Implicit Neural Representations NeurIPS 2024
Implicit Neural Representations (INRs) have peaked interest in recent years due to their ability to encode natural signals using neural networks. While INRs allow for useful applications such as interpolating new coordinates and signal compression, their black-box nature makes it difficult to modify them post-training. In this paper we explore the idea of editable INRs, and specifically focus on the widely used cropping operation. To this end, we present Local-Global SIRENs -- a novel INR architecture that supports cropping by design. Local-Global SIRENs are based on combining local and global feature extraction for signal encoding. What makes their design unique is the ability to effortlessly remove specific portions of an encoded signal, with a proportional weight decrease. This is achieved by eliminating the corresponding weights from the network, without the need for retraining. We further show how this architecture can be used to support the straightforward extension of previously encoded signals. Beyond signal editing, we examine how the Local-Global approach can accelerate training, enhance encoding of various signals, improve downstream performance, and be applied to modern INRs such as INCODE, highlighting its potential and flexibility. Code is available at https://github.com/maorash/Local-Global-INRs.
comment: Accepted to NeurIPS 2024
♻ ☆ Breaking Class Barriers: Efficient Dataset Distillation via Inter-Class Feature Compensator
Dataset distillation has emerged as a technique aiming to condense informative features from large, natural datasets into a compact and synthetic form. While recent advancements have refined this technique, its performance is bottlenecked by the prevailing class-specific synthesis paradigm. Under this paradigm, synthetic data is optimized exclusively for a pre-assigned one-hot label, creating an implicit class barrier in feature condensation. This leads to inefficient utilization of the distillation budget and oversight of inter-class feature distributions, which ultimately limits the effectiveness and efficiency, as demonstrated in our analysis. To overcome these constraints, this paper presents the Inter-class Feature Compensator (INFER), an innovative distillation approach that transcends the class-specific data-label framework widely utilized in current dataset distillation methods. Specifically, INFER leverages a Universal Feature Compensator (UFC) to enhance feature integration across classes, enabling the generation of multiple additional synthetic instances from a single UFC input. This significantly improves the efficiency of the distillation budget. Moreover, INFER enriches inter-class interactions during the distillation, thereby enhancing the effectiveness and generalizability of the distilled data. By allowing for the linear interpolation of labels similar to those in the original dataset, INFER meticulously optimizes the synthetic data and dramatically reduces the size of soft labels in the synthetic dataset to almost zero, establishing a new benchmark for efficiency and effectiveness in dataset distillation.
♻ ☆ Exploring Stronger Transformer Representation Learning for Occluded Person Re-Identification
Due to some complex factors (e.g., occlusion, pose variation and diverse camera perspectives), extracting stronger feature representation in person re-identification remains a challenging task. In this paper, we proposed a novel self-supervision and supervision combining transformer-based person re-identification framework, namely SSSC-TransReID. Different from the general transformer-based person re-identification models, we designed a self-supervised contrastive learning branch, which can enhance the feature representation for person re-identification without negative samples or additional pre-training. In order to train the contrastive learning branch, we also proposed a novel random rectangle mask strategy to simulate the occlusion in real scenes, so as to enhance the feature representation for occlusion. Finally, we utilized the joint-training loss function to integrate the advantages of supervised learning with ID tags and self-supervised contrastive learning without negative samples, which can reinforce the ability of our model to excavate stronger discriminative features, especially for occlusion. Extensive experimental results on several benchmark datasets show our proposed model obtains superior Re-ID performance consistently and outperforms the state-of-the-art ReID methods by large margins on the mean average accuracy (mAP) and Rank-1 accuracy.
♻ ☆ From Real Artifacts to Virtual Reference: A Robust Framework for Translating Endoscopic Images
Domain adaptation, which bridges the distributions across different modalities, plays a crucial role in multimodal medical image analysis. In endoscopic imaging, combining pre-operative data with intra-operative imaging is important for surgical planning and navigation. However, existing domain adaptation methods are hampered by distribution shift caused by in vivo artifacts, necessitating robust techniques for aligning noisy and artifact abundant patient endoscopic videos with clean virtual images reconstructed from pre-operative tomographic data for pose estimation during intraoperative guidance. This paper presents an artifact-resilient image translation method and an associated benchmark for this purpose. The method incorporates a novel ``local-global'' translation framework and a noise-resilient feature extraction strategy. For the former, it decouples the image translation process into a local step for feature denoising, and a global step for global style transfer. For feature extraction, a new contrastive learning strategy is proposed, which can extract noise-resilient features for establishing robust correspondence across domains. Detailed validation on both public and in-house clinical datasets has been conducted, demonstrating significantly improved performance compared to the current state-of-the-art.
♻ ☆ Few-Shot Adversarial Prompt Learning on Vision-Language Models NeurIPS 2024
The vulnerability of deep neural networks to imperceptible adversarial perturbations has attracted widespread attention. Inspired by the success of vision-language foundation models, previous efforts achieved zero-shot adversarial robustness by aligning adversarial visual features with text supervision. However, in practice, they are still unsatisfactory due to several issues, including heavy adaptation cost, suboptimal text supervision, and uncontrolled natural generalization capacity. In this paper, to address these issues, we propose a few-shot adversarial prompt framework where adapting input sequences with limited data makes significant adversarial robustness improvement. Specifically, we achieve this by providing adversarially correlated text supervision that is end-to-end learned from adversarial examples. We also propose a novel training objective that enhances the consistency of multi-modal features while encourages differentiated uni-modal features between natural and adversarial examples. The proposed framework gives access to learn adversarial text supervision, which provides superior cross-modal adversarial alignment and matches state-of-the-art zero-shot adversarial robustness with only 1% training data. Code is available at: https://github.com/lionel-w2/FAP.
comment: NeurIPS 2024
♻ ☆ Enhancing Interaction Modeling with Agent Selection and Physical Coefficient for Trajectory Prediction SP
A thorough understanding of the interaction between the target agent and surrounding agents is a prerequisite for accurate trajectory prediction. Although many methods have been explored, they all assign correlation coefficients to surrounding agents in a purely learning-based manner. In this study, we present ASPILin, which manually selects interacting agents and calculates their correlations instead of attention scores. Surprisingly, these simple modifications can significantly improve prediction performance and substantially reduce computational costs. Additionally, ASPILin models the interacting agents at each past time step separately, rather than only modeling the interacting agents at the current time step. This clarifies the causal chain of the target agent's historical trajectory and helps the model better understand dynamic interactions. We intentionally simplified our model in other aspects, such as map encoding. Remarkably, experiments conducted on the INTERACTION, highD, and CitySim datasets demonstrate that our method is efficient and straightforward, outperforming other state-of-the-art methods.
comment: code:https://github.com/kkk00714/ASPILin
♻ ☆ Conquering the Communication Constraints to Enable Large Pre-Trained Models in Federated Learning
Federated learning (FL) has emerged as a promising paradigm for enabling the collaborative training of models without centralized access to the raw data on local devices. In the typical FL paradigm (e.g., FedAvg), model weights are sent to and from the server each round to participating clients. Recently, the use of small pre-trained models has been shown effective in federated learning optimization and improving convergence. However, recent state-of-the-art pre-trained models are getting more capable but also have more parameters. In conventional FL, sharing the enormous model weights can quickly put a massive communication burden on the system, especially if more capable models are employed. Can we find a solution to enable those strong and readily-available pre-trained models in FL to achieve excellent performance while simultaneously reducing the communication burden? To this end, we investigate the use of parameter-efficient fine-tuning in federated learning and thus introduce a new framework: FedPEFT. Specifically, we systemically evaluate the performance of FedPEFT across a variety of client stability, data distribution, and differential privacy settings. By only locally tuning and globally sharing a small portion of the model weights, significant reductions in the total communication overhead can be achieved while maintaining competitive or even better performance in a wide range of federated learning scenarios, providing insight into a new paradigm for practical and effective federated systems.
♻ ☆ Latent Noise Segmentation: How Neural Noise Leads to the Emergence of Segmentation and Grouping ICML 2024
Humans are able to segment images effortlessly without supervision using perceptual grouping. Here, we propose a counter-intuitive computational approach to solving unsupervised perceptual grouping and segmentation: that they arise because of neural noise, rather than in spite of it. We (1) mathematically demonstrate that under realistic assumptions, neural noise can be used to separate objects from each other; (2) that adding noise in a DNN enables the network to segment images even though it was never trained on any segmentation labels; and (3) that segmenting objects using noise results in segmentation performance that aligns with the perceptual grouping phenomena observed in humans, and is sample-efficient. We introduce the Good Gestalt (GG) datasets -- six datasets designed to specifically test perceptual grouping, and show that our DNN models reproduce many important phenomena in human perception, such as illusory contours, closure, continuity, proximity, and occlusion. Finally, we (4) show that our model improves performance on our GG datasets compared to other tested unsupervised models by $24.9\%$. Together, our results suggest a novel unsupervised segmentation method requiring few assumptions, a new explanation for the formation of perceptual grouping, and a novel potential benefit of neural noise.
comment: ICML 2024 camera ready version
♻ ☆ STBA: Towards Evaluating the Robustness of DNNs for Query-Limited Black-box Scenario
Many attack techniques have been proposed to explore the vulnerability of DNNs and further help to improve their robustness. Despite the significant progress made recently, existing black-box attack methods still suffer from unsatisfactory performance due to the vast number of queries needed to optimize desired perturbations. Besides, the other critical challenge is that adversarial examples built in a noise-adding manner are abnormal and struggle to successfully attack robust models, whose robustness is enhanced by adversarial training against small perturbations. There is no doubt that these two issues mentioned above will significantly increase the risk of exposure and result in a failure to dig deeply into the vulnerability of DNNs. Hence, it is necessary to evaluate DNNs' fragility sufficiently under query-limited settings in a non-additional way. In this paper, we propose the Spatial Transform Black-box Attack (STBA), a novel framework to craft formidable adversarial examples in the query-limited scenario. Specifically, STBA introduces a flow field to the high-frequency part of clean images to generate adversarial examples and adopts the following two processes to enhance their naturalness and significantly improve the query efficiency: a) we apply an estimated flow field to the high-frequency part of clean images to generate adversarial examples instead of introducing external noise to the benign image, and b) we leverage an efficient gradient estimation method based on a batch of samples to optimize such an ideal flow field under query-limited settings. Compared to existing score-based black-box baselines, extensive experiments indicated that STBA could effectively improve the imperceptibility of the adversarial examples and remarkably boost the attack success rate under query-limited settings.
comment: Accepted by T-MM
♻ ☆ ODTFormer: Efficient Obstacle Detection and Tracking with Stereo Cameras Based on Transformer IROS 2024
Obstacle detection and tracking represent a critical component in robot autonomous navigation. In this paper, we propose ODTFormer, a Transformer-based model to address both obstacle detection and tracking problems. For the detection task, our approach leverages deformable attention to construct a 3D cost volume, which is decoded progressively in the form of voxel occupancy grids. We further track the obstacles by matching the voxels between consecutive frames. The entire model can be optimized in an end-to-end manner. Through extensive experiments on DrivingStereo and KITTI benchmarks, our model achieves state-of-the-art performance in the obstacle detection task. We also report comparable accuracy to state-of-the-art obstacle tracking models while requiring only a fraction of their computation cost, typically ten-fold to twenty-fold less. The code and model weights will be publicly released.
comment: 8 pages. Accepted by IROS 2024
♻ ☆ Meteor: Mamba-based Traversal of Rationale for Large Language and Vision Models
The rapid development of large language and vision models (LLVMs) has been driven by advances in visual instruction tuning. Recently, open-source LLVMs have curated high-quality visual instruction tuning datasets and utilized additional vision encoders or multiple computer vision models in order to narrow the performance gap with powerful closed-source LLVMs. These advancements are attributed to multifaceted information required for diverse capabilities, including fundamental image understanding, real-world knowledge about common-sense and non-object concepts (e.g., charts, diagrams, symbols, signs, and math problems), and step-by-step procedures for solving complex questions. Drawing from the multifaceted information, we present a new efficient LLVM, Mamba-based traversal of rationales (Meteor), which leverages multifaceted rationale to enhance understanding and answering capabilities. To embed lengthy rationales containing abundant information, we employ the Mamba architecture, capable of processing sequential data with linear time complexity. We introduce a new concept of traversal of rationale that facilitates efficient embedding of rationale. Subsequently, the backbone multimodal language model (MLM) is trained to generate answers with the aid of rationale. Through these steps, Meteor achieves significant improvements in vision language performances across multiple evaluation benchmarks requiring diverse capabilities, without scaling up the model size or employing additional vision encoders and computer vision models.
comment: Code is available in https://github.com/ByungKwanLee/Meteor
♻ ☆ LLaVA-MoD: Making LLaVA Tiny via MoE Knowledge Distillation
We introduce LLaVA-MoD, a novel framework designed to enable the efficient training of small-scale Multimodal Language Models (s-MLLM) by distilling knowledge from large-scale MLLM (l-MLLM). Our approach tackles two fundamental challenges in MLLM distillation. First, we optimize the network structure of s-MLLM by integrating a sparse Mixture of Experts (MoE) architecture into the language model, striking a balance between computational efficiency and model expressiveness. Second, we propose a progressive knowledge transfer strategy to ensure comprehensive knowledge migration. This strategy begins with mimic distillation, where we minimize the Kullback-Leibler (KL) divergence between output distributions to enable the student model to emulate the teacher network's understanding. Following this, we introduce preference distillation via Direct Preference Optimization (DPO), where the key lies in treating l-MLLM as the reference model. During this phase, the s-MLLM's ability to discriminate between superior and inferior examples is significantly enhanced beyond l-MLLM, leading to a better student that surpasses its teacher, particularly in hallucination benchmarks. Extensive experiments demonstrate that LLaVA-MoD outperforms existing models across various multimodal benchmarks while maintaining a minimal number of activated parameters and low computational costs. Remarkably, LLaVA-MoD, with only 2B activated parameters, surpasses Qwen-VL-Chat-7B by an average of 8.8% across benchmarks, using merely 0.3% of the training data and 23% trainable parameters. These results underscore LLaVA-MoD's ability to effectively distill comprehensive knowledge from its teacher model, paving the way for the development of more efficient MLLMs. The code will be available on: https://github.com/shufangxun/LLaVA-MoD.
♻ ☆ DIP-Watermark: A Double Identity Protection Method Based on Robust Adversarial Watermark
The wide deployment of Face Recognition (FR) systems poses privacy risks. One countermeasure is adversarial attack, deceiving unauthorized malicious FR, but it also disrupts regular identity verification of trusted authorizers, exacerbating the potential threat of identity impersonation. To address this, we propose the first double identity protection scheme based on traceable adversarial watermarking, termed DIP-Watermark. DIP-Watermark employs a one-time watermark embedding to deceive unauthorized FR models and allows authorizers to perform identity verification by extracting the watermark. Specifically, we propose an information-guided adversarial attack against FR models. The encoder embeds an identity-specific watermark into the deep feature space of the carrier, guiding recognizable features of the image to deviate from the source identity. We further adopt a collaborative meta-optimization strategy compatible with sub-tasks, which regularizes the joint optimization direction of the encoder and decoder. This strategy enhances the representation of universal carrier features, mitigating multi-objective optimization conflicts in watermarking. Experiments confirm that DIP-Watermark achieves significant attack success rates and traceability accuracy on state-of-the-art FR models, exhibiting remarkable robustness that outperforms the existing privacy protection methods using adversarial attacks and deep watermarking, or simple combinations of the two. Our work potentially opens up new insights into proactive protection for FR privacy.
♻ ☆ Hierarchical Light Transformer Ensembles for Multimodal Trajectory Forecasting
Accurate trajectory forecasting is crucial for the performance of various systems, such as advanced driver-assistance systems and self-driving vehicles. These forecasts allow to anticipate events leading to collisions and, therefore, to mitigate them. Deep Neural Networks have excelled in motion forecasting, but issues like overconfidence and uncertainty quantification persist. Deep Ensembles address these concerns, yet applying them to multimodal distributions remains challenging. In this paper, we propose a novel approach named Hierarchical Light Transformer Ensembles (HLT-Ens), aimed at efficiently training an ensemble of Transformer architectures using a novel hierarchical loss function. HLT-Ens leverages grouped fully connected layers, inspired by grouped convolution techniques, to capture multimodal distributions, effectively. Through extensive experimentation, we demonstrate that HLT-Ens achieves state-of-the-art performance levels, offering a promising avenue for improving trajectory forecasting techniques.
comment: acknowledgement added
♻ ☆ Pulling Target to Source: A New Perspective on Domain Adaptive Semantic Segmentation
Domain adaptive semantic segmentation aims to transfer knowledge from a labeled source domain to an unlabeled target domain. However, existing methods primarily focus on directly learning qualified target features, making it challenging to guarantee their discrimination in the absence of target labels. This work provides a new perspective. We observe that the features learned with source data manage to keep categorically discriminative during training, thereby enabling us to implicitly learn adequate target representations by simply \textbf{pulling target features close to source features for each category}. To this end, we propose T2S-DA, which we interpret as a form of pulling Target to Source for Domain Adaptation, encouraging the model in learning similar cross-domain features. Also, considering the pixel categories are heavily imbalanced for segmentation datasets, we come up with a dynamic re-weighting strategy to help the model concentrate on those underperforming classes. Extensive experiments confirm that T2S-DA learns a more discriminative and generalizable representation, significantly surpassing the state-of-the-art. We further show that our method is quite qualified for the domain generalization task, verifying its domain-invariant property.
comment: Accepted by IJCV
♻ ☆ Improving Text Generation on Images with Synthetic Captions
The recent emergence of latent diffusion models such as SDXL and SD 1.5 has shown significant capability in generating highly detailed and realistic images. Despite their remarkable ability to produce images, generating accurate text within images still remains a challenging task. In this paper, we examine the validity of fine-tuning approaches in generating legible text within the image. We propose a low-cost approach by leveraging SDXL without any time-consuming training on large-scale datasets. The proposed strategy employs a fine-tuning technique that examines the effects of data refinement levels and synthetic captions. Moreover, our results demonstrate how our small scale fine-tuning approach can improve the accuracy of text generation in different scenarios without the need of additional multimodal encoders. Our experiments show that with the addition of random letters to our raw dataset, our model's performance improves in producing well-formed visual text.
comment: 2024 16th IIAI International Congress on Advanced Applied Informatics (IIAI-AAI)
♻ ☆ Harmonizing Visual Text Comprehension and Generation NeurIPS 2024
In this work, we present TextHarmony, a unified and versatile multimodal generative model proficient in comprehending and generating visual text. Simultaneously generating images and texts typically results in performance degradation due to the inherent inconsistency between vision and language modalities. To overcome this challenge, existing approaches resort to modality-specific data for supervised fine-tuning, necessitating distinct model instances. We propose Slide-LoRA, which dynamically aggregates modality-specific and modality-agnostic LoRA experts, partially decoupling the multimodal generation space. Slide-LoRA harmonizes the generation of vision and language within a singular model instance, thereby facilitating a more unified generative process. Additionally, we develop a high-quality image caption dataset, DetailedTextCaps-100K, synthesized with a sophisticated closed-source MLLM to enhance visual text generation capabilities further. Comprehensive experiments across various benchmarks demonstrate the effectiveness of the proposed approach. Empowered by Slide-LoRA, TextHarmony achieves comparable performance to modality-specific fine-tuning results with only a 2% increase in parameters and shows an average improvement of 2.5% in visual text comprehension tasks and 4.0% in visual text generation tasks. Our work delineates the viability of an integrated approach to multimodal generation within the visual text domain, setting a foundation for subsequent inquiries. Code is available at https://github.com/bytedance/TextHarmony.
comment: accepted by NeurIPS 2024
♻ ☆ Lightning-Fast Image Inversion and Editing for Text-to-Image Diffusion Models
Diffusion inversion is the problem of taking an image and a text prompt that describes it and finding a noise latent that would generate the exact same image. Most current deterministic inversion techniques operate by approximately solving an implicit equation and may converge slowly or yield poor reconstructed images. We formulate the problem by finding the roots of an implicit equation and devlop a method to solve it efficiently. Our solution is based on Newton-Raphson (NR), a well-known technique in numerical analysis. We show that a vanilla application of NR is computationally infeasible while naively transforming it to a computationally tractable alternative tends to converge to out-of-distribution solutions, resulting in poor reconstruction and editing. We therefore derive an efficient guided formulation that fastly converges and provides high-quality reconstructions and editing. We showcase our method on real image editing with three popular open-sourced diffusion models: Stable Diffusion, SDXL-Turbo, and Flux with different deterministic schedulers. Our solution, Guided Newton-Raphson Inversion, inverts an image within 0.4 sec (on an A100 GPU) for few-step models (SDXL-Turbo and Flux.1), opening the door for interactive image editing. We further show improved results in image interpolation and generation of rare objects.
♻ ☆ Improving Instance Optimization in Deformable Image Registration with Gradient Projection MICCAI 2024
Deformable image registration is inherently a multi-objective optimization (MOO) problem, requiring a delicate balance between image similarity and deformation regularity. These conflicting objectives often lead to poor optimization outcomes, such as being trapped in unsatisfactory local minima or experiencing slow convergence. Deep learning methods have recently gained popularity in this domain due to their efficiency in processing large datasets and achieving high accuracy. However, they often underperform during test time compared to traditional optimization techniques, which further explore iterative, instance-specific gradient-based optimization. This performance gap is more pronounced when a distribution shift between training and test data exists. To address this issue, we focus on the instance optimization (IO) paradigm, which involves additional optimization for test-time instances based on a pre-trained model. IO effectively combines the generalization capabilities of deep learning with the fine-tuning advantages of instance-specific optimization. Within this framework, we emphasize the use of gradient projection to mitigate conflicting updates in MOO. This technique projects conflicting gradients into a common space, better aligning the dual objectives and enhancing optimization stability. We validate our method using a state-of-the-art foundation model on the 3D Brain inter-subject registration task (LUMIR) from the Learn2Reg 2024 Challenge. Our results show significant improvements over standard gradient descent, leading to more accurate and reliable registration results.
comment: Learn2Reg Challenge at MICCAI 2024
♻ ☆ ERX: A Fast Real-Time Anomaly Detection Algorithm for Hyperspectral Line Scanning
Detecting unexpected objects (anomalies) in real time has great potential for monitoring, managing, and protecting the environment. Hyperspectral line-scan cameras are a low-cost solution that enhance confidence in anomaly detection over RGB and multispectral imagery. However, existing line-scan algorithms are too slow when using small computers (e.g. those onboard a drone or small satellite), do not adapt to changing scenery, or lack robustness against geometric distortions. This paper introduces the Exponentially moving RX algorithm (ERX) to address these issues, and compares it with existing RX-based anomaly detection methods for hyperspectral line scanning. Three large and more complex datasets are also introduced to better assess the practical challenges when using line-scan cameras (two hyperspectral and one multispectral). ERX is evaluated using a Jetson Xavier NX compute module, achieving the best combination of speed and detection performance. This research paves the way for future studies in grouping and locating anomalous objects, adaptive and automatic threshold selection, and real-time field tests. The datasets and the Python code are available at: https://github.com/WiseGamgee/HyperAD.
comment: 17 pages, 13 figures, 4 tables, code and datasets accessible at https://github.com/WiseGamgee/HyperAD
♻ ☆ CAT: Contrastive Adapter Training for Personalized Image Generation CVPR
The emergence of various adapters, including Low-Rank Adaptation (LoRA) applied from the field of natural language processing, has allowed diffusion models to personalize image generation at a low cost. However, due to the various challenges including limited datasets and shortage of regularization and computation resources, adapter training often results in unsatisfactory outcomes, leading to the corruption of the backbone model's prior knowledge. One of the well known phenomena is the loss of diversity in object generation, especially within the same class which leads to generating almost identical objects with minor variations. This poses challenges in generation capabilities. To solve this issue, we present Contrastive Adapter Training (CAT), a simple yet effective strategy to enhance adapter training through the application of CAT loss. Our approach facilitates the preservation of the base model's original knowledge when the model initiates adapters. Furthermore, we introduce the Knowledge Preservation Score (KPS) to evaluate CAT's ability to keep the former information. We qualitatively and quantitatively compare CAT's improvement. Finally, we mention the possibility of CAT in the aspects of multi-concept adapter and optimization.
comment: CVPRW 2024
♻ ☆ Gaussian-Informed Continuum for Physical Property Identification and Simulation NeurIPS 2024
This paper studies the problem of estimating physical properties (system identification) through visual observations. To facilitate geometry-aware guidance in physical property estimation, we introduce a novel hybrid framework that leverages 3D Gaussian representation to not only capture explicit shapes but also enable the simulated continuum to render object masks as 2D shape surrogates during training. We propose a new dynamic 3D Gaussian framework based on motion factorization to recover the object as 3D Gaussian point sets across different time states. Furthermore, we develop a coarse-to-fine filling strategy to generate the density fields of the object from the Gaussian reconstruction, allowing for the extraction of object continuums along with their surfaces and the integration of Gaussian attributes into these continuums. In addition to the extracted object surfaces, the Gaussian-informed continuum also enables the rendering of object masks during simulations, serving as 2D-shape guidance for physical property estimation. Extensive experimental evaluations demonstrate that our pipeline achieves state-of-the-art performance across multiple benchmarks and metrics. Additionally, we illustrate the effectiveness of the proposed method through real-world demonstrations, showcasing its practical utility. Our project page is at https://jukgei.github.io/project/gic.
comment: 21 pages, 8 figures, NeurIPS 2024
♻ ☆ Toward Fairer Face Recognition Datasets
Face recognition and verification are two computer vision tasks whose performance has progressed with the introduction of deep representations. However, ethical, legal, and technical challenges due to the sensitive character of face data and biases in real training datasets hinder their development. Generative AI addresses privacy by creating fictitious identities, but fairness problems persist. We promote fairness by introducing a demographic attributes balancing mechanism in generated training datasets. We experiment with an existing real dataset, three generated training datasets, and the balanced versions of a diffusion-based dataset. We propose a comprehensive evaluation that considers accuracy and fairness equally and includes a rigorous regression-based statistical analysis of attributes. The analysis shows that balancing reduces demographic unfairness. Also, a performance gap persists despite generation becoming more accurate with time. The proposed balancing method and comprehensive verification evaluation promote fairer and transparent face recognition and verification.
♻ ☆ LVBench: An Extreme Long Video Understanding Benchmark
Recent progress in multimodal large language models has markedly enhanced the understanding of short videos (typically under one minute), and several evaluation datasets have emerged accordingly. However, these advancements fall short of meeting the demands of real-world applications such as embodied intelligence for long-term decision-making, in-depth movie reviews and discussions, and live sports commentary, all of which require comprehension of long videos spanning several hours. To address this gap, we introduce LVBench, a benchmark specifically designed for long video understanding. Our dataset comprises publicly sourced videos and encompasses a diverse set of tasks aimed at long video comprehension and information extraction. LVBench is designed to challenge multimodal models to demonstrate long-term memory and extended comprehension capabilities. Our extensive evaluations reveal that current multimodal models still underperform on these demanding long video understanding tasks. Through LVBench, we aim to spur the development of more advanced models capable of tackling the complexities of long video comprehension. Our data and code are publicly available at: https://lvbench.github.io.
♻ ☆ Advancing Open-Set Domain Generalization Using Evidential Bi-Level Hardest Domain Scheduler NeurIPS 2024
In Open-Set Domain Generalization (OSDG), the model is exposed to both new variations of data appearance (domains) and open-set conditions, where both known and novel categories are present at test time. The challenges of this task arise from the dual need to generalize across diverse domains and accurately quantify category novelty, which is critical for applications in dynamic environments. Recently, meta-learning techniques have demonstrated superior results in OSDG, effectively orchestrating the meta-train and -test tasks by employing varied random categories and predefined domain partition strategies. These approaches prioritize a well-designed training schedule over traditional methods that focus primarily on data augmentation and the enhancement of discriminative feature learning. The prevailing meta-learning models in OSDG typically utilize a predefined sequential domain scheduler to structure data partitions. However, a crucial aspect that remains inadequately explored is the influence brought by strategies of domain schedulers during training. In this paper, we observe that an adaptive domain scheduler benefits more in OSDG compared with prefixed sequential and random domain schedulers. We propose the Evidential Bi-Level Hardest Domain Scheduler (EBiL-HaDS) to achieve an adaptive domain scheduler. This method strategically sequences domains by assessing their reliabilities in utilizing a follower network, trained with confidence scores learned in an evidential manner, regularized by max rebiasing discrepancy, and optimized in a bi-level manner. The results show that our method substantially improves OSDG performance and achieves more discriminative embeddings for both the seen and unseen categories. The source code is publicly available at https://github.com/KPeng9510/EBiL-HaDS.
comment: Accepted to NeurIPS 2024. The source code is publicly available at https://github.com/KPeng9510/EBiL-HaDS
♻ ☆ Learning to Manipulate Anywhere: A Visual Generalizable Framework For Reinforcement Learning
Can we endow visuomotor robots with generalization capabilities to operate in diverse open-world scenarios? In this paper, we propose \textbf{Maniwhere}, a generalizable framework tailored for visual reinforcement learning, enabling the trained robot policies to generalize across a combination of multiple visual disturbance types. Specifically, we introduce a multi-view representation learning approach fused with Spatial Transformer Network (STN) module to capture shared semantic information and correspondences among different viewpoints. In addition, we employ a curriculum-based randomization and augmentation approach to stabilize the RL training process and strengthen the visual generalization ability. To exhibit the effectiveness of Maniwhere, we meticulously design 8 tasks encompassing articulate objects, bi-manual, and dexterous hand manipulation tasks, demonstrating Maniwhere's strong visual generalization and sim2real transfer abilities across 3 hardware platforms. Our experiments show that Maniwhere significantly outperforms existing state-of-the-art methods. Videos are provided at https://gemcollector.github.io/maniwhere/.
comment: Webpage: https://gemcollector.github.io/maniwhere/
♻ ☆ The Ultimate Combo: Boosting Adversarial Example Transferability by Composing Data Augmentations
To help adversarial examples generalize from surrogate machine-learning (ML) models to targets, certain transferability-based black-box evasion attacks incorporate data augmentations (e.g., random resizing). Yet, prior work has explored limited augmentations and their composition. To fill the gap, we systematically studied how data augmentation affects transferability. Specifically, we explored 46 augmentation techniques originally proposed to help ML models generalize to unseen benign samples, and assessed how they impact transferability, when applied individually or composed. Performing exhaustive search on a small subset of augmentation techniques and genetic search on all techniques, we identified augmentation combinations that help promote transferability. Extensive experiments with the ImageNet and CIFAR-10 datasets and 18 models showed that simple color-space augmentations (e.g., color to greyscale) attain high transferability when combined with standard augmentations. Furthermore, we discovered that composing augmentations impacts transferability mostly monotonically (i.e., more augmentations $\rightarrow$ $\ge$transferability). We also found that the best composition significantly outperformed the state of the art (e.g., 91.8% vs. $\le$82.5% average transferability to adversarially trained targets on ImageNet). Lastly, our theoretical analysis, backed by empirical evidence, intuitively explains why certain augmentations promote transferability.
comment: Accepted by AISec'24
♻ ☆ Diffusion Models are Certifiably Robust Classifiers NeurIPS 2024
Generative learning, recognized for its effective modeling of data distributions, offers inherent advantages in handling out-of-distribution instances, especially for enhancing robustness to adversarial attacks. Among these, diffusion classifiers, utilizing powerful diffusion models, have demonstrated superior empirical robustness. However, a comprehensive theoretical understanding of their robustness is still lacking, raising concerns about their vulnerability to stronger future attacks. In this study, we prove that diffusion classifiers possess $O(1)$ Lipschitzness, and establish their certified robustness, demonstrating their inherent resilience. To achieve non-constant Lipschitzness, thereby obtaining much tighter certified robustness, we generalize diffusion classifiers to classify Gaussian-corrupted data. This involves deriving the evidence lower bounds (ELBOs) for these distributions, approximating the likelihood using the ELBO, and calculating classification probabilities via Bayes' theorem. Experimental results show the superior certified robustness of these Noised Diffusion Classifiers (NDCs). Notably, we achieve over 80% and 70% certified robustness on CIFAR-10 under adversarial perturbations with \(\ell_2\) norms less than 0.25 and 0.5, respectively, using a single off-the-shelf diffusion model without any additional data.
comment: Accepted by NeurIPS 2024
♻ ☆ SemiSAM: Enhancing Semi-Supervised Medical Image Segmentation via SAM-Assisted Consistency Regularization
Semi-supervised learning has attracted much attention due to its less dependence on acquiring abundant annotations from experts compared to fully supervised methods, which is especially important for medical image segmentation which typically requires intensive pixel/voxel-wise labeling by domain experts. Although semi-supervised methods can improve the performance by utilizing unlabeled data, there are still gaps between fully supervised methods under extremely limited annotation scenarios. In this paper, we propose a simple yet efficient strategy to explore the usage of the Segment Anything Model (SAM) for enhancing semi-supervised medical image segmentation. Concretely, the segmentation model trained with domain knowledge provides information for localization and generating input prompts to the SAM. Then the generated pseudo-labels of SAM are utilized as additional supervision to assist in the learning procedure of the semi-supervised framework. Extensive experiments demonstrate that SemiSAM significantly improves the performance of existing semi-supervised frameworks when only one or a few labeled images are available and shows strong efficiency as a plug-and-play strategy for semi-supervised medical image segmentation.
comment: Accept for BIBM 2024
♻ ☆ RotCAtt-TransUNet++: Novel Deep Neural Network for Sophisticated Cardiac Segmentation
Cardiovascular disease remains a predominant global health concern, responsible for a significant portion of mortality worldwide. Accurate segmentation of cardiac medical imaging data is pivotal in mitigating fatality rates associated with cardiovascular conditions. However, existing state-of-the-art (SOTA) neural networks, including both CNN-based and Transformer-based approaches, exhibit limitations in practical applicability due to their inability to effectively capture inter-slice connections alongside intra-slice information. This deficiency is particularly evident in datasets featuring intricate, long-range details along the z-axis, such as coronary arteries in axial views. Additionally, SOTA methods fail to differentiate non-cardiac components from myocardium in segmentation, leading to the "spraying" phenomenon. To address these challenges, we present RotCAtt-TransUNet++, a novel architecture tailored for robust segmentation of complex cardiac structures. Our approach emphasizes modeling global contexts by aggregating multiscale features with nested skip connections in the encoder. It integrates transformer layers to capture interactions between patches and employs a rotatory attention mechanism to capture connectivity between multiple slices (inter-slice information). Additionally, a channel-wise cross-attention gate guides the fused multi-scale channel-wise information and features from decoder stages to bridge semantic gaps. Experimental results demonstrate that our proposed model outperforms existing SOTA approaches across four cardiac datasets and one abdominal dataset. Importantly, coronary arteries and myocardium are annotated with near-perfect accuracy during inference. An ablation study shows that the rotatory attention mechanism effectively transforms embedded vectorized patches in the semantic dimensional space, enhancing segmentation accuracy.
comment: 11 pages, 11 figures
♻ ☆ Exploring Self-Supervised Skeleton-Based Human Action Recognition under Occlusions
To integrate self-supervised skeleton-based action recognition methods into autonomous robotic systems, it is crucial to consider adverse situations involving target occlusions. Such a scenario, despite its practical relevance, is rarely addressed in existing self-supervised skeleton-based action recognition methods. To empower models with the capacity to address occlusion, we propose a simple and effective method. We first pre-train using occluded skeleton sequences, then use k-means clustering (KMeans) on sequence embeddings to group semantically similar samples. Next, we propose KNN-Imputation to fill in missing skeleton data based on the closest sample neighbors. Imputing incomplete skeleton sequences to create relatively complete sequences as input provides significant benefits to existing skeleton-based self-supervised methods. Meanwhile, building on the state-of-the-art Partial Spatio-Temporal Learning (PSTL), we introduce an Occluded Partial Spatio-Temporal Learning (OPSTL) framework. This enhancement utilizes Adaptive Spatial Masking (ASM) for better use of high-quality, intact skeletons. The new proposed method is verified on the challenging occluded versions of the NTURGB+D 60 and NTURGB+D 120. The source code is publicly available at https://github.com/cyfml/OPSTL.
comment: The source code is publicly available at https://github.com/cyfml/OPSTL
♻ ☆ RealignDiff: Boosting Text-to-Image Diffusion Model with Coarse-to-fine Semantic Re-alignment
Recent advances in text-to-image diffusion models have achieved remarkable success in generating high-quality, realistic images from textual descriptions. However, these approaches have faced challenges in precisely aligning the generated visual content with the textual concepts described in the prompts. In this paper, we propose a two-stage coarse-to-fine semantic re-alignment method, named RealignDiff, aimed at improving the alignment between text and images in text-to-image diffusion models. In the coarse semantic re-alignment phase, a novel caption reward, leveraging the BLIP-2 model, is proposed to evaluate the semantic discrepancy between the generated image caption and the given text prompt. Subsequently, the fine semantic re-alignment stage employs a local dense caption generation module and a re-weighting attention modulation module to refine the previously generated images from a local semantic view. Experimental results on the MS-COCO and ViLG-300 datasets demonstrate that the proposed two-stage coarse-to-fine semantic re-alignment method outperforms other baseline re-alignment techniques by a substantial margin in both visual quality and semantic similarity with the input prompt.
♻ ☆ Real-World Robot Applications of Foundation Models: A Review
Recent developments in foundation models, like Large Language Models (LLMs) and Vision-Language Models (VLMs), trained on extensive data, facilitate flexible application across different tasks and modalities. Their impact spans various fields, including healthcare, education, and robotics. This paper provides an overview of the practical application of foundation models in real-world robotics, with a primary emphasis on the replacement of specific components within existing robot systems. The summary encompasses the perspective of input-output relationships in foundation models, as well as their role in perception, motion planning, and control within the field of robotics. This paper concludes with a discussion of future challenges and implications for practical robot applications.
♻ ☆ MMBench-Video: A Long-Form Multi-Shot Benchmark for Holistic Video Understanding NeurIPS 2024
The advent of large vision-language models (LVLMs) has spurred research into their applications in multi-modal contexts, particularly in video understanding. Traditional VideoQA benchmarks, despite providing quantitative metrics, often fail to encompass the full spectrum of video content and inadequately assess models' temporal comprehension. To address these limitations, we introduce MMBench-Video, a quantitative benchmark designed to rigorously evaluate LVLMs' proficiency in video understanding. MMBench-Video incorporates lengthy videos from YouTube and employs free-form questions, mirroring practical use cases. The benchmark is meticulously crafted to probe the models' temporal reasoning skills, with all questions human-annotated according to a carefully constructed ability taxonomy. We employ GPT-4 for automated assessment, demonstrating superior accuracy and robustness over earlier LLM-based evaluations. Utilizing MMBench-Video, we have conducted comprehensive evaluations that include both proprietary and open-source LVLMs for images and videos. MMBench-Video stands as a valuable resource for the research community, facilitating improved evaluation of LVLMs and catalyzing progress in the field of video understanding. The evalutation code of MMBench-Video will be integrated into VLMEvalKit: https://github.com/open-compass/VLMEvalKit.
comment: Accepted in NeurIPS 2024 Datasets and Benchmarks Track
♻ ☆ Selective Vision is the Challenge for Visual Reasoning: A Benchmark for Visual Argument Understanding EMNLP 2024
Visual arguments, often used in advertising or social causes, rely on images to persuade viewers to do or believe something. Understanding these arguments requires selective vision: only specific visual stimuli within an image are relevant to the argument, and relevance can only be understood within the context of a broader argumentative structure. While visual arguments are readily appreciated by human audiences, we ask: are today's AI capable of similar understanding? We present VisArgs, a dataset of 1,611 images annotated with 5,112 visual premises (with regions), 5,574 commonsense premises, and reasoning trees connecting them into structured arguments. We propose three tasks for evaluating visual argument understanding: premise localization, premise identification, and conclusion deduction. Experiments show that 1) machines struggle to capture visual cues: GPT-4-O achieved 78.5% accuracy, while humans reached 98.0%. Models also performed 19.5% worse when distinguishing between irrelevant objects within the image compared to external objects. 2) Providing relevant visual premises improved model performance significantly.
comment: 12 pages, 6 figures. Accepted as main paper in EMNLP 2024
♻ ☆ Can visual language models resolve textual ambiguity with visual cues? Let visual puns tell you! EMNLP 2024
Humans possess multimodal literacy, allowing them to actively integrate information from various modalities to form reasoning. Faced with challenges like lexical ambiguity in text, we supplement this with other modalities, such as thumbnail images or textbook illustrations. Is it possible for machines to achieve a similar multimodal understanding capability? In response, we present Understanding Pun with Image Explanations (UNPIE), a novel benchmark designed to assess the impact of multimodal inputs in resolving lexical ambiguities. Puns serve as the ideal subject for this evaluation due to their intrinsic ambiguity. Our dataset includes 1,000 puns, each accompanied by an image that explains both meanings. We pose three multimodal challenges with the annotations to assess different aspects of multimodal literacy; Pun Grounding, Disambiguation, and Reconstruction. The results indicate that various Socratic Models and Visual-Language Models improve over the text-only models when given visual context, particularly as the complexity of the tasks increases.
comment: Accepted as main paper in EMNLP 2024
♻ ☆ CV-VAE: A Compatible Video VAE for Latent Generative Video Models
Spatio-temporal compression of videos, utilizing networks such as Variational Autoencoders (VAE), plays a crucial role in OpenAI's SORA and numerous other video generative models. For instance, many LLM-like video models learn the distribution of discrete tokens derived from 3D VAEs within the VQVAE framework, while most diffusion-based video models capture the distribution of continuous latent extracted by 2D VAEs without quantization. The temporal compression is simply realized by uniform frame sampling which results in unsmooth motion between consecutive frames. Currently, there lacks of a commonly used continuous video (3D) VAE for latent diffusion-based video models in the research community. Moreover, since current diffusion-based approaches are often implemented using pre-trained text-to-image (T2I) models, directly training a video VAE without considering the compatibility with existing T2I models will result in a latent space gap between them, which will take huge computational resources for training to bridge the gap even with the T2I models as initialization. To address this issue, we propose a method for training a video VAE of latent video models, namely CV-VAE, whose latent space is compatible with that of a given image VAE, e.g., image VAE of Stable Diffusion (SD). The compatibility is achieved by the proposed novel latent space regularization, which involves formulating a regularization loss using the image VAE. Benefiting from the latent space compatibility, video models can be trained seamlessly from pre-trained T2I or video models in a truly spatio-temporally compressed latent space, rather than simply sampling video frames at equal intervals. With our CV-VAE, existing video models can generate four times more frames with minimal finetuning. Extensive experiments are conducted to demonstrate the effectiveness of the proposed video VAE.
comment: Project Page: https://ailab-cvc.github.io/cvvae/index.html
♻ ☆ GPHM: Gaussian Parametric Head Model for Monocular Head Avatar Reconstruction
Creating high-fidelity 3D human head avatars is crucial for applications in VR/AR, digital human, and film production. Recent advances have leveraged morphable face models to generate animated head avatars from easily accessible data, representing varying identities and expressions within a low-dimensional parametric space. However, existing methods often struggle with modeling complex appearance details, e.g., hairstyles, and suffer from low rendering quality and efficiency. In this paper we introduce a novel approach, 3D Gaussian Parametric Head Model, which employs 3D Gaussians to accurately represent the complexities of the human head, allowing precise control over both identity and expression. The Gaussian model can handle intricate details, enabling realistic representations of varying appearances and complex expressions. Furthermore, we presents a well-designed training framework to ensure smooth convergence, providing a robust guarantee for learning the rich content. Our method achieves high-quality, photo-realistic rendering with real-time efficiency, making it a valuable contribution to the field of parametric head models. Finally, we apply the 3D Gaussian Parametric Head Model to monocular video or few-shot head avatar reconstruction tasks, which enables instant reconstruction of high-quality 3D head avatars even when input data is extremely limited, surpassing previous methods in terms of reconstruction quality and training speed.
comment: Project page: https://yuelangx.github.io/gphmv2/
♻ ☆ CoIN: A Benchmark of Continual Instruction tuNing for Multimodel Large Language Model
Instruction tuning represents a prevalent strategy employed by Multimodal Large Language Models (MLLMs) to align with human instructions and adapt to new tasks. Nevertheless, MLLMs encounter the challenge of adapting to users' evolving knowledge and demands. Therefore, how to retain existing skills while acquiring new knowledge needs to be investigated. In this paper, we present a comprehensive benchmark, namely Continual Instruction tuNing (CoIN), to assess existing MLLMs in the sequential instruction tuning paradigm. CoIN comprises 10 commonly used datasets spanning 8 task categories, ensuring a diverse range of instructions and tasks. Besides, the trained model is evaluated from two aspects: Instruction Following and General Knowledge, which assess the alignment with human intention and knowledge preserved for reasoning, respectively. Experiments on CoIN demonstrate that current powerful MLLMs still suffer catastrophic forgetting, and the failure in intention alignment assumes the main responsibility, instead of the knowledge forgetting. To this end, we introduce MoELoRA to MLLMs which is effective to retain the previous instruction alignment. Experimental results consistently illustrate the forgetting decreased from this method on CoIN.
♻ ☆ CD-NGP: A Fast Scalable Continual Representation for Dynamic Scenes
Current methodologies for novel view synthesis (NVS) in dynamic scenes encounter significant challenges in harmonizing memory consumption, model complexity, training efficiency, and rendering fidelity. Existing offline techniques, while delivering high-quality results, are often characterized by substantial memory demands and limited scalability. In contrast, online methods grapple with the challenge of balancing rapid convergence with model compactness. To address these issues, we propose continual dynamic neural graphics primitives (CD-NGP). Our approach synergizes features from both temporal and spatial hash encodings to achieve high rendering quality, employs parameter reuse to enhance scalability, and leverages a continual learning framework to mitigate memory overhead. Furthermore, we introduce a novel dataset comprising multi-view, exceptionally long video sequences with substantial rigid and non-rigid motion, thereby substantiating the scalability of our method.
comment: new template, editing
♻ ☆ Hybrid Spatial Representations for Species Distribution Modeling SDM
We address an important problem in ecology called Species Distribution Modeling (SDM), whose goal is to predict whether a species exists at a certain position on Earth. In particular, we tackle a challenging version of this task, where we learn from presence-only data in a community-sourced dataset, model a large number of species simultaneously, and do not use any additional environmental information. Previous work has used neural implicit representations to construct models that achieve promising results. However, implicit representations often generate predictions of limited spatial precision. We attribute this limitation to their inherently global formulation and inability to effectively capture local feature variations. This issue is especially pronounced with presence-only data and a large number of species. To address this, we propose a hybrid embedding scheme that combines both implicit and explicit embeddings. Specifically, the explicit embedding is implemented with a multiresolution hashgrid, enabling our models to better capture local information. Experiments demonstrate that our results exceed other works by a large margin on various standard benchmarks, and that the hybrid representation is better than both purely implicit and explicit ones. Qualitative visualizations and comprehensive ablation studies reveal that our hybrid representation successfully addresses the two main challenges. Our code is open-sourced at https://github.com/Shiran-Yuan/HSR-SDM.
comment: Project codebase https://github.com/Shiran-Yuan/HSR-SDM
Information Retrieval 8
☆ SimRAG: Self-Improving Retrieval-Augmented Generation for Adapting Large Language Models to Specialized Domains
Retrieval-augmented generation (RAG) enhances the question-answering (QA) abilities of large language models (LLMs) by integrating external knowledge. However, adapting general-purpose RAG systems to specialized fields such as science and medicine poses unique challenges due to distribution shifts and limited access to domain-specific data. To tackle this, we propose SimRAG, a self-training approach that equips the LLM with joint capabilities of question answering and question generation for domain adaptation. Our method first fine-tunes the LLM on instruction-following, question-answering, and search-related data. Then, it prompts the same LLM to generate diverse domain-relevant questions from unlabeled corpora, with an additional filtering strategy to retain high-quality synthetic examples. By leveraging these synthetic examples, the LLM can improve their performance on domain-specific RAG tasks. Experiments on 11 datasets, spanning two backbone sizes and three domains, demonstrate that SimRAG outperforms baselines by 1.2\%--8.6\%.
comment: Work in Progress
☆ YOLO-Vehicle-Pro: A Cloud-Edge Collaborative Framework for Object Detection in Autonomous Driving under Adverse Weather Conditions
With the rapid advancement of autonomous driving technology, efficient and accurate object detection capabilities have become crucial factors in ensuring the safety and reliability of autonomous driving systems. However, in low-visibility environments such as hazy conditions, the performance of traditional object detection algorithms often degrades significantly, failing to meet the demands of autonomous driving. To address this challenge, this paper proposes two innovative deep learning models: YOLO-Vehicle and YOLO-Vehicle-Pro. YOLO-Vehicle is an object detection model tailored specifically for autonomous driving scenarios, employing multimodal fusion techniques to combine image and textual information for object detection. YOLO-Vehicle-Pro builds upon this foundation by introducing an improved image dehazing algorithm, enhancing detection performance in low-visibility environments. In addition to model innovation, this paper also designs and implements a cloud-edge collaborative object detection system, deploying models on edge devices and offloading partial computational tasks to the cloud in complex situations. Experimental results demonstrate that on the KITTI dataset, the YOLO-Vehicle-v1s model achieved 92.1% accuracy while maintaining a detection speed of 226 FPS and an inference time of 12ms, meeting the real-time requirements of autonomous driving. When processing hazy images, the YOLO-Vehicle-Pro model achieved a high accuracy of 82.3% mAP@50 on the Foggy Cityscapes dataset while maintaining a detection speed of 43 FPS.
☆ Testing Deep Learning Recommender Systems Models on Synthetic GAN-Generated Datasets
The published method Generative Adversarial Networks for Recommender Systems (GANRS) allows generating data sets for collaborative filtering recommendation systems. The GANRS source code is available along with a representative set of generated datasets. We have tested the GANRS method by creating multiple synthetic datasets from three different real datasets taken as a source. Experiments include variations in the number of users in the synthetic datasets, as well as a different number of samples. We have also selected six state-of-the-art collaborative filtering deep learning models to test both their comparative performance and the GANRS method. The results show a consistent behavior of the generated datasets compared to the source ones; particularly, in the obtained values and trends of the precision and recall quality measures. The tested deep learning models have also performed as expected on all synthetic datasets, making it possible to compare the results with those obtained from the real source data. Future work is proposed, including different cold start scenarios, unbalanced data, and demographic fairness.
comment: 10 pages, 7 figures, In press
☆ Comprehensive Evaluation of Matrix Factorization Models for Collaborative Filtering Recommender Systems
Matrix factorization models are the core of current commercial collaborative filtering Recommender Systems. This paper tested six representative matrix factorization models, using four collaborative filtering datasets. Experiments have tested a variety of accuracy and beyond accuracy quality measures, including prediction, recommendation of ordered and unordered lists, novelty, and diversity. Results show each convenient matrix factorization model attending to their simplicity, the required prediction quality, the necessary recommendation quality, the desired recommendation novelty and diversity, the need to explain recommendations, the adequacy of assigning semantic interpretations to hidden factors, the advisability of recommending to groups of users, and the need to obtain reliability values. To ensure the reproducibility of the experiments, an open framework has been used, and the implementation code is provided.
comment: 10 pages, 5 figures
☆ Extending and Applying Automated HERMES Software Publication Workflows
Research software is an import output of research and must be published according to the FAIR Principles for Research Software. This can be achieved by publishing software with metadata under a persistent identifier. HERMES is a tool that leverages continuous integration to automate the publication of software with rich metadata. In this work, we describe the HERMES workflow itself, and how to extend it to meet the needs of specific research software metadata or infrastructure. We introduce the HERMES plugin architecture and provide the example of creating a new HERMES plugin that harvests metadata from a metadata source in source code repositories. We show how to use HERMES as an end user, both via the command line interface, and as a step in a continuous integration pipeline. Finally, we report three informal case studies whose results provide a preliminary evaluation of the feasibility and applicability of HERMES workflows, and the extensibility of the hermes software package.
comment: 17 pages, 2 figures, 2 tables, submitted to a special issue of Electronic Communications of the EASST collecting submissions of deRSE24, Conference for Research Software Engineers
♻ ☆ LLM-Assisted Multi-Teacher Continual Learning for Visual Question Answering in Robotic Surgery ICRA
Visual question answering (VQA) is crucial for promoting surgical education. In practice, the needs of trainees are constantly evolving, such as learning more surgical types, adapting to different robots, and learning new surgical instruments and techniques for various surgeries. However, patient data privacy often restricts the availability of old data when updating the model, necessitating an exemplar-free continual learning (CL) setup. Prior CL studies overlooked two vital problems in the surgical domain: 1) large domain shifts from diverse surgical operations collected from multiple sources, and 2) severe data imbalance arising from the uneven presence of surgical instruments or activities. This paper proposes addressing these problems with a multimodal large language model (LLM) and an adaptive weight assignment methodology. We first develop a new multi-teacher CL framework that leverages a multimodal LLM as the additional teacher. The strong generalization ability of the LLM can bridge the knowledge gap when domain shifts and data imbalances occur. We then put forth a novel data processing method that transforms complex LLM embeddings into logits compatible with our CL framework. We further design an adaptive weight assignment approach that balances the generalization ability of the LLM and the domain expertise of the old CL model. Finally, to comprehensively test the effectiveness of our proposed method, we have also constructed two new surgical VQA datasets that are largely different from existing ones and could be valuable resources for future research. Extensive experimental results on the tested datasets demonstrate the superiority of our method to other advanced CL schemes.
comment: This paper has been accapted by 2024 IEEE International Conference on Robotics and Automation (ICRA)
♻ ☆ From Keywords to Structured Summaries: Streamlining Scholarly Information Access ISWC 2024
This paper highlights the growing importance of information retrieval (IR) engines in the scientific community, addressing the inefficiency of traditional keyword-based search engines due to the rising volume of publications. The proposed solution involves structured records, underpinning advanced information technology (IT) tools, including visualization dashboards, to revolutionize how researchers access and filter articles, replacing the traditional text-heavy approach. This vision is exemplified through a proof of concept centered on the "reproductive number estimate of infectious diseases" research theme, using a fine-tuned large language model (LLM) to automate the creation of structured records to populate a backend database that now goes beyond keywords. The result is a next-generation information access system as an IR method accessible at https://orkg.org/usecases/r0-estimates.
comment: 8 pages, 3 figures | Accepted for publication as a poster paper at the International Semantic Web Conference (ISWC 2024)
♻ ☆ Post-Training Attribute Unlearning in Recommender Systems
With the growing privacy concerns in recommender systems, recommendation unlearning is getting increasing attention. Existing studies predominantly use training data, i.e., model inputs, as unlearning target. However, attackers can extract private information from the model even if it has not been explicitly encountered during training. We name this unseen information as \textit{attribute} and treat it as unlearning target. To protect the sensitive attribute of users, Attribute Unlearning (AU) aims to make target attributes indistinguishable. In this paper, we focus on a strict but practical setting of AU, namely Post-Training Attribute Unlearning (PoT-AU), where unlearning can only be performed after the training of the recommendation model is completed. To address the PoT-AU problem in recommender systems, we propose a two-component loss function. The first component is distinguishability loss, where we design a distribution-based measurement to make attribute labels indistinguishable from attackers. We further extend this measurement to handle multi-class attribute cases with efficient computational overhead. The second component is regularization loss, where we explore a function-space measurement that effectively maintains recommendation performance compared to parameter-space regularization. We use stochastic gradient descent algorithm to optimize our proposed loss. Extensive experiments on four real-world datasets demonstrate the effectiveness of our proposed methods.
comment: Accepted by TOIS. arXiv admin note: text overlap with arXiv:2310.05847
Machine Learning 192
☆ Prioritized Generative Replay
Sample-efficient online reinforcement learning often uses replay buffers to store experience for reuse when updating the value function. However, uniform replay is inefficient, since certain classes of transitions can be more relevant to learning. While prioritization of more useful samples is helpful, this strategy can also lead to overfitting, as useful samples are likely to be more rare. In this work, we instead propose a prioritized, parametric version of an agent's memory, using generative models to capture online experience. This paradigm enables (1) densification of past experience, with new generations that benefit from the generative model's generalization capacity and (2) guidance via a family of "relevance functions" that push these generations towards more useful parts of an agent's acquired history. We show this recipe can be instantiated using conditional diffusion models and simple relevance functions such as curiosity- or value-based metrics. Our approach consistently improves performance and sample efficiency in both state- and pixel-based domains. We expose the mechanisms underlying these gains, showing how guidance promotes diversity in our generated transitions and reduces overfitting. We also showcase how our approach can train policies with even higher update-to-data ratios than before, opening up avenues to better scale online RL agents.
☆ ALTA: Compiler-Based Analysis of Transformers
We propose a new programming language called ALTA and a compiler that can map ALTA programs to Transformer weights. ALTA is inspired by RASP, a language proposed by Weiss et al. (2021), and Tracr (Lindner et al., 2023), a compiler from RASP programs to Transformer weights. ALTA complements and extends this prior work, offering the ability to express loops and to compile programs to Universal Transformers, among other advantages. ALTA allows us to constructively show how Transformers can represent length-invariant algorithms for computing parity and addition, as well as a solution to the SCAN benchmark of compositional generalization tasks, without requiring intermediate scratchpad decoding steps. We also propose tools to analyze cases where the expressibility of an algorithm is established, but end-to-end training on a given training set fails to induce behavior consistent with the desired algorithm. To this end, we explore training from ALTA execution traces as a more fine-grained supervision signal. This enables additional experiments and theoretical analyses relating the learnability of various algorithms to data availability and modeling decisions, such as positional encodings. We make the ALTA framework -- language specification, symbolic interpreter, and weight compiler -- available to the community to enable further applications and insights.
☆ Leveraging Skills from Unlabeled Prior Data for Efficient Online Exploration
Unsupervised pretraining has been transformative in many supervised domains. However, applying such ideas to reinforcement learning (RL) presents a unique challenge in that fine-tuning does not involve mimicking task-specific data, but rather exploring and locating the solution through iterative self-improvement. In this work, we study how unlabeled prior trajectory data can be leveraged to learn efficient exploration strategies. While prior data can be used to pretrain a set of low-level skills, or as additional off-policy data for online RL, it has been unclear how to combine these ideas effectively for online exploration. Our method SUPE (Skills from Unlabeled Prior data for Exploration) demonstrates that a careful combination of these ideas compounds their benefits. Our method first extracts low-level skills using a variational autoencoder (VAE), and then pseudo-relabels unlabeled trajectories using an optimistic reward model, transforming prior data into high-level, task-relevant examples. Finally, SUPE uses these transformed examples as additional off-policy data for online RL to learn a high-level policy that composes pretrained low-level skills to explore efficiently. We empirically show that SUPE reliably outperforms prior strategies, successfully solving a suite of long-horizon, sparse-reward tasks. Code: https://github.com/rail-berkeley/supe.
comment: 23 pages, 10 figures
☆ ProFL: Performative Robust Optimal Federated Learning
Performative prediction (PP) is a framework that captures distribution shifts that occur during the training of machine learning models due to their deployment. As the trained model is used, its generated data could cause the model to evolve, leading to deviations from the original data distribution. The impact of such model-induced distribution shifts in the federated learning (FL) setup remains unexplored despite being increasingly likely to transpire in real-life use cases. Although Jin et al. (2024) recently extended PP to FL in a straightforward manner, the resulting model only converges to a performative stable point, which may be far from optimal. The methods in Izzo et al. (2021); Miller et al. (2021) can find a performative optimal point in centralized settings, but they require the performative risk to be convex and the training data to be noiseless, assumptions often violated in realistic FL systems. This paper overcomes all of these shortcomings and proposes Performative robust optimal Federated Learning (ProFL), an algorithm that finds performative optimal points in FL from noisy and contaminated data. We present the convergence analysis under the Polyak-Lojasiewicz condition, which applies to non-convex objectives. Extensive experiments on multiple datasets validate our proposed algorithms' efficiency.
comment: 27 pages with Appendix, 18 figures. The paper has been submitted and is currently under review
☆ UnCLe: Unsupervised Continual Learning of Depth Completion
We propose UnCLe, a standardized benchmark for Unsupervised Continual Learning of a multimodal depth estimation task: Depth completion aims to infer a dense depth map from a pair of synchronized RGB image and sparse depth map. We benchmark depth completion models under the practical scenario of unsupervised learning over continuous streams of data. Existing methods are typically trained on a static, or stationary, dataset. However, when adapting to novel non-stationary distributions, they "catastrophically forget" previously learned information. UnCLe simulates these non-stationary distributions by adapting depth completion models to sequences of datasets containing diverse scenes captured from distinct domains using different visual and range sensors. We adopt representative methods from continual learning paradigms and translate them to enable unsupervised continual learning of depth completion. We benchmark these models for indoor and outdoor and investigate the degree of catastrophic forgetting through standard quantitative metrics. Furthermore, we introduce model inversion quality as an additional measure of forgetting. We find that unsupervised continual learning of depth completion is an open problem, and we invite researchers to leverage UnCLe as a development platform.
comment: Preprint
☆ Training Free Guided Flow Matching with Optimal Control
Controlled generation with pre-trained Diffusion and Flow Matching models has vast applications. One strategy for guiding ODE-based generative models is through optimizing a target loss $R(x_1)$ while staying close to the prior distribution. Along this line, some recent work showed the effectiveness of guiding flow model by differentiating through its ODE sampling process. Despite the superior performance, the theoretical understanding of this line of methods is still preliminary, leaving space for algorithm improvement. Moreover, existing methods predominately focus on Euclidean data manifold, and there is a compelling need for guided flow methods on complex geometries such as SO(3), which prevails in high-stake scientific applications like protein design. We present OC-Flow, a general and theoretically grounded training-free framework for guided flow matching using optimal control. Building upon advances in optimal control theory, we develop effective and practical algorithms for solving optimal control in guided ODE-based generation and provide a systematic theoretical analysis of the convergence guarantee in both Euclidean and SO(3). We show that existing backprop-through-ODE methods can be interpreted as special cases of Euclidean OC-Flow. OC-Flow achieved superior performance in extensive experiments on text-guided image manipulation, conditional molecule generation, and all-atom peptide design.
☆ Beyond position: how rotary embeddings shape representations and memory in autoregressive transfomers
Rotary Positional Embeddings (RoPE) enhance positional encoding in Transformer models, yet their full impact on model dynamics remains underexplored. This paper studies how RoPE introduces position-dependent rotations, causing phase shifts in token embeddings that influence higher-frequency components within the model's internal representations. Through spectral analysis, we demonstrate that RoPE's rotation matrices induce oscillatory behaviors in embeddings, affecting information retention across layers and shaping temporal modeling capabilities. We show that activation functions in feed-forward networks interact with RoPE-modulated embeddings to generate harmonics, leading to constructive or destructive interference based on phase alignment. Our findings reveal that phase alignment amplifies activations and sharpens attention, while misalignment weakens activations and disrupts focus on positional patterns. This study underscores the importance of frequency components as intrinsic elements of model behavior, offering new insights beyond traditional analyses.
☆ The Double-Edged Sword of Behavioral Responses in Strategic Classification: Theory and User Studies
When humans are subject to an algorithmic decision system, they can strategically adjust their behavior accordingly (``game'' the system). While a growing line of literature on strategic classification has used game-theoretic modeling to understand and mitigate such gaming, these existing works consider standard models of fully rational agents. In this paper, we propose a strategic classification model that considers behavioral biases in human responses to algorithms. We show how misperceptions of a classifier (specifically, of its feature weights) can lead to different types of discrepancies between biased and rational agents' responses, and identify when behavioral agents over- or under-invest in different features. We also show that strategic agents with behavioral biases can benefit or (perhaps, unexpectedly) harm the firm compared to fully rational strategic agents. We complement our analytical results with user studies, which support our hypothesis of behavioral biases in human responses to the algorithm. Together, our findings highlight the need to account for human (cognitive) biases when designing AI systems, and providing explanations of them, to strategic human in the loop.
☆ SPIRE: Synergistic Planning, Imitation, and Reinforcement Learning for Long-Horizon Manipulation
Robot learning has proven to be a general and effective technique for programming manipulators. Imitation learning is able to teach robots solely from human demonstrations but is bottlenecked by the capabilities of the demonstrations. Reinforcement learning uses exploration to discover better behaviors; however, the space of possible improvements can be too large to start from scratch. And for both techniques, the learning difficulty increases proportional to the length of the manipulation task. Accounting for this, we propose SPIRE, a system that first uses Task and Motion Planning (TAMP) to decompose tasks into smaller learning subproblems and second combines imitation and reinforcement learning to maximize their strengths. We develop novel strategies to train learning agents when deployed in the context of a planning system. We evaluate SPIRE on a suite of long-horizon and contact-rich robot manipulation problems. We find that SPIRE outperforms prior approaches that integrate imitation learning, reinforcement learning, and planning by 35% to 50% in average task performance, is 6 times more data efficient in the number of human demonstrations needed to train proficient agents, and learns to complete tasks nearly twice as efficiently. View https://sites.google.com/view/spire-corl-2024 for more details.
comment: Conference on Robot Learning (CoRL) 2024
☆ POD-Attention: Unlocking Full Prefill-Decode Overlap for Faster LLM Inference
Each request in LLM inference goes through two phases: compute-bound prefill and memory-bandwidth-bound decode. To improve GPU utilization, recent systems use hybrid batching that combines the prefill and decode phases of different requests into the same batch. Hybrid batching works well for linear operations as it amortizes the cost of loading model weights from HBM. However, attention computation in hybrid batches remains inefficient because existing attention kernels are optimized for either prefill or decode. In this paper, we present POD-Attention -- the first GPU kernel that efficiently computes attention for hybrid batches. POD-Attention aims to maximize the utilization of both compute and memory bandwidth by carefully allocating the GPU's resources such that prefill and decode operations happen concurrently on the same multiprocessor. We integrate POD-Attention in a state-of-the-art LLM inference scheduler Sarathi-Serve. POD-Attention speeds up attention computation by up to 75% (mean 28%) and increases LLM serving throughput by up to 22% in offline inference. In online inference, POD-Attention enables lower time-to-first-token (TTFT), time-between-tokens (TBT), and request execution latency versus Sarathi-Serve.
☆ Inferring stability properties of chaotic systems on autoencoders' latent spaces
The data-driven learning of solutions of partial differential equations can be based on a divide-and-conquer strategy. First, the high dimensional data is compressed to a latent space with an autoencoder; and, second, the temporal dynamics are inferred on the latent space with a form of recurrent neural network. In chaotic systems and turbulence, convolutional autoencoders and echo state networks (CAE-ESN) successfully forecast the dynamics, but little is known about whether the stability properties can also be inferred. We show that the CAE-ESN model infers the invariant stability properties and the geometry of the tangent space in the low-dimensional manifold (i.e. the latent space) through Lyapunov exponents and covariant Lyapunov vectors. This work opens up new opportunities for inferring the stability of high-dimensional chaotic systems in latent spaces.
☆ Estimating the Spectral Moments of the Kernel Integral Operator from Finite Sample Matrices
Analyzing the structure of sampled features from an input data distribution is challenging when constrained by limited measurements in both the number of inputs and features. Traditional approaches often rely on the eigenvalue spectrum of the sample covariance matrix derived from finite measurement matrices; however, these spectra are sensitive to the size of the measurement matrix, leading to biased insights. In this paper, we introduce a novel algorithm that provides unbiased estimates of the spectral moments of the kernel integral operator in the limit of infinite inputs and features from finitely sampled measurement matrices. Our method, based upon dynamic programming, is efficient and capable of estimating the moments of the operator spectrum. We demonstrate the accuracy of our estimator on radial basis function (RBF) kernels, highlighting its consistency with the theoretical spectra. Furthermore, we showcase the practical utility and robustness of our method in understanding the geometry of learned representations in neural networks.
☆ Federated Transformer: Multi-Party Vertical Federated Learning on Practical Fuzzily Linked Data
Federated Learning (FL) is an evolving paradigm that enables multiple parties to collaboratively train models without sharing raw data. Among its variants, Vertical Federated Learning (VFL) is particularly relevant in real-world, cross-organizational collaborations, where distinct features of a shared instance group are contributed by different parties. In these scenarios, parties are often linked using fuzzy identifiers, leading to a common practice termed as multi-party fuzzy VFL. Existing models generally address either multi-party VFL or fuzzy VFL between two parties. Extending these models to practical multi-party fuzzy VFL typically results in significant performance degradation and increased costs for maintaining privacy. To overcome these limitations, we introduce the Federated Transformer (FeT), a novel framework that supports multi-party VFL with fuzzy identifiers. FeT innovatively encodes these identifiers into data representations and employs a transformer architecture distributed across different parties, incorporating three new techniques to enhance performance. Furthermore, we have developed a multi-party privacy framework for VFL that integrates differential privacy with secure multi-party computation, effectively protecting local representations while minimizing associated utility costs. Our experiments demonstrate that the FeT surpasses the baseline models by up to 46\% in terms of accuracy when scaled to 50 parties. Additionally, in two-party fuzzy VFL settings, FeT also shows improved performance and privacy over cutting-edge VFL models.
☆ Stick-breaking Attention
The self-attention mechanism traditionally relies on the softmax operator, necessitating positional embeddings like RoPE, or position biases to account for token order. But current methods using still face length generalisation challenges. We propose an alternative attention mechanism based on the stick-breaking process: For each token before the current, we determine a break point $\beta_{i,j}$, which represents the proportion of the remaining stick to allocate to the current token. We repeat the process until the stick is fully allocated, resulting in a sequence of attention weights. This process naturally incorporates recency bias, which has linguistic motivations for grammar parsing (Shen et. al., 2017). We study the implications of replacing the conventional softmax-based attention mechanism with stick-breaking attention. We then discuss implementation of numerically stable stick-breaking attention and adapt Flash Attention to accommodate this mechanism. When used as a drop-in replacement for current softmax+RoPE attention systems, we find that stick-breaking attention performs competitively with current methods on length generalisation and downstream tasks. Stick-breaking also performs well at length generalisation, allowing a model trained with $2^{11}$ context window to perform well at $2^{14}$ with perplexity improvements.
☆ metasnf: Meta Clustering with Similarity Network Fusion in R
metasnf is an R package that enables users to apply meta clustering, a method for efficiently searching a broad space of cluster solutions by clustering the solutions themselves, to clustering workflows based on similarity network fusion (SNF). SNF is a multi-modal data integration algorithm commonly used for biomedical subtype discovery. The package also contains functions to assist with cluster visualization, characterization, and validation. This package can help researchers identify SNF-derived cluster solutions that are guided by context-specific utility over context-agnostic measures of quality.
comment: 72 pages, 22 figures, submitted to Journal of Statistical Software
☆ Optical Generative Models
Generative models cover various application areas, including image, video and music synthesis, natural language processing, and molecular design, among many others. As digital generative models become larger, scalable inference in a fast and energy-efficient manner becomes a challenge. Here, we present optical generative models inspired by diffusion models, where a shallow and fast digital encoder first maps random noise into phase patterns that serve as optical generative seeds for a desired data distribution; a jointly-trained free-space-based reconfigurable decoder all-optically processes these generative seeds to create novel images (never seen before) following the target data distribution. Except for the illumination power and the random seed generation through a shallow encoder, these optical generative models do not consume computing power during the synthesis of novel images. We report the optical generation of monochrome and multi-color novel images of handwritten digits, fashion products, butterflies, and human faces, following the data distributions of MNIST, Fashion MNIST, Butterflies-100, and Celeb-A datasets, respectively, achieving an overall performance comparable to digital neural network-based generative models. To experimentally demonstrate optical generative models, we used visible light to generate, in a snapshot, novel images of handwritten digits and fashion products. These optical generative models might pave the way for energy-efficient, scalable and rapid inference tasks, further exploiting the potentials of optics and photonics for artificial intelligence-generated content.
comment: 24 Pages, 9 Figures
☆ POMDP-Driven Cognitive Massive MIMO Radar: Joint Target Detection-Tracking In Unknown Disturbances
The joint detection and tracking of a moving target embedded in an unknown disturbance represents a key feature that motivates the development of the cognitive radar paradigm. Building upon recent advancements in robust target detection with multiple-input multiple-output (MIMO) radars, this work explores the application of a Partially Observable Markov Decision Process (POMDP) framework to enhance the tracking and detection tasks in a statistically unknown environment. In the POMDP setup, the radar system is considered as an intelligent agent that continuously senses the surrounding environment, optimizing its actions to maximize the probability of detection $(P_D)$ and improve the target position and velocity estimation, all this while keeping a constant probability of false alarm $(P_{FA})$. The proposed approach employs an online algorithm that does not require any apriori knowledge of the noise statistics, and it relies on a much more general observation model than the traditional range-azimuth-elevation model employed by conventional tracking algorithms. Simulation results clearly show substantial performance improvement of the POMDP-based algorithm compared to the State-Action-Reward-State-Action (SARSA)-based one that has been recently investigated in the context of massive MIMO (MMIMO) radar systems.
comment: The paper has been submitted to ieee Transactions on radar systems
☆ A Time-Aware Approach to Early Detection of Anorexia: UNSL at eRisk 2024
The eRisk laboratory aims to address issues related to early risk detection on the Web. In this year's edition, three tasks were proposed, where Task 2 was about early detection of signs of anorexia. Early risk detection is a problem where precision and speed are two crucial objectives. Our research group solved Task 2 by defining a CPI+DMC approach, addressing both objectives independently, and a time-aware approach, where precision and speed are considered a combined single-objective. We implemented the last approach by explicitly integrating time during the learning process, considering the ERDE{\theta} metric as the training objective. It also allowed us to incorporate temporal metrics to validate and select the optimal models. We achieved outstanding results for the ERDE50 metric and ranking-based metrics, demonstrating consistency in solving ERD problems.
comment: In Conference and Labs of the Evaluation Forum (CLEF 2024), Grenoble, France
☆ Closed-form merging of parameter-efficient modules for Federated Continual Learning
Model merging has emerged as a crucial technique in Deep Learning, enabling the integration of multiple models into a unified system while preserving performance and scalability. In this respect, the compositional properties of low-rank adaptation techniques (e.g., LoRA) have proven beneficial, as simple averaging LoRA modules yields a single model that mostly integrates the capabilities of all individual modules. Building on LoRA, we take a step further by imposing that the merged model matches the responses of all learned modules. Solving this objective in closed form yields an indeterminate system with A and B as unknown variables, indicating the existence of infinitely many closed-form solutions. To address this challenge, we introduce LoRM, an alternating optimization strategy that trains one LoRA matrix at a time. This allows solving for each unknown variable individually, thus finding a unique solution. We apply our proposed methodology to Federated Class-Incremental Learning (FCIL), ensuring alignment of model responses both between clients and across tasks. Our method demonstrates state-of-the-art performance across a range of FCIL scenarios.
☆ Medical Imaging Complexity and its Effects on GAN Performance ACCV
The proliferation of machine learning models in diverse clinical applications has led to a growing need for high-fidelity, medical image training data. Such data is often scarce due to cost constraints and privacy concerns. Alleviating this burden, medical image synthesis via generative adversarial networks (GANs) emerged as a powerful method for synthetically generating photo-realistic images based on existing sets of real medical images. However, the exact image set size required to efficiently train such a GAN is unclear. In this work, we experimentally establish benchmarks that measure the relationship between a sample dataset size and the fidelity of the generated images, given the dataset's distribution of image complexities. We analyze statistical metrics based on delentropy, an image complexity measure rooted in Shannon's entropy in information theory. For our pipeline, we conduct experiments with two state-of-the-art GANs, StyleGAN 3 and SPADE-GAN, trained on multiple medical imaging datasets with variable sample sizes. Across both GANs, general performance improved with increasing training set size but suffered with increasing complexity.
comment: Accepted to ACCV, Workshop on Generative AI for Synthetic Medical Data
☆ MCUBERT: Memory-Efficient BERT Inference on Commodity Microcontrollers
In this paper, we propose MCUBERT to enable language models like BERT on tiny microcontroller units (MCUs) through network and scheduling co-optimization. We observe the embedding table contributes to the major storage bottleneck for tiny BERT models. Hence, at the network level, we propose an MCU-aware two-stage neural architecture search algorithm based on clustered low-rank approximation for embedding compression. To reduce the inference memory requirements, we further propose a novel fine-grained MCU-friendly scheduling strategy. Through careful computation tiling and re-ordering as well as kernel design, we drastically increase the input sequence lengths supported on MCUs without any latency or accuracy penalty. MCUBERT reduces the parameter size of BERT-tiny and BERT-mini by 5.7$\times$ and 3.0$\times$ and the execution memory by 3.5$\times$ and 4.3$\times$, respectively. MCUBERT also achieves 1.5$\times$ latency reduction. For the first time, MCUBERT enables lightweight BERT models on commodity MCUs and processing more than 512 tokens with less than 256KB of memory.
comment: ICCAD 2024
☆ SimRAG: Self-Improving Retrieval-Augmented Generation for Adapting Large Language Models to Specialized Domains
Retrieval-augmented generation (RAG) enhances the question-answering (QA) abilities of large language models (LLMs) by integrating external knowledge. However, adapting general-purpose RAG systems to specialized fields such as science and medicine poses unique challenges due to distribution shifts and limited access to domain-specific data. To tackle this, we propose SimRAG, a self-training approach that equips the LLM with joint capabilities of question answering and question generation for domain adaptation. Our method first fine-tunes the LLM on instruction-following, question-answering, and search-related data. Then, it prompts the same LLM to generate diverse domain-relevant questions from unlabeled corpora, with an additional filtering strategy to retain high-quality synthetic examples. By leveraging these synthetic examples, the LLM can improve their performance on domain-specific RAG tasks. Experiments on 11 datasets, spanning two backbone sizes and three domains, demonstrate that SimRAG outperforms baselines by 1.2\%--8.6\%.
comment: Work in Progress
☆ Generalized Resubstitution for Regression Error Estimation
We propose generalized resubstitution error estimators for regression, a broad family of estimators, each corresponding to a choice of empirical probability measures and loss function. The usual sum of squares criterion is a special case corresponding to the standard empirical probability measure and the quadratic loss. Other choices of empirical probability measure lead to more general estimators with superior bias and variance properties. We prove that these error estimators are consistent under broad assumptions. In addition, procedures for choosing the empirical measure based on the method of moments and maximum pseudo-likelihood are proposed and investigated. Detailed experimental results using polynomial regression demonstrate empirically the superior finite-sample bias and variance properties of the proposed estimators. The R code for the experiments is provided.
☆ Theoretically Grounded Pruning of Large Ground Sets for Constrained, Discrete Optimization
Modern instances of combinatorial optimization problems often exhibit billion-scale ground sets, which have many uninformative or redundant elements. In this work, we develop light-weight pruning algorithms to quickly discard elements that are unlikely to be part of an optimal solution. Under mild assumptions on the instance, we prove theoretical guarantees on the fraction of the optimal value retained and the size of the resulting pruned ground set. Through extensive experiments on real-world datasets for various applications, we demonstrate that our algorithm, QuickPrune, efficiently prunes over 90% of the ground set and outperforms state-of-the-art classical and machine learning heuristics for pruning.
☆ Optimizing Travel Itineraries with AI Algorithms in a Microservices Architecture: Balancing Cost, Time, Preferences, and Sustainability
The objective of this research is how an implementation of AI algorithms in the microservices architecture enhances travel itineraries by cost, time, user preferences, and environmental sustainability. It uses machine learning models for both cost forecasting and personalization, genetic algorithm for optimization of the itinerary, and heuristics for sustainability checking. Primary evaluated parameters consist of latency, ability to satisfy user preferences, cost and environmental concern. The experimental results demonstrate an average of 4.5 seconds of response time on 1000 concurrent users and 92% of user preferences accuracy. The cost efficiency is proved, with 95% of provided trips being within the limits of the budget declared by the user. The system also implements some measures to alleviate negative externalities related to travel and 60% of offered travel plans had green options incorporated, resulting in the average 15% lower carbon emissions than the traditional travel plans offered. The genetic algorithm with time complexity O(g.p.f) provides the optimal solution in 100 generations. Every iteration improves the quality of the solution by 5%, thus enabling its effective use in optimization problems where time is measured in seconds. Finally, the system is designed to be fault-tolerant with functional 99.9% availability which allows the provision of services even when requirements are exceeded. Travel optimization platform is turned dynamic and efficient by this microservices based architecture which provides enhanced scaling, allows asynchronous communication and real time changes. Because of the incorporation of Ai, cost control and eco-friendliness approaches, the system addresses the different user needs in the present days travel business.
comment: 18 pages, 6 figures
☆ Spiking Graph Neural Network on Riemannian Manifolds NeurIPS 2024
Graph neural networks (GNNs) have become the dominant solution for learning on graphs, the typical non-Euclidean structures. Conventional GNNs, constructed with the Artificial Neuron Network (ANN), have achieved impressive performance at the cost of high computation and energy consumption. In parallel, spiking GNNs with brain-like spiking neurons are drawing increasing research attention owing to the energy efficiency. So far, existing spiking GNNs consider graphs in Euclidean space, ignoring the structural geometry, and suffer from the high latency issue due to Back-Propagation-Through-Time (BPTT) with the surrogate gradient. In light of the aforementioned issues, we are devoted to exploring spiking GNN on Riemannian manifolds, and present a Manifold-valued Spiking GNN (MSG). In particular, we design a new spiking neuron on geodesically complete manifolds with the diffeomorphism, so that BPTT regarding the spikes is replaced by the proposed differentiation via manifold. Theoretically, we show that MSG approximates a solver of the manifold ordinary differential equation. Extensive experiments on common graphs show the proposed MSG achieves superior performance to previous spiking GNNs and energy efficiency to conventional GNNs.
comment: Accepted by NeurIPS 2024, 30 pages
☆ Semi-Implicit Functional Gradient Flow
Particle-based variational inference methods (ParVIs) use non-parametric variational families represented by particles to approximate the target distribution according to the kernelized Wasserstein gradient flow for the Kullback-Leibler (KL) divergence. Recent works introduce functional gradient flows to substitute the kernel for better flexibility. However, the deterministic updating mechanism may suffer from limited exploration and require expensive repetitive runs for new samples. In this paper, we propose Semi-Implicit Functional Gradient flow (SIFG), a functional gradient ParVI method that uses perturbed particles as the approximation family. The corresponding functional gradient flow, which can be estimated via denoising score matching, exhibits strong theoretical convergence guarantee. We also present an adaptive version of our method to automatically choose the suitable noise magnitude. Extensive experiments demonstrate the effectiveness and efficiency of the proposed framework on both simulated and real data problems.
comment: 31 pages, 12 figures
☆ Retrieving snow depth distribution by downscaling ERA5 Reanalysis with ICESat-2 laser altimetry
Estimating the variability of seasonal snow cover, in particular snow depth in remote areas, poses significant challenges due to limited spatial and temporal data availability. This study uses snow depth measurements from the ICESat-2 satellite laser altimeter, which are sparse in both space and time, and incorporates them with climate reanalysis data into a downscaling-calibration scheme to produce monthly gridded snow depth maps at microscale (10 m). Snow surface elevation measurements from ICESat-2 along profiles are compared to a digital elevation model to determine snow depth at each point. To efficiently turn sparse measurements into snow depth maps, a regression model is fitted to establish a relationship between the retrieved snow depth and the corresponding ERA5 Land snow depth. This relationship, referred to as subgrid variability, is then applied to downscale the monthly ERA5 Land snow depth data. The method can provide timeseries of monthly snow depth maps for the entire ERA5 time range (since 1950). The validation of downscaled snow depth data was performed at an intermediate scale (100 m x 500 m) using datasets from airborne laser scanning (ALS) in the Hardangervidda region of southern Norway. Results show that snow depth prediction achieved R2 values ranging from 0.74 to 0.88 (post-calibration). The method relies on globally available data and is applicable to other snow regions above the treeline. Though requiring area-specific calibration, our approach has the potential to provide snow depth maps in areas where no such data exist and can be used to extrapolate existing snow surveys in time and over larger areas. With this, it can offer valuable input data for hydrological, ecological or permafrost modeling tasks.
☆ Multi-Continental Healthcare Modelling Using Blockchain-Enabled Federated Learning
One of the biggest challenges of building artificial intelligence (AI) model in healthcare area is the data sharing. Since healthcare data is private, sensitive, and heterogeneous, collecting sufficient data for modelling is exhausted, costly, and sometimes impossible. In this paper, we propose a framework for global healthcare modelling using datasets from multi-continents (Europe, North America and Asia) while without sharing the local datasets, and choose glucose management as a study model to verify its effectiveness. Technically, blockchain-enabled federated learning is implemented with adaption to make it meet with the privacy and safety requirements of healthcare data, meanwhile rewards honest participation and penalize malicious activities using its on-chain incentive mechanism. Experimental results show that the proposed framework is effective, efficient, and privacy preserved. Its prediction accuracy is much better than the models trained from limited personal data and is similar to, and even slightly better than, the results from a centralized dataset. This work paves the way for international collaborations on healthcare projects, where additional data is crucial for reducing bias and providing benefits to humanity.
comment: Accepted by IEEE Global Blockchain Conference
☆ Addressing Asynchronicity in Clinical Multimodal Fusion via Individualized Chest X-ray Generation NeurIPS-24
Integrating multi-modal clinical data, such as electronic health records (EHR) and chest X-ray images (CXR), is particularly beneficial for clinical prediction tasks. However, in a temporal setting, multi-modal data are often inherently asynchronous. EHR can be continuously collected but CXR is generally taken with a much longer interval due to its high cost and radiation dose. When clinical prediction is needed, the last available CXR image might have been outdated, leading to suboptimal predictions. To address this challenge, we propose DDL-CXR, a method that dynamically generates an up-to-date latent representation of the individualized CXR images. Our approach leverages latent diffusion models for patient-specific generation strategically conditioned on a previous CXR image and EHR time series, providing information regarding anatomical structures and disease progressions, respectively. In this way, the interaction across modalities could be better captured by the latent CXR generation process, ultimately improving the prediction performance. Experiments using MIMIC datasets show that the proposed model could effectively address asynchronicity in multimodal fusion and consistently outperform existing methods.
comment: Accepted by NeurIPS-24
☆ regAL: Python Package for Active Learning of Regression Problems
Increasingly more research areas rely on machine learning methods to accelerate discovery while saving resources. Machine learning models, however, usually require large datasets of experimental or computational results, which in certain fields, such as (bio)chemistry, materials science, or medicine, are rarely given and often prohibitively expensive to obtain. To bypass that obstacle, active learning methods are employed to develop machine learning models with a desired performance while requiring the least possible number of computational or experimental results from the domain of application. For this purpose, the model's knowledge about certain regions of the application domain is estimated to guide the choice of the model's training set. Although active learning is widely studied for classification problems (discrete outcomes), comparatively few works handle this method for regression problems (continuous outcomes). In this work, we present our Python package regAL, which allows users to evaluate different active learning strategies for regression problems. With a minimal input of just the dataset in question, but many additional customization and insight options, this package is intended for anyone who aims to perform and understand active learning in their problem-specific scope.
☆ Deep learning for model correction of dynamical systems with data scarcity
We present a deep learning framework for correcting existing dynamical system models utilizing only a scarce high-fidelity data set. In many practical situations, one has a low-fidelity model that can capture the dynamics reasonably well but lacks high resolution, due to the inherent limitation of the model and the complexity of the underlying physics. When high resolution data become available, it is natural to seek model correction to improve the resolution of the model predictions. We focus on the case when the amount of high-fidelity data is so small that most of the existing data driven modeling methods cannot be applied. In this paper, we address these challenges with a model-correction method which only requires a scarce high-fidelity data set. Our method first seeks a deep neural network (DNN) model to approximate the existing low-fidelity model. By using the scarce high-fidelity data, the method then corrects the DNN model via transfer learning (TL). After TL, an improved DNN model with high prediction accuracy to the underlying dynamics is obtained. One distinct feature of the propose method is that it does not assume a specific form of the model correction terms. Instead, it offers an inherent correction to the low-fidelity model via TL. A set of numerical examples are presented to demonstrate the effectiveness of the proposed method.
☆ Reinforcement Learning under Latent Dynamics: Toward Statistical and Algorithmic Modularity
Real-world applications of reinforcement learning often involve environments where agents operate on complex, high-dimensional observations, but the underlying (''latent'') dynamics are comparatively simple. However, outside of restrictive settings such as small latent spaces, the fundamental statistical requirements and algorithmic principles for reinforcement learning under latent dynamics are poorly understood. This paper addresses the question of reinforcement learning under $\textit{general}$ latent dynamics from a statistical and algorithmic perspective. On the statistical side, our main negative result shows that most well-studied settings for reinforcement learning with function approximation become intractable when composed with rich observations; we complement this with a positive result, identifying latent pushforward coverability as a general condition that enables statistical tractability. Algorithmically, we develop provably efficient observable-to-latent reductions -- that is, reductions that transform an arbitrary algorithm for the latent MDP into an algorithm that can operate on rich observations -- in two settings: one where the agent has access to hindsight observations of the latent dynamics [LADZ23], and one where the agent can estimate self-predictive latent models [SAGHCB20]. Together, our results serve as a first step toward a unified statistical and algorithmic theory for reinforcement learning under latent dynamics.
☆ Scalable Offline Reinforcement Learning for Mean Field Games AAMAS
Reinforcement learning algorithms for mean-field games offer a scalable framework for optimizing policies in large populations of interacting agents. Existing methods often depend on online interactions or access to system dynamics, limiting their practicality in real-world scenarios where such interactions are infeasible or difficult to model. In this paper, we present Offline Munchausen Mirror Descent (Off-MMD), a novel mean-field RL algorithm that approximates equilibrium policies in mean-field games using purely offline data. By leveraging iterative mirror descent and importance sampling techniques, Off-MMD estimates the mean-field distribution from static datasets without relying on simulation or environment dynamics. Additionally, we incorporate techniques from offline reinforcement learning to address common issues like Q-value overestimation, ensuring robust policy learning even with limited data coverage. Our algorithm scales to complex environments and demonstrates strong performance on benchmark tasks like crowd exploration or navigation, highlighting its applicability to real-world multi-agent systems where online experimentation is infeasible. We empirically demonstrate the robustness of Off-MMD to low-quality datasets and conduct experiments to investigate its sensitivity to hyperparameter choices.
comment: Submitted to AAMAS
☆ Identifiable Representation and Model Learning for Latent Dynamic Systems
Learning identifiable representations and models from low-level observations is useful for an intelligent spacecraft to reliability finish downstream tasks. For temporal observations, to ensure that the data generating process is provably inverted, most existing works either assume the noise variables in the dynamic mechanisms are (conditionally) independent, or require interventions which can directly affect each latent variable. However, in practice, the relationship between the exogenous inputs/interventions and the latent variables may follow some complex deterministic mechanisms. In this work, we study the problem of identifiable representation and model learning for latent dynamic systems. The key idea is that we use an inductive bias inspired by controllable canonical forms, which is invariant, sparse, and input dependent by definition. We prove that, for linear or affine nonlinear latent dynamic systems, it is possible to identify the representations up to scaling and determine the models up to some simple transformations. The results have potential to provide some theoretical guarantees for developing more trustworthy decision-making and control methods for intelligent spacecrafts.
☆ AdaRankGrad: Adaptive Gradient-Rank and Moments for Memory-Efficient LLMs Training and Fine-Tuning
Training and fine-tuning large language models (LLMs) come with challenges related to memory and computational requirements due to the increasing size of the model weights and the optimizer states. Various techniques have been developed to tackle these challenges, such as low-rank adaptation (LoRA), which involves introducing a parallel trainable low-rank matrix to the fixed pre-trained weights at each layer. However, these methods often fall short compared to the full-rank weight training approach, as they restrict the parameter search to a low-rank subspace. This limitation can disrupt training dynamics and require a full-rank warm start to mitigate the impact. In this paper, we introduce a new method inspired by a phenomenon we formally prove: as training progresses, the rank of the estimated layer gradients gradually decreases, and asymptotically approaches rank one. Leveraging this, our approach involves adaptively reducing the rank of the gradients during Adam optimization steps, using an efficient online-updating low-rank projections rule. We further present a randomized SVD scheme for efficiently finding the projection matrix. Our technique enables full-parameter fine-tuning with adaptive low-rank gradient updates, significantly reducing overall memory requirements during training compared to state-of-the-art methods while improving model performance in both pretraining and fine-tuning. Finally, we provide a convergence analysis of our method and demonstrate its merits for training and fine-tuning language and biological foundation models.
☆ Relaxed Equivariance via Multitask Learning
Incorporating equivariance as an inductive bias into deep learning architectures to take advantage of the data symmetry has been successful in multiple applications, such as chemistry and dynamical systems. In particular, roto-translations are crucial for effectively modeling geometric graphs and molecules, where understanding the 3D structures enhances generalization. However, equivariant models often pose challenges due to their high computational complexity. In this paper, we introduce REMUL, a training procedure for approximating equivariance with multitask learning. We show that unconstrained models (which do not build equivariance into the architecture) can learn approximate symmetries by minimizing an additional simple equivariance loss. By formulating equivariance as a new learning objective, we can control the level of approximate equivariance in the model. Our method achieves competitive performance compared to equivariant baselines while being $10 \times$ faster at inference and $2.5 \times$ at training.
☆ Population stratification for prediction of mortality in post-AKI patients
Acute kidney injury (AKI) is a serious clinical condition that affects up to 20% of hospitalised patients. AKI is associated with short term unplanned hospital readmission and post-discharge mortality risk. Patient risk and healthcare expenditures can be minimised by followup planning grounded on predictive models and machine learning. Since AKI is multi-factorial, predictive models specialised in different categories of patients can increase accuracy of predictions. In the present article we present some results following this approach.
☆ CASCRNet: An Atrous Spatial Pyramid Pooling and Shared Channel Residual based Network for Capsule Endoscopy
This manuscript summarizes work on the Capsule Vision Challenge 2024 by MISAHUB. To address the multi-class disease classification task, which is challenging due to the complexity and imbalance in the Capsule Vision challenge dataset, this paper proposes CASCRNet (Capsule endoscopy-Aspp-SCR-Network), a parameter-efficient and novel model that uses Shared Channel Residual (SCR) blocks and Atrous Spatial Pyramid Pooling (ASPP) blocks. Further, the performance of the proposed model is compared with other well-known approaches. The experimental results yield that proposed model provides better disease classification results. The proposed model was successful in classifying diseases with an F1 Score of 78.5% and a Mean AUC of 98.3%, which is promising given its compact architecture.
comment: 8 pages, 4 figures
☆ The Probabilistic Tsetlin Machine: A Novel Approach to Uncertainty Quantification
Tsetlin Machines (TMs) have emerged as a compelling alternative to conventional deep learning methods, offering notable advantages such as smaller memory footprint, faster inference, fault-tolerant properties, and interpretability. Although various adaptations of TMs have expanded their applicability across diverse domains, a fundamental gap remains in understanding how TMs quantify uncertainty in their predictions. In response, this paper introduces the Probabilistic Tsetlin Machine (PTM) framework, aimed at providing a robust, reliable, and interpretable approach for uncertainty quantification. Unlike the original TM, the PTM learns the probability of staying on each state of each Tsetlin Automaton (TA) across all clauses. These probabilities are updated using the feedback tables that are part of the TM framework: Type I and Type II feedback. During inference, TAs decide their actions by sampling states based on learned probability distributions, akin to Bayesian neural networks when generating weight values. In our experimental analysis, we first illustrate the spread of the probabilities across TA states for the noisy-XOR dataset. Then we evaluate the PTM alongside benchmark models using both simulated and real-world datasets. The experiments on the simulated dataset reveal the PTM's effectiveness in uncertainty quantification, particularly in delineating decision boundaries and identifying regions of high uncertainty. Moreover, when applied to multiclass classification tasks using the Iris dataset, the PTM demonstrates competitive performance in terms of predictive entropy and expected calibration error, showcasing its potential as a reliable tool for uncertainty estimation. Our findings underscore the importance of selecting appropriate models for accurate uncertainty quantification in predictive tasks, with the PTM offering a particularly interpretable and effective solution.
comment: 12 pages, 5 figures, 6 tables, accepted and presented at ICAAI 2024, London
☆ Is the GPU Half-Empty or Half-Full? Practical Scheduling Techniques for LLMs
Serving systems for Large Language Models (LLMs) improve throughput by processing several requests concurrently. However, multiplexing hardware resources between concurrent requests involves non-trivial scheduling decisions. Practical serving systems typically implement these decisions at two levels: First, a load balancer routes requests to different servers which each hold a replica of the LLM. Then, on each server, an engine-level scheduler decides when to run a request, or when to queue or preempt it. Improved scheduling policies may benefit a wide range of LLM deployments and can often be implemented as "drop-in replacements" to a system's current policy. In this work, we survey scheduling techniques from the literature and from practical serving systems. We find that schedulers from the literature often achieve good performance but introduce significant complexity. In contrast, schedulers in practical deployments often leave easy performance gains on the table but are easy to implement, deploy and configure. This finding motivates us to introduce two new scheduling techniques, which are both easy to implement, and outperform current techniques on production workload traces.
comment: 12 pages, 11 figures
☆ Optimal Streaming Algorithms for Multi-Armed Bandits
This paper studies two variants of the best arm identification (BAI) problem under the streaming model, where we have a stream of $n$ arms with reward distributions supported on $[0,1]$ with unknown means. The arms in the stream are arriving one by one, and the algorithm cannot access an arm unless it is stored in a limited size memory. We first study the streaming \eps-$top$-$k$ arms identification problem, which asks for $k$ arms whose reward means are lower than that of the $k$-th best arm by at most $\eps$ with probability at least $1-\delta$. For general $\eps \in (0,1)$, the existing solution for this problem assumes $k = 1$ and achieves the optimal sample complexity $O(\frac{n}{\eps^2} \log \frac{1}{\delta})$ using $O(\log^*(n))$ ($\log^*(n)$ equals the number of times that we need to apply the logarithm function on $n$ before the results is no more than 1.) memory and a single pass of the stream. We propose an algorithm that works for any $k$ and achieves the optimal sample complexity $O(\frac{n}{\eps^2} \log\frac{k}{\delta})$ using a single-arm memory and a single pass of the stream. Second, we study the streaming BAI problem, where the objective is to identify the arm with the maximum reward mean with at least $1-\delta$ probability, using a single-arm memory and as few passes of the input stream as possible. We present a single-arm-memory algorithm that achieves a near instance-dependent optimal sample complexity within $O(\log \Delta_2^{-1})$ passes, where $\Delta_2$ is the gap between the mean of the best arm and that of the second best arm.
comment: 24pages
☆ Non-intrusive Speech Quality Assessment with Diffusion Models Trained on Clean Speech
Diffusion models have found great success in generating high quality, natural samples of speech, but their potential for density estimation for speech has so far remained largely unexplored. In this work, we leverage an unconditional diffusion model trained only on clean speech for the assessment of speech quality. We show that the quality of a speech utterance can be assessed by estimating the likelihood of a corresponding sample in the terminating Gaussian distribution, obtained via a deterministic noising process. The resulting method is purely unsupervised, trained only on clean speech, and therefore does not rely on annotations. Our diffusion-based approach leverages clean speech priors to assess quality based on how the input relates to the learned distribution of clean data. Our proposed log-likelihoods show promising results, correlating well with intrusive speech quality metrics such as POLQA and SI-SDR.
☆ Att2CPC: Attention-Guided Lossy Attribute Compression of Point Clouds
With the great progress of 3D sensing and acquisition technology, the volume of point cloud data has grown dramatically, which urges the development of efficient point cloud compression methods. In this paper, we focus on the task of learned lossy point cloud attribute compression (PCAC). We propose an efficient attention-based method for lossy compression of point cloud attributes leveraging on an autoencoder architecture. Specifically, at the encoding side, we conduct multiple downsampling to best exploit the local attribute patterns, in which effective External Cross Attention (ECA) is devised to hierarchically aggregate features by intergrating attributes and geometry contexts. At the decoding side, the attributes of the point cloud are progressively reconstructed based on the multi-scale representation and the zero-padding upsampling tactic. To the best of our knowledge, this is the first approach to introduce attention mechanism to point-based lossy PCAC task. We verify the compression efficiency of our model on various sequences, including human body frames, sparse objects, and large-scale point cloud scenes. Experiments show that our method achieves an average improvement of 1.15 dB and 2.13 dB in BD-PSNR of Y channel and YUV channel, respectively, when comparing with the state-of-the-art point-based method Deep-PCAC. Codes of this paper are available at https://github.com/I2-Multimedia-Lab/Att2CPC.
☆ Learning Lossless Compression for High Bit-Depth Volumetric Medical Image
Recent advances in learning-based methods have markedly enhanced the capabilities of image compression. However, these methods struggle with high bit-depth volumetric medical images, facing issues such as degraded performance, increased memory demand, and reduced processing speed. To address these challenges, this paper presents the Bit-Division based Lossless Volumetric Image Compression (BD-LVIC) framework, which is tailored for high bit-depth medical volume compression. The BD-LVIC framework skillfully divides the high bit-depth volume into two lower bit-depth segments: the Most Significant Bit-Volume (MSBV) and the Least Significant Bit-Volume (LSBV). The MSBV concentrates on the most significant bits of the volumetric medical image, capturing vital structural details in a compact manner. This reduction in complexity greatly improves compression efficiency using traditional codecs. Conversely, the LSBV deals with the least significant bits, which encapsulate intricate texture details. To compress this detailed information effectively, we introduce an effective learning-based compression model equipped with a Transformer-Based Feature Alignment Module, which exploits both intra-slice and inter-slice redundancies to accurately align features. Subsequently, a Parallel Autoregressive Coding Module merges these features to precisely estimate the probability distribution of the least significant bit-planes. Our extensive testing demonstrates that the BD-LVIC framework not only sets new performance benchmarks across various datasets but also maintains a competitive coding speed, highlighting its significant potential and practical utility in the realm of volumetric medical image compression.
comment: 13 pages
☆ A Comprehensive Analysis on the Learning Curve in Kernel Ridge Regression
This paper conducts a comprehensive study of the learning curves of kernel ridge regression (KRR) under minimal assumptions. Our contributions are three-fold: 1) we analyze the role of key properties of the kernel, such as its spectral eigen-decay, the characteristics of the eigenfunctions, and the smoothness of the kernel; 2) we demonstrate the validity of the Gaussian Equivalent Property (GEP), which states that the generalization performance of KRR remains the same when the whitened features are replaced by standard Gaussian vectors, thereby shedding light on the success of previous analyzes under the Gaussian Design Assumption; 3) we derive novel bounds that improve over existing bounds across a broad range of setting such as (in)dependent feature vectors and various combinations of eigen-decay rates in the over/underparameterized regimes.
☆ Enhancing Federated Learning Convergence with Dynamic Data Queue and Data Entropy-driven Participant Selection
Federated Learning (FL) is a decentralized approach for collaborative model training on edge devices. This distributed method of model training offers advantages in privacy, security, regulatory compliance, and cost-efficiency. Our emphasis in this research lies in addressing statistical complexity in FL, especially when the data stored locally across devices is not identically and independently distributed (non-IID). We have observed an accuracy reduction of up to approximately 10\% to 30\%, particularly in skewed scenarios where each edge device trains with only 1 class of data. This reduction is attributed to weight divergence, quantified using the Euclidean distance between device-level class distributions and the population distribution, resulting in a bias term (\(\delta_k\)). As a solution, we present a method to improve convergence in FL by creating a global subset of data on the server and dynamically distributing it across devices using a Dynamic Data queue-driven Federated Learning (DDFL). Next, we leverage Data Entropy metrics to observe the process during each training round and enable reasonable device selection for aggregation. Furthermore, we provide a convergence analysis of our proposed DDFL to justify their viability in practical FL scenarios, aiming for better device selection, a non-sub-optimal global model, and faster convergence. We observe that our approach results in a substantial accuracy boost of approximately 5\% for the MNIST dataset, around 18\% for CIFAR-10, and 20\% for CIFAR-100 with a 10\% global subset of data, outperforming the state-of-the-art (SOTA) aggregation algorithms.
comment: The Journal is submitted to IEEE Transactions in the Internet of Things
☆ Large Language Models Engineer Too Many Simple Features For Tabular Data
Tabular machine learning problems often require time-consuming and labor-intensive feature engineering. Recent efforts have focused on using large language models (LLMs) to capitalize on their potential domain knowledge. At the same time, researchers have observed ethically concerning negative biases in other LLM-related use cases, such as text generation. These developments motivated us to investigate whether LLMs exhibit a bias that negatively impacts the performance of feature engineering. While not ethically concerning, such a bias could hinder practitioners from fully utilizing LLMs for automated data science. Therefore, we propose a method to detect potential biases by detecting anomalies in the frequency of operators (e.g., adding two features) suggested by LLMs when engineering new features. Our experiments evaluate the bias of four LLMs, two big frontier and two small open-source models, across 27 tabular datasets. Our results indicate that LLMs are biased toward simple operators, such as addition, and can fail to utilize more complex operators, such as grouping followed by aggregations. Furthermore, the bias can negatively impact the predictive performance when using LLM-generated features. Our results call for mitigating bias when using LLMs for feature engineering.
comment: Preprint
☆ Scaling Robot Policy Learning via Zero-Shot Labeling with Foundation Models
A central challenge towards developing robots that can relate human language to their perception and actions is the scarcity of natural language annotations in diverse robot datasets. Moreover, robot policies that follow natural language instructions are typically trained on either templated language or expensive human-labeled instructions, hindering their scalability. To this end, we introduce NILS: Natural language Instruction Labeling for Scalability. NILS automatically labels uncurated, long-horizon robot data at scale in a zero-shot manner without any human intervention. NILS combines pretrained vision-language foundation models in order to detect objects in a scene, detect object-centric changes, segment tasks from large datasets of unlabelled interaction data and ultimately label behavior datasets. Evaluations on BridgeV2, Fractal, and a kitchen play dataset show that NILS can autonomously annotate diverse robot demonstrations of unlabeled and unstructured datasets while alleviating several shortcomings of crowdsourced human annotations, such as low data quality and diversity. We use NILS to label over 115k trajectories obtained from over 430 hours of robot data. We open-source our auto-labeling code and generated annotations on our website: http://robottasklabeling.github.io.
comment: Project Website at https://robottasklabeling.github.io/
☆ Locating Information in Large Language Models via Random Matrix Theory
As large language models (LLMs) become central to AI applications, gaining a deeper understanding of their inner workings is increasingly important. In this work, we analyze the weight matrices of pretrained transformer models -- specifically BERT and Llama -- using random matrix theory (RMT) as a zero-information hypothesis. While randomly initialized weights perfectly agree with RMT predictions, deviations emerge after training, allowing us to locate learned structures within the models. We identify layer-type specific behaviors that are consistent across all blocks and architectures considered. By pinpointing regions that deviate from RMT predictions, we highlight areas of feature learning and confirm this through comparisons with the activation covariance matrices of the corresponding layers. Our method provides a diagnostic tool for identifying relevant regions in transformer weights using only the trained matrices. Additionally, we address the ongoing debate regarding the significance of small singular values in the context of fine-tuning and alignment in LLMs. Our findings reveal that, after fine-tuning, small singular values play a crucial role in the models' capabilities, suggesting that removing them in an already aligned transformer can be detrimental, as it may compromise model alignment.
comment: 17 pages, 14 figures
☆ Faster Language Models with Better Multi-Token Prediction Using Tensor Decomposition
We propose a new model for multi-token prediction in transformers, aiming to enhance sampling efficiency without compromising accuracy. Motivated by recent work that predicts the probabilities of subsequent tokens using multiple heads, we connect this approach to rank-$1$ canonical tensor decomposition. By generalizing it to a rank-$r$ canonical probability decomposition, we develop an improved model that predicts multiple tokens simultaneously. This model can also be interpreted as a mixture of experts, allowing us to leverage successful techniques from that domain for efficient and robust training. Importantly, the overall overhead for training and sampling remains low. Our method demonstrates significant improvements in inference speed for both text and code generation tasks, proving particularly beneficial within the self-speculative decoding paradigm. It maintains its effectiveness across various model sizes and training epochs, highlighting its robustness and scalability.
☆ Beyond Backpropagation: Optimization with Multi-Tangent Forward Gradients
The gradients used to train neural networks are typically computed using backpropagation. While an efficient way to obtain exact gradients, backpropagation is computationally expensive, hinders parallelization, and is biologically implausible. Forward gradients are an approach to approximate the gradients from directional derivatives along random tangents computed by forward-mode automatic differentiation. So far, research has focused on using a single tangent per step. This paper provides an in-depth analysis of multi-tangent forward gradients and introduces an improved approach to combining the forward gradients from multiple tangents based on orthogonal projections. We demonstrate that increasing the number of tangents improves both approximation quality and optimization performance across various tasks.
☆ Anomaly Resilient Temporal QoS Prediction using Hypergraph Convoluted Transformer Network
Quality-of-Service (QoS) prediction is a critical task in the service lifecycle, enabling precise and adaptive service recommendations by anticipating performance variations over time in response to evolving network uncertainties and user preferences. However, contemporary QoS prediction methods frequently encounter data sparsity and cold-start issues, which hinder accurate QoS predictions and limit the ability to capture diverse user preferences. Additionally, these methods often assume QoS data reliability, neglecting potential credibility issues such as outliers and the presence of greysheep users and services with atypical invocation patterns. Furthermore, traditional approaches fail to leverage diverse features, including domain-specific knowledge and complex higher-order patterns, essential for accurate QoS predictions. In this paper, we introduce a real-time, trust-aware framework for temporal QoS prediction to address the aforementioned challenges, featuring an end-to-end deep architecture called the Hypergraph Convoluted Transformer Network (HCTN). HCTN combines a hypergraph structure with graph convolution over hyper-edges to effectively address high-sparsity issues by capturing complex, high-order correlations. Complementing this, the transformer network utilizes multi-head attention along with parallel 1D convolutional layers and fully connected dense blocks to capture both fine-grained and coarse-grained dynamic patterns. Additionally, our approach includes a sparsity-resilient solution for detecting greysheep users and services, incorporating their unique characteristics to improve prediction accuracy. Trained with a robust loss function resistant to outliers, HCTN demonstrated state-of-the-art performance on the large-scale WSDREAM-2 datasets for response time and throughput.
comment: 16 pages, 12 figures
☆ Topology meets Machine Learning: An Introduction using the Euler Characteristic Transform
This overview article makes the case for how topological concepts can enrich research in machine learning. Using the Euler Characteristic Transform (ECT), a geometrical-topological invariant, as a running example, I present different use cases that result in more efficient models for analyzing point clouds, graphs, and meshes. Moreover, I outline a vision for how topological concepts could be used in the future, comprising (1) the learning of functions on topological spaces, (2) the building of hybrid models that imbue neural networks with knowledge about the topological information in data, and (3) the analysis of qualitative properties of neural networks. With current research already addressing some of these aspects, this article thus serves as an introduction and invitation to this nascent area of research.
☆ Escaping the Forest: Sparse Interpretable Neural Networks for Tabular Data
Tabular datasets are widely used in scientific disciplines such as biology. While these disciplines have already adopted AI methods to enhance their findings and analysis, they mainly use tree-based methods due to their interpretability. At the same time, artificial neural networks have been shown to offer superior flexibility and depth for rich and complex non-tabular problems, but they are falling behind tree-based models for tabular data in terms of performance and interpretability. Although sparsity has been shown to improve the interpretability and performance of ANN models for complex non-tabular datasets, enforcing sparsity structurally and formatively for tabular data before training the model, remains an open question. To address this question, we establish a method that infuses sparsity in neural networks by utilising attention mechanisms to capture the features' importance in tabular datasets. We show that our models, Sparse TABular NET or sTAB-Net with attention mechanisms, are more effective than tree-based models, reaching the state-of-the-art on biological datasets. They further permit the extraction of insights from these datasets and achieve better performance than post-hoc methods like SHAP.
☆ VISAGE: Video Synthesis using Action Graphs for Surgery MICCAI 2024
Surgical data science (SDS) is a field that analyzes patient data before, during, and after surgery to improve surgical outcomes and skills. However, surgical data is scarce, heterogeneous, and complex, which limits the applicability of existing machine learning methods. In this work, we introduce the novel task of future video generation in laparoscopic surgery. This task can augment and enrich the existing surgical data and enable various applications, such as simulation, analysis, and robot-aided surgery. Ultimately, it involves not only understanding the current state of the operation but also accurately predicting the dynamic and often unpredictable nature of surgical procedures. Our proposed method, VISAGE (VIdeo Synthesis using Action Graphs for Surgery), leverages the power of action scene graphs to capture the sequential nature of laparoscopic procedures and utilizes diffusion models to synthesize temporally coherent video sequences. VISAGE predicts the future frames given only a single initial frame, and the action graph triplets. By incorporating domain-specific knowledge through the action graph, VISAGE ensures the generated videos adhere to the expected visual and motion patterns observed in real laparoscopic procedures. The results of our experiments demonstrate high-fidelity video generation for laparoscopy procedures, which enables various applications in SDS.
comment: Accepted at MICCAI 2024 Embodied AI and Robotics for HealTHcare (EARTH) Workshop
☆ Can Uncertainty Quantification Enable Better Learning-based Index Tuning?
Index tuning is crucial for optimizing database performance by selecting optimal indexes based on workload. The key to this process lies in an accurate and efficient benefit estimator. Traditional methods relying on what-if tools often suffer from inefficiency and inaccuracy. In contrast, learning-based models provide a promising alternative but face challenges such as instability, lack of interpretability, and complex management. To overcome these limitations, we adopt a novel approach: quantifying the uncertainty in learning-based models' results, thereby combining the strengths of both traditional and learning-based methods for reliable index tuning. We propose Beauty, the first uncertainty-aware framework that enhances learning-based models with uncertainty quantification and uses what-if tools as a complementary mechanism to improve reliability and reduce management complexity. Specifically, we introduce a novel method that combines AutoEncoder and Monte Carlo Dropout to jointly quantify uncertainty, tailored to the characteristics of benefit estimation tasks. In experiments involving sixteen models, our approach outperformed existing uncertainty quantification methods in the majority of cases. We also conducted index tuning tests on six datasets. By applying the Beauty framework, we eliminated worst-case scenarios and more than tripled the occurrence of best-case scenarios.
comment: 14 pages, 11 figures
☆ Learning Versatile Skills with Curriculum Masking NeurIPS 2024
Masked prediction has emerged as a promising pretraining paradigm in offline reinforcement learning (RL) due to its versatile masking schemes, enabling flexible inference across various downstream tasks with a unified model. Despite the versatility of masked prediction, it remains unclear how to balance the learning of skills at different levels of complexity. To address this, we propose CurrMask, a curriculum masking pretraining paradigm for sequential decision making. Motivated by how humans learn by organizing knowledge in a curriculum, CurrMask adjusts its masking scheme during pretraining for learning versatile skills. Through extensive experiments, we show that CurrMask exhibits superior zero-shot performance on skill prompting tasks, goal-conditioned planning tasks, and competitive finetuning performance on offline RL tasks. Additionally, our analysis of training dynamics reveals that CurrMask gradually acquires skills of varying complexity by dynamically adjusting its masking scheme.
comment: NeurIPS 2024 poster, 21 pages, 7 figures
☆ Continual Learning on a Data Diet
Continual Learning (CL) methods usually learn from all available data. However, this is not the case in human cognition which efficiently focuses on key experiences while disregarding the redundant information. Similarly, not all data points in a dataset have equal potential; some can be more informative than others. This disparity may significantly impact the performance, as both the quality and quantity of samples directly influence the model's generalizability and efficiency. Drawing inspiration from this, we explore the potential of learning from important samples and present an empirical study for evaluating coreset selection techniques in the context of CL to stimulate research in this unexplored area. We train different continual learners on increasing amounts of selected samples and investigate the learning-forgetting dynamics by shedding light on the underlying mechanisms driving their improved stability-plasticity balance. We present several significant observations: learning from selectively chosen samples (i) enhances incremental accuracy, (ii) improves knowledge retention of previous tasks, and (iii) refines learned representations. This analysis contributes to a deeper understanding of selective learning strategies in CL scenarios.
comment: 18 pages, 6 figures
☆ Beware of Calibration Data for Pruning Large Language Models
As large language models (LLMs) are widely applied across various fields, model compression has become increasingly crucial for reducing costs and improving inference efficiency. Post-training pruning is a promising method that does not require resource-intensive iterative training and only needs a small amount of calibration data to assess the importance of parameters. Previous research has primarily focused on designing advanced pruning methods, while different calibration data's impact on pruning performance still lacks systematical exploration. We fill this blank and surprisingly observe that the effects of calibration data even value more than designing advanced pruning strategies, especially for high sparsity. Our preliminary exploration also discloses that using calibration data similar to the training data can yield better performance. As pre-training data is usually inaccessible for advanced LLMs, we further provide a self-generating calibration data synthesis strategy to construct feasible calibration data. We conduct experiments on the recent strong open-source LLMs (e.g., DCLM, and LLaMA-3), and the results show that the proposed method outperforms commonly used calibration data and can effectively enhance strong pruning methods (e.g., Wanda, OWL).
comment: under review
☆ Scalable Random Feature Latent Variable Models
Random feature latent variable models (RFLVMs) represent the state-of-the-art in latent variable models, capable of handling non-Gaussian likelihoods and effectively uncovering patterns in high-dimensional data. However, their heavy reliance on Monte Carlo sampling results in scalability issues which makes it difficult to use these models for datasets with a massive number of observations. To scale up RFLVMs, we turn to the optimization-based variational Bayesian inference (VBI) algorithm which is known for its scalability compared to sampling-based methods. However, implementing VBI for RFLVMs poses challenges, such as the lack of explicit probability distribution functions (PDFs) for the Dirichlet process (DP) in the kernel learning component, and the incompatibility of existing VBI algorithms with RFLVMs. To address these issues, we introduce a stick-breaking construction for DP to obtain an explicit PDF and a novel VBI algorithm called ``block coordinate descent variational inference" (BCD-VI). This enables the development of a scalable version of RFLVMs, or in short, SRFLVM. Our proposed method shows scalability, computational efficiency, superior performance in generating informative latent representations and the ability of imputing missing data across various real-world datasets, outperforming state-of-the-art competitors.
☆ Optimizing Load Scheduling in Power Grids Using Reinforcement Learning and Markov Decision Processes
Power grid load scheduling is a critical task that ensures the balance between electricity generation and consumption while minimizing operational costs and maintaining grid stability. Traditional optimization methods often struggle with the dynamic and stochastic nature of power systems, especially when faced with renewable energy sources and fluctuating demand. This paper proposes a reinforcement learning (RL) approach using a Markov Decision Process (MDP) framework to address the challenges of dynamic load scheduling. The MDP is defined by a state space representing grid conditions, an action space covering control operations like generator adjustments and storage management, and a reward function balancing economic efficiency and system reliability. We investigate the application of various RL algorithms, from basic Q-Learning to more advanced Deep Q-Networks (DQN) and Actor-Critic methods, to determine optimal scheduling policies. The proposed approach is evaluated through a simulated power grid environment, demonstrating its potential to improve scheduling efficiency and adapt to variable demand patterns. Our results show that the RL-based method provides a robust and scalable solution for real-time load scheduling, contributing to the efficient management of modern power grids.
☆ PETAH: Parameter Efficient Task Adaptation for Hybrid Transformers in a resource-limited Context
Following their success in natural language processing (NLP), there has been a shift towards transformer models in computer vision. While transformers perform well and offer promising multi-tasking performance, due to their high compute requirements, many resource-constrained applications still rely on convolutional or hybrid models that combine the benefits of convolution and attention layers and achieve the best results in the sub 100M parameter range. Simultaneously, task adaptation techniques that allow for the use of one shared transformer backbone for multiple downstream tasks, resulting in great storage savings at negligible cost in performance, have not yet been adopted for hybrid transformers. In this work, we investigate how to achieve the best task-adaptation performance and introduce PETAH: Parameter Efficient Task Adaptation for Hybrid Transformers. We further combine PETAH adaptation with pruning to achieve highly performant and storage friendly models for multi-tasking. In our extensive evaluation on classification and other vision tasks, we demonstrate that our PETAH-adapted hybrid models outperform established task-adaptation techniques for ViTs while requiring fewer parameters and being more efficient on mobile hardware.
☆ Mapping the Media Landscape: Predicting Factual Reporting and Political Bias Through Web Interactions
Bias assessment of news sources is paramount for professionals, organizations, and researchers who rely on truthful evidence for information gathering and reporting. While certain bias indicators are discernible from content analysis, descriptors like political bias and fake news pose greater challenges. In this paper, we propose an extension to a recently presented news media reliability estimation method that focuses on modeling outlets and their longitudinal web interactions. Concretely, we assess the classification performance of four reinforcement learning strategies on a large news media hyperlink graph. Our experiments, targeting two challenging bias descriptors, factual reporting and political bias, showed a significant performance improvement at the source media level. Additionally, we validate our methods on the CLEF 2023 CheckThat! Lab challenge, outperforming the reported results in both, F1-score and the official MAE metric. Furthermore, we contribute by releasing the largest annotated dataset of news source media, categorized with factual reporting and political bias labels. Our findings suggest that profiling news media sources based on their hyperlink interactions over time is feasible, offering a bird's-eye view of evolving media landscapes.
comment: Accepted to CLEF 2024
☆ Towards Active Participant-Centric Vertical Federated Learning: Some Representations May Be All You Need
Vertical Federated Learning (VFL) enables collaborative model training across different participants with distinct features and common samples, while preserving data privacy. Existing VFL methodologies often struggle with realistic data partitions, typically incurring high communication costs and significant operational complexity. In this work, we introduce a novel simplified approach to VFL, Active Participant-Centric VFL (APC-VFL), that, to the best of our knowledge, is the first to require only a single communication round between participants, and allows the active participant to do inference in a non collaborative fashion. This method integrates unsupervised representation learning with knowledge distillation to achieve comparable accuracy to traditional VFL methods based on vertical split learning in classical settings, reducing required communication rounds by up to $4200\times$, while being more flexible. Our approach also shows improvements compared to non-federated local models, as well as a comparable VFL proposal, VFedTrans, offering an efficient and flexible solution for collaborative learning.
☆ Entity-based Reinforcement Learning for Autonomous Cyber Defence CCS 2024
A significant challenge for autonomous cyber defence is ensuring a defensive agent's ability to generalise across diverse network topologies and configurations. This capability is necessary for agents to remain effective when deployed in dynamically changing environments, such as an enterprise network where devices may frequently join and leave. Standard approaches to deep reinforcement learning, where policies are parameterised using a fixed-input multi-layer perceptron (MLP) expect fixed-size observation and action spaces. In autonomous cyber defence, this makes it hard to develop agents that generalise to environments with network topologies different from those trained on, as the number of nodes affects the natural size of the observation and action spaces. To overcome this limitation, we reframe the problem of autonomous network defence using entity-based reinforcement learning, where the observation and action space of an agent are decomposed into a collection of discrete entities. This framework enables the use of policy parameterisations specialised in compositional generalisation. Namely, we train a Transformer-based policy on the Yawning Titan cyber-security simulation environment and test its generalisation capabilities across various network topologies. We demonstrate that this approach significantly outperforms an MLP-based policy on fixed networks, and has the ability for zero-shot generalisation to networks of a different size to those seen in training. These findings highlight the potential for entity-based reinforcement learning to advance the field of autonomous cyber defence by providing more generalisable policies capable of handling variations in real-world network environments.
comment: Material to appear in the proceedings of the 1st International Workshop on Autonomous Cybersecurity at ACM CCS 2024
☆ Exploring structure diversity in atomic resolution microscopy with graph neural networks
The emergence of deep learning (DL) has provided great opportunities for the high-throughput analysis of atomic-resolution micrographs. However, the DL models trained by image patches in fixed size generally lack efficiency and flexibility when processing micrographs containing diversified atomic configurations. Herein, inspired by the similarity between the atomic structures and graphs, we describe a few-shot learning framework based on an equivariant graph neural network (EGNN) to analyze a library of atomic structures (e.g., vacancies, phases, grain boundaries, doping, etc.), showing significantly promoted robustness and three orders of magnitude reduced computing parameters compared to the image-driven DL models, which is especially evident for those aggregated vacancy lines with flexible lattice distortion. Besides, the intuitiveness of graphs enables quantitative and straightforward extraction of the atomic-scale structural features in batches, thus statistically unveiling the self-assembly dynamics of vacancy lines under electron beam irradiation. A versatile model toolkit is established by integrating EGNN sub-models for single structure recognition to process images involving varied configurations in the form of a task chain, leading to the discovery of novel doping configurations with superior electrocatalytic properties for hydrogen evolution reactions. This work provides a powerful tool to explore structure diversity in a fast, accurate, and intelligent manner.
☆ Feature Learning in Attention Mechanisms Is More Compact and Stable Than in Convolution
Attention and convolution are fundamental techniques in machine learning. While they use different approaches to learn features - attention mechanisms capture both global and local data relathionships, while convolutional layers focus on local patterns - both methods are effective for various tasks. Although the feature learning of both models is well-studied individually, there has not been a direct comparison of their feature learning dynamics. In this paper, we compare their Lipschitz continuity with respect to the Wasserstein distance and covering numbers under similar settings. We demonstrate that attention processes data in a more compact and stable manner. Compactness refers to the lower variance and intrinsic dimensionality of the activation outputs, while stability refers to the changes between inputs and outputs. We validate our findings through experiments using topological data analysis, measuring the 1-, 2-, and infinity-Wasserstein distances between the outputs of each layer from both models. Furthermore, we extend our comparison to Vision Transformers (ViTs) and ResNets, showing that while ViTs have higher output variance, their feature learning is more stable than that of ResNets.
☆ Incremental Learning of Affordances using Markov Logic Networks
Affordances enable robots to have a semantic understanding of their surroundings. This allows them to have more acting flexibility when completing a given task. Capturing object affordances in a machine learning model is a difficult task, because of their dependence on contextual information. Markov Logic Networks (MLN) combine probabilistic reasoning with logic that is able to capture such context. Mobile robots operate in partially known environments wherein unseen object affordances can be observed. This new information must be incorporated into the existing knowledge, without having to retrain the MLN from scratch. We introduce the MLN Cumulative Learning Algorithm (MLN-CLA). MLN-CLA learns new relations in various knowledge domains by retaining knowledge and only updating the changed knowledge, for which the MLN is retrained. We show that MLN-CLA is effective for accumulative learning and zero-shot affordance inference, outperforming strong baselines.
comment: accepted at IEEE IRC 2024
Self-Supervised Graph Neural Networks for Enhanced Feature Extraction in Heterogeneous Information Networks
This paper explores the applications and challenges of graph neural networks (GNNs) in processing complex graph data brought about by the rapid development of the Internet. Given the heterogeneity and redundancy problems that graph data often have, traditional GNN methods may be overly dependent on the initial structure and attribute information of the graph, which limits their ability to accurately simulate more complex relationships and patterns in the graph. Therefore, this study proposes a graph neural network model under a self-supervised learning framework, which can flexibly combine different types of additional information of the attribute graph and its nodes, so as to better mine the deep features in the graph data. By introducing a self-supervisory mechanism, it is expected to improve the adaptability of existing models to the diversity and complexity of graph data and improve the overall performance of the model.
☆ A Kernel Perspective on Distillation-based Collaborative Learning NeurIPS 2024
Over the past decade, there is a growing interest in collaborative learning that can enhance AI models of multiple parties. However, it is still challenging to enhance performance them without sharing private data and models from individual parties. One recent promising approach is to develop distillation-based algorithms that exploit unlabeled public data but the results are still unsatisfactory in both theory and practice. To tackle this problem, we rigorously analyze a representative distillation-based algorithm in the view of kernel regression. This work provides the first theoretical results to prove the (nearly) minimax optimality of the nonparametric collaborative learning algorithm that does not directly share local data or models in massively distributed statistically heterogeneous environments. Inspired by our theoretical results, we also propose a practical distillation-based collaborative learning algorithm based on neural network architecture. Our algorithm successfully bridges the gap between our theoretical assumptions and practical settings with neural networks through feature kernel matching. We simulate various regression tasks to verify our theory and demonstrate the practical feasibility of our proposed algorithm.
comment: Accepted to NeurIPS 2024
☆ Challenge on Sound Scene Synthesis: Evaluating Text-to-Audio Generation NeurIPS 2024
Despite significant advancements in neural text-to-audio generation, challenges persist in controllability and evaluation. This paper addresses these issues through the Sound Scene Synthesis challenge held as part of the Detection and Classification of Acoustic Scenes and Events 2024. We present an evaluation protocol combining objective metric, namely Fr\'echet Audio Distance, with perceptual assessments, utilizing a structured prompt format to enable diverse captions and effective evaluation. Our analysis reveals varying performance across sound categories and model architectures, with larger models generally excelling but innovative lightweight approaches also showing promise. The strong correlation between objective metrics and human ratings validates our evaluation approach. We discuss outcomes in terms of audio quality, controllability, and architectural considerations for text-to-audio synthesizers, providing direction for future research.
comment: accepted to NeurIPS 2024 Workshop: Audio Imagination
☆ Predicting Company Growth by Econophysics informed Machine Learning
Predicting company growth is crucial for strategic adjustment, operational decision-making, risk assessment, and loan eligibility reviews. Traditional models for company growth often focus too much on theory, overlooking practical forecasting, or they rely solely on time series forecasting techniques, ignoring interpretability and the inherent mechanisms of company growth. In this paper, we propose a machine learning-based prediction framework that incorporates an econophysics model for company growth. Our model captures both the intrinsic growth mechanisms of companies led by scaling laws and the fluctuations influenced by random factors and individual decisions, demonstrating superior predictive performance compared with methods that use time series techniques alone. Its advantages are more pronounced in long-range prediction tasks. By explicitly modeling the baseline growth and volatility components, our model is more interpretable.
comment: 18 pages, 12 figures
☆ Bonsai: Gradient-free Graph Distillation for Node Classification
Graph distillation has emerged as a promising avenue to enable scalable training of GNNs by compressing the training dataset while preserving essential graph characteristics. Our study uncovers significant shortcomings in current graph distillation techniques. First, the majority of the algorithms paradoxically require training on the full dataset to perform distillation. Second, due to their gradient-emulating approach, these methods require fresh distillation for any change in hyperparameters or GNN architecture, limiting their flexibility and reusability. Finally, they fail to achieve substantial size reduction due to synthesizing fully-connected, edge-weighted graphs. To address these challenges, we present Bonsai, a novel graph distillation method empowered by the observation that \textit{computation trees} form the fundamental processing units of message-passing GNNs. Bonsai distills datasets by encoding a careful selection of \textit{exemplar} trees that maximize the representation of all computation trees in the training set. This unique approach imparts Bonsai as the first linear-time, model-agnostic graph distillation algorithm for node classification that outperforms existing baselines across $6$ real-world datasets on accuracy, while being $22$ times faster on average. Bonsai is grounded in rigorous mathematical guarantees on the adopted approximation strategies making it robust to GNN architectures, datasets, and parameters.
☆ Adversarial Domain Adaptation for Metal Cutting Sound Detection: Leveraging Abundant Lab Data for Scarce Industry Data
Cutting state monitoring in the milling process is crucial for improving manufacturing efficiency and tool life. Cutting sound detection using machine learning (ML) models, inspired by experienced machinists, can be employed as a cost-effective and non-intrusive monitoring method in a complex manufacturing environment. However, labeling industry data for training is costly and time-consuming. Moreover, industry data is often scarce. In this study, we propose a novel adversarial domain adaptation (DA) approach to leverage abundant lab data to learn from scarce industry data, both labeled, for training a cutting-sound detection model. Rather than adapting the features from separate domains directly, we project them first into two separate latent spaces that jointly work as the feature space for learning domain-independent representations. We also analyze two different mechanisms for adversarial learning where the discriminator works as an adversary and a critic in separate settings, enabling our model to learn expressive domain-invariant and domain-ingrained features, respectively. We collected cutting sound data from multiple sensors in different locations, prepared datasets from lab and industry domain, and evaluated our learning models on them. Experiments showed that our models outperformed the multi-layer perceptron based vanilla domain adaptation models in labeling tasks on the curated datasets, achieving near 92%, 82% and 85% accuracy respectively for three different sensors installed in industry settings.
comment: 8 pages, 3 figures, 3 tables, First two named Authors have equal contribution (Co-first author)
☆ Securing Federated Learning Against Novel and Classic Backdoor Threats During Foundation Model Integration
Federated learning (FL) enables decentralized model training while preserving privacy. Recently, integrating Foundation Models (FMs) into FL has boosted performance but also introduced a novel backdoor attack mechanism. Attackers can exploit the FM's capabilities to embed backdoors into synthetic data generated by FMs used for model fusion, subsequently infecting all client models through knowledge sharing without involvement in the long-lasting FL process. These novel attacks render existing FL backdoor defenses ineffective, as they primarily detect anomalies among client updates, which may appear uniformly malicious under this attack. Our work proposes a novel data-free defense strategy by constraining abnormal activations in the hidden feature space during model aggregation on the server. The activation constraints, optimized using synthetic data alongside FL training, mitigate the attack while barely affecting model performance, as the parameters remain untouched. Extensive experiments demonstrate its effectiveness against both novel and classic backdoor attacks, outperforming existing defenses while maintaining model performance.
☆ Differentially Private Learning Needs Better Model Initialization and Self-Distillation
Differentially private SGD (DPSGD) enables privacy-preserving training of language models, but often reduces utility, diversity, and linguistic quality. We introduce DPRefine, a three-phase method that initializes a model using data synthesis from a small pre-trained LM with rigorous filtering, applies DP finetuning on private data, and performs self-distillation to refine outputs. This approach significantly outperforms vanilla DPSGD, with AlpacaEval preferring DPRefine's generations in 78.4% of cases across all datasets. Our analysis reveals that DPRefine reduces linguistic errors in generated text by 84.0%, mitigating grammar and spelling errors, commonly associated with DPSGD. It also reduces inconsistencies of non-private models, such as hallucinated details and misattributed quotes. We find that small models like GPT-2 can be effective for initialization and distillation, highlighting their potential in enabling scalable and efficient deployment of privacy-preserving language.
comment: 18 pages
☆ DisenGCD: A Meta Multigraph-assisted Disentangled Graph Learning Framework for Cognitive Diagnosis NeurIPS 2024
Existing graph learning-based cognitive diagnosis (CD) methods have made relatively good results, but their student, exercise, and concept representations are learned and exchanged in an implicit unified graph, which makes the interaction-agnostic exercise and concept representations be learned poorly, failing to provide high robustness against noise in students' interactions. Besides, lower-order exercise latent representations obtained in shallow layers are not well explored when learning the student representation. To tackle the issues, this paper suggests a meta multigraph-assisted disentangled graph learning framework for CD (DisenGCD), which learns three types of representations on three disentangled graphs: student-exercise-concept interaction, exercise-concept relation, and concept dependency graphs, respectively. Specifically, the latter two graphs are first disentangled from the interaction graph. Then, the student representation is learned from the interaction graph by a devised meta multigraph learning module; multiple learnable propagation paths in this module enable current student latent representation to access lower-order exercise latent representations, which can lead to more effective nad robust student representations learned; the exercise and concept representations are learned on the relation and dependency graphs by graph attention modules. Finally, a novel diagnostic function is devised to handle three disentangled representations for prediction. Experiments show better performance and robustness of DisenGCD than state-of-the-art CD methods and demonstrate the effectiveness of the disentangled learning framework and meta multigraph module. The source code is available at \textcolor{red}{\url{https://github.com/BIMK/Intelligent-Education/tree/main/DisenGCD}}.
comment: 21 pages, Accepted by NeurIPS 2024 as a poster
☆ BlurryScope: a cost-effective and compact scanning microscope for automated HER2 scoring using deep learning on blurry image data
We developed a rapid scanning optical microscope, termed "BlurryScope", that leverages continuous image acquisition and deep learning to provide a cost-effective and compact solution for automated inspection and analysis of tissue sections. BlurryScope integrates specialized hardware with a neural network-based model to quickly process motion-blurred histological images and perform automated pathology classification. This device offers comparable speed to commercial digital pathology scanners, but at a significantly lower price point and smaller size/weight, making it ideal for fast triaging in small clinics, as well as for resource-limited settings. To demonstrate the proof-of-concept of BlurryScope, we implemented automated classification of human epidermal growth factor receptor 2 (HER2) scores on immunohistochemically (IHC) stained breast tissue sections, achieving concordant results with those obtained from a high-end digital scanning microscope. We evaluated this approach by scanning HER2-stained tissue microarrays (TMAs) at a continuous speed of 5 mm/s, which introduces bidirectional motion blur artifacts. These compromised images were then used to train our network models. Using a test set of 284 unique patient cores, we achieved blind testing accuracies of 79.3% and 89.7% for 4-class (0, 1+, 2+, 3+) and 2-class (0/1+ , 2+/3+) HER2 score classification, respectively. BlurryScope automates the entire workflow, from image scanning to stitching and cropping of regions of interest, as well as HER2 score classification. We believe BlurryScope has the potential to enhance the current pathology infrastructure in resource-scarce environments, save diagnostician time and bolster cancer identification and classification across various clinical environments.
comment: 18 Pages, 6 Figures
☆ LEADS: Lightweight Embedded Assisted Driving System
With the rapid development of electric vehicles, formula races that face high school and university students have become more popular than ever as the threshold for design and manufacturing has been lowered. In many cases, we see teams inspired by or directly using toolkits and technologies inherited from standardized commercial vehicles. These architectures are usually overly complicated for amateur applications like the races. In order to improve the efficiency and simplify the development of instrumentation, control, and analysis systems, we propose LEADS (Lightweight Embedded Assisted Driving System), a dedicated solution for such scenarios.
comment: 12 pages
☆ Multimodal Information Bottleneck for Deep Reinforcement Learning with Multiple Sensors
Reinforcement learning has achieved promising results on robotic control tasks but struggles to leverage information effectively from multiple sensory modalities that differ in many characteristics. Recent works construct auxiliary losses based on reconstruction or mutual information to extract joint representations from multiple sensory inputs to improve the sample efficiency and performance of reinforcement learning algorithms. However, the representations learned by these methods could capture information irrelevant to learning a policy and may degrade the performance. We argue that compressing information in the learned joint representations about raw multimodal observations is helpful, and propose a multimodal information bottleneck model to learn task-relevant joint representations from egocentric images and proprioception. Our model compresses and retains the predictive information in multimodal observations for learning a compressed joint representation, which fuses complementary information from visual and proprioceptive feedback and meanwhile filters out task-irrelevant information in raw multimodal observations. We propose to minimize the upper bound of our multimodal information bottleneck objective for computationally tractable optimization. Experimental evaluations on several challenging locomotion tasks with egocentric images and proprioception show that our method achieves better sample efficiency and zero-shot robustness to unseen white noise than leading baselines. We also empirically demonstrate that leveraging information from egocentric images and proprioception is more helpful for learning policies on locomotion tasks than solely using one single modality.
comment: 31 pages
☆ Predicting 30-Day Hospital Readmission in Medicare Patients: Insights from an LSTM Deep Learning Model
Readmissions among Medicare beneficiaries are a major problem for the US healthcare system from a perspective of both healthcare operations and patient caregiving outcomes. Our study analyzes Medicare hospital readmissions using LSTM networks with feature engineering to assess feature contributions. We selected variables from admission-level data, inpatient medical history and patient demography. The LSTM model is designed to capture temporal dynamics from admission-level and patient-level data. On a case study on the MIMIC dataset, the LSTM model outperformed the logistic regression baseline, accurately leveraging temporal features to predict readmission. The major features were the Charlson Comorbidity Index, hospital length of stay, the hospital admissions over the past 6 months, while demographic variables were less impactful. This work suggests that LSTM networks offers a more promising approach to improve Medicare patient readmission prediction. It captures temporal interactions in patient databases, enhancing current prediction models for healthcare providers. Adoption of predictive models into clinical practice may be more effective in identifying Medicare patients to provide early and targeted interventions to improve patient outcomes.
comment: 5 pages, 1 table, 5 figures, Accepted by 2024 3rd International Conference on Cloud Computing, Big Data Application and Software Engineering(CBASE 2024), the final version will be published on on IEEE Conference proceeding
☆ Primal-Dual Spectral Representation for Off-policy Evaluation
Off-policy evaluation (OPE) is one of the most fundamental problems in reinforcement learning (RL) to estimate the expected long-term payoff of a given target policy with only experiences from another behavior policy that is potentially unknown. The distribution correction estimation (DICE) family of estimators have advanced the state of the art in OPE by breaking the curse of horizon. However, the major bottleneck of applying DICE estimators lies in the difficulty of solving the saddle-point optimization involved, especially with neural network implementations. In this paper, we tackle this challenge by establishing a linear representation of value function and stationary distribution correction ratio, i.e., primal and dual variables in the DICE framework, using the spectral decomposition of the transition operator. Such primal-dual representation not only bypasses the non-convex non-concave optimization in vanilla DICE, therefore enabling an computational efficient algorithm, but also paves the way for more efficient utilization of historical data. We highlight that our algorithm, SpectralDICE, is the first to leverage the linear representation of primal-dual variables that is both computation and sample efficient, the performance of which is supported by a rigorous theoretical sample complexity guarantee and a thorough empirical evaluation on various benchmarks.
comment: 29 pages, 5 figures
☆ GDDA: Semantic OOD Detection on Graphs under Covariate Shift via Score-Based Diffusion Models
Out-of-distribution (OOD) detection poses a significant challenge for Graph Neural Networks (GNNs), particularly in open-world scenarios with varying distribution shifts. Most existing OOD detection methods on graphs primarily focus on identifying instances in test data domains caused by either semantic shifts (changes in data classes) or covariate shifts (changes in data features), while leaving the simultaneous occurrence of both distribution shifts under-explored. In this work, we address both types of shifts simultaneously and introduce a novel challenge for OOD detection on graphs: graph-level semantic OOD detection under covariate shift. In this scenario, variations between the training and test domains result from the concurrent presence of both covariate and semantic shifts, where only graphs associated with unknown classes are identified as OOD samples (OODs). To tackle this challenge, we propose a novel two-phase framework called Graph Disentangled Diffusion Augmentation (GDDA). The first phase focuses on disentangling graph representations into domain-invariant semantic factors and domain-specific style factors. In the second phase, we introduce a novel distribution-shift-controlled score-based generative diffusion model that generates latent factors outside the training semantic and style spaces. Additionally, auxiliary pseudo-in-distribution (InD) and pseudo-OOD graph representations are employed to enhance the effectiveness of the energy-based semantic OOD detector. Extensive empirical studies on three benchmark datasets demonstrate that our approach outperforms state-of-the-art baselines.
comment: 4 pages, 6 figures
☆ MobileSafetyBench: Evaluating Safety of Autonomous Agents in Mobile Device Control
Autonomous agents powered by large language models (LLMs) show promising potential in assistive tasks across various domains, including mobile device control. As these agents interact directly with personal information and device settings, ensuring their safe and reliable behavior is crucial to prevent undesirable outcomes. However, no benchmark exists for standardized evaluation of the safety of mobile device-control agents. In this work, we introduce MobileSafetyBench, a benchmark designed to evaluate the safety of device-control agents within a realistic mobile environment based on Android emulators. We develop a diverse set of tasks involving interactions with various mobile applications, including messaging and banking applications. To clearly evaluate safety apart from general capabilities, we design separate tasks measuring safety and tasks evaluating helpfulness. The safety tasks challenge agents with managing potential risks prevalent in daily life and include tests to evaluate robustness against indirect prompt injections. Our experiments demonstrate that while baseline agents, based on state-of-the-art LLMs, perform well in executing helpful tasks, they show poor performance in safety tasks. To mitigate these safety concerns, we propose a prompting method that encourages agents to prioritize safety considerations. While this method shows promise in promoting safer behaviors, there is still considerable room for improvement to fully earn user trust. This highlights the urgent need for continued research to develop more robust safety mechanisms in mobile environments. We open-source our benchmark at: https://mobilesafetybench.github.io/.
☆ Univariate Conditional Variational Autoencoder for Morphogenic Patterns Design in Frontal Polymerization-Based Manufacturing
Rapid reaction-thermal diffusion during frontal polymerization (FP) with variations in initial and boundary conditions destabilizes the planar mode of front propagation, leading to spatially varying complex hierarchical patterns in polymeric materials. Although modern reaction-diffusion models can predict the patterns resulting from unstable FP, the inverse design of patterns, which aims to retrieve process conditions that produce a desired pattern, remains an open challenge due to the nonunique and nonintuitive mapping between process conditions and patterns. In this work, we propose a novel probabilistic generative model named univariate conditional variational autoencoder (UcVAE) for the inverse design of hierarchical patterns in FP-based manufacturing. Unlike the cVAE, which encodes both the design space and the design target, the UcVAE encodes only the design space. In the encoder of the UcVAE, the number of training parameters is significantly reduced compared to the cVAE, resulting in a shorter training time while maintaining comparable performance. Given desired pattern images, the trained UcVAE can generate multiple process condition solutions that produce high-fidelity hierarchical patterns.
☆ Time and Frequency Synergy for Source-Free Time-Series Domain Adaptations
The issue of source-free time-series domain adaptations still gains scarce research attentions. On the other hand, existing approaches rely solely on time-domain features ignoring frequency components providing complementary information. This paper proposes Time Frequency Domain Adaptation (TFDA), a method to cope with the source-free time-series domain adaptation problems. TFDA is developed with a dual branch network structure fully utilizing both time and frequency features in delivering final predictions. It induces pseudo-labels based on a neighborhood concept where predictions of a sample group are aggregated to generate reliable pseudo labels. The concept of contrastive learning is carried out in both time and frequency domains with pseudo label information and a negative pair exclusion strategy to make valid neighborhood assumptions. In addition, the time-frequency consistency technique is proposed using the self-distillation strategy while the uncertainty reduction strategy is implemented to alleviate uncertainties due to the domain shift problem. Last but not least, the curriculum learning strategy is integrated to combat noisy pseudo labels. Our experiments demonstrate the advantage of our approach over prior arts with noticeable margins in benchmark problems.
☆ Congestion Forecast for Trains with Railroad-Graph-based Semi-Supervised Learning using Sparse Passenger Reports SP
Forecasting rail congestion is crucial for efficient mobility in transport systems. We present rail congestion forecasting using reports from passengers collected through a transit application. Although reports from passengers have received attention from researchers, ensuring a sufficient volume of reports is challenging due to passenger's reluctance. The limited number of reports results in the sparsity of the congestion label, which can be an issue in building a stable prediction model. To address this issue, we propose a semi-supervised method for congestion forecasting for trains, or SURCONFORT. Our key idea is twofold: firstly, we adopt semi-supervised learning to leverage sparsely labeled data and many unlabeled data. Secondly, in order to complement the unlabeled data from nearby stations, we design a railway network-oriented graph and apply the graph to semi-supervised graph regularization. Empirical experiments with actual reporting data show that SURCONFORT improved the forecasting performance by 14.9% over state-of-the-art methods under the label sparsity.
comment: Accepted in ACM SIGSPATIAL 2024
☆ WAGLE: Strategic Weight Attribution for Effective and Modular Unlearning in Large Language Models
The need for effective unlearning mechanisms in large language models (LLMs) is increasingly urgent, driven by the necessity to adhere to data regulations and foster ethical generative AI practices. Despite growing interest of LLM unlearning, much of the existing research has focused on varied unlearning method designs to boost effectiveness and efficiency. However, the inherent relationship between model weights and LLM unlearning has not been extensively examined. In this paper, we systematically explore how model weights interact with unlearning processes in LLMs and we design the weight attribution-guided LLM unlearning method, WAGLE, which unveils the interconnections between 'influence' of weights and 'influence' of data to forget and retain in LLM generation. By strategically guiding the LLM unlearning across different types of unlearning methods and tasks, WAGLE can erase the undesired content, while maintaining the performance of the original tasks. We refer to the weight attribution-guided LLM unlearning method as WAGLE, which unveils the interconnections between 'influence' of weights and 'influence' of data to forget and retain in LLM generation. Our extensive experiments show that WAGLE boosts unlearning performance across a range of LLM unlearning methods such as gradient difference and (negative) preference optimization, applications such as fictitious unlearning, malicious use prevention, and copyrighted information removal, and models including Zephyr-7b-beta and Llama2-7b. To the best of our knowledge, our work offers the first principled method for attributing and pinpointing the influential weights in enhancing LLM unlearning. It stands in contrast to previous methods that lack weight attribution and simpler weight attribution techniques.
☆ Mitigating Graph Covariate Shift via Score-based Out-of-distribution Augmentation
Distribution shifts between training and testing datasets significantly impair the model performance on graph learning. A commonly-taken causal view in graph invariant learning suggests that stable predictive features of graphs are causally associated with labels, whereas varying environmental features lead to distribution shifts. In particular, covariate shifts caused by unseen environments in test graphs underscore the critical need for out-of-distribution (OOD) generalization. Existing graph augmentation methods designed to address the covariate shift often disentangle the stable and environmental features in the input space, and selectively perturb or mixup the environmental features. However, such perturbation-based methods heavily rely on an accurate separation of stable and environmental features, and their exploration ability is confined to existing environmental features in the training distribution. To overcome these limitations, we introduce a novel approach using score-based graph generation strategies that synthesize unseen environmental features while preserving the validity and stable features of overall graph patterns. Our comprehensive empirical evaluations demonstrate the enhanced effectiveness of our method in improving graph OOD generalization.
comment: 17 pages, 5 figures, 4 tables
☆ BadFair: Backdoored Fairness Attacks with Group-conditioned Triggers EMNLP 2024
Attacking fairness is crucial because compromised models can introduce biased outcomes, undermining trust and amplifying inequalities in sensitive applications like hiring, healthcare, and law enforcement. This highlights the urgent need to understand how fairness mechanisms can be exploited and to develop defenses that ensure both fairness and robustness. We introduce BadFair, a novel backdoored fairness attack methodology. BadFair stealthily crafts a model that operates with accuracy and fairness under regular conditions but, when activated by certain triggers, discriminates and produces incorrect results for specific groups. This type of attack is particularly stealthy and dangerous, as it circumvents existing fairness detection methods, maintaining an appearance of fairness in normal use. Our findings reveal that BadFair achieves a more than 85% attack success rate in attacks aimed at target groups on average while only incurring a minimal accuracy loss. Moreover, it consistently exhibits a significant discrimination score, distinguishing between pre-defined target and non-target attacked groups across various datasets and models.
comment: Accepted by EMNLP 2024
☆ GenDP: 3D Semantic Fields for Category-Level Generalizable Diffusion Policy
Diffusion-based policies have shown remarkable capability in executing complex robotic manipulation tasks but lack explicit characterization of geometry and semantics, which often limits their ability to generalize to unseen objects and layouts. To enhance the generalization capabilities of Diffusion Policy, we introduce a novel framework that incorporates explicit spatial and semantic information via 3D semantic fields. We generate 3D descriptor fields from multi-view RGBD observations with large foundational vision models, then compare these descriptor fields against reference descriptors to obtain semantic fields. The proposed method explicitly considers geometry and semantics, enabling strong generalization capabilities in tasks requiring category-level generalization, resolving geometric ambiguities, and attention to subtle geometric details. We evaluate our method across eight tasks involving articulated objects and instances with varying shapes and textures from multiple object categories. Our method demonstrates its effectiveness by increasing Diffusion Policy's average success rate on unseen instances from 20% to 93%. Additionally, we provide a detailed analysis and visualization to interpret the sources of performance gain and explain how our method can generalize to novel instances.
comment: Accepted to Conference on Robot Learning (CoRL 2024). Project Page: https://robopil.github.io/GenDP/
☆ Which Client is Reliable?: A Reliable and Personalized Prompt-based Federated Learning for Medical Image Question Answering
Conventional medical artificial intelligence (AI) models face barriers in clinical application and ethical issues owing to their inability to handle the privacy-sensitive characteristics of medical data. We present a novel personalized federated learning (pFL) method for medical visual question answering (VQA) models, addressing privacy reliability challenges in the medical domain. Our method introduces learnable prompts into a Transformer architecture to efficiently train it on diverse medical datasets without massive computational costs. Then we introduce a reliable client VQA model that incorporates Dempster-Shafer evidence theory to quantify uncertainty in predictions, enhancing the model's reliability. Furthermore, we propose a novel inter-client communication mechanism that uses maximum likelihood estimation to balance accuracy and uncertainty, fostering efficient integration of insights across clients.
♻ ☆ Pruning By Explaining Revisited: Optimizing Attribution Methods to Prune CNNs and Transformers ECCV 2024
To solve ever more complex problems, Deep Neural Networks are scaled to billions of parameters, leading to huge computational costs. An effective approach to reduce computational requirements and increase efficiency is to prune unnecessary components of these often over-parameterized networks. Previous work has shown that attribution methods from the field of eXplainable AI serve as effective means to extract and prune the least relevant network components in a few-shot fashion. We extend the current state by proposing to explicitly optimize hyperparameters of attribution methods for the task of pruning, and further include transformer-based networks in our analysis. Our approach yields higher model compression rates of large transformer- and convolutional architectures (VGG, ResNet, ViT) compared to previous works, while still attaining high performance on ImageNet classification tasks. Here, our experiments indicate that transformers have a higher degree of over-parameterization compared to convolutional neural networks. Code is available at https://github.com/erfanhatefi/Pruning-by-eXplaining-in-PyTorch.
comment: Accepted as a workshop paper at ECCV 2024, 26 pages (11 pages manuscript, 3 pages references, 12 pages appendix)
♻ ☆ Correlated Proxies: A New Definition and Improved Mitigation for Reward Hacking
Because it is difficult to precisely specify complex objectives, reinforcement learning policies are often optimized using flawed proxy rewards that seem to capture the true objective. However, optimizing proxy rewards frequently leads to reward hacking: the optimized reward function ceases to be a good proxy, and the resulting policy performs poorly with respect to the unspecified true reward. Principled solutions to reward hacking have been impeded by the lack of a good definition for the problem. To address this, we introduce a definition of reward hacking based on the correlation between proxy and true rewards for states and actions seen by a "base policy" that breaks down under optimization. We show that this definition captures reward hacking behavior across several realistic settings, including in reinforcement learning from human feedback (RLHF). We then show theoretically that regularization to the base policy can effectively prevent reward hacking. While current RLHF approaches apply a KL penalty between the action distributions of policies, our theory suggests that it is more effective to regularize using the $\chi^2$ divergence between the policies' occupancy measures. We intuitively show why this type of regularization is superior and demonstrate that it better mitigates reward hacking in practice across four realistic domains, including RLHF for LLMs. Our code is available at https://github.com/cassidylaidlaw/orpo.
♻ ☆ Conditional Language Policy: A General Framework for Steerable Multi-Objective Finetuning EMNLP 2024
Reward-based finetuning is crucial for aligning language policies with intended behaviors (e.g., creativity and safety). A key challenge is to develop steerable language models that trade-off multiple (conflicting) objectives in a flexible and efficient manner. This paper presents Conditional Language Policy (CLP), a general framework for finetuning language models on multiple objectives. Building on techniques from multi-task training and parameter-efficient finetuning, CLP learn steerable models that effectively trade-off conflicting objectives at inference time. Notably, this does not require training or maintaining multiple models to achieve different trade-offs between the objectives. Through extensive experiments and ablations on two summarization datasets, we show that CLP learns steerable language models that outperform and Pareto-dominate the existing approaches for multi-objective finetuning.
comment: 40 pages. Findings of EMNLP 2024
♻ ☆ VILA-U: a Unified Foundation Model Integrating Visual Understanding and Generation
VILA-U is a Unified foundation model that integrates Video, Image, Language understanding and generation. Traditional visual language models (VLMs) use separate modules for understanding and generating visual content, which can lead to misalignment and increased complexity. In contrast, VILA-U employs a single autoregressive next-token prediction framework for both tasks, eliminating the need for additional components like diffusion models. This approach not only simplifies the model but also achieves near state-of-the-art performance in visual language understanding and generation. The success of VILA-U is attributed to two main factors: the unified vision tower that aligns discrete visual tokens with textual inputs during pretraining, which enhances visual perception, and autoregressive image generation can achieve similar quality as diffusion models with high-quality dataset. This allows VILA-U to perform comparably to more complex models using a fully token-based autoregressive framework.
comment: Code: https://github.com/mit-han-lab/vila-u. The first two authors contributed equally to this work
♻ ☆ Counter-Current Learning: A Biologically Plausible Dual Network Approach for Deep Learning NeurIPS 2024
Despite its widespread use in neural networks, error backpropagation has faced criticism for its lack of biological plausibility, suffering from issues such as the backward locking problem and the weight transport problem. These limitations have motivated researchers to explore more biologically plausible learning algorithms that could potentially shed light on how biological neural systems adapt and learn. Inspired by the counter-current exchange mechanisms observed in biological systems, we propose counter-current learning (CCL), a biologically plausible framework for credit assignment in neural networks. This framework employs a feedforward network to process input data and a feedback network to process targets, with each network enhancing the other through anti-parallel signal propagation. By leveraging the more informative signals from the bottom layer of the feedback network to guide the updates of the top layer of the feedforward network and vice versa, CCL enables the simultaneous transformation of source inputs to target outputs and the dynamic mutual influence of these transformations. Experimental results on MNIST, FashionMNIST, CIFAR10, and CIFAR100 datasets using multi-layer perceptrons and convolutional neural networks demonstrate that CCL achieves comparable performance to other biologically plausible algorithms while offering a more biologically realistic learning mechanism. Furthermore, we showcase the applicability of our approach to an autoencoder task, underscoring its potential for unsupervised representation learning. Our work presents a direction for biologically inspired and plausible learning algorithms, offering an alternative mechanism of learning and adaptation in neural networks.
comment: Accepted at NeurIPS 2024. Code available at https://github.com/IandRover/CCL-NeurIPS24
♻ ☆ Proof of Thought : Neurosymbolic Program Synthesis allows Robust and Interpretable Reasoning NeurIPS 2024
Large Language Models (LLMs) have revolutionized natural language processing, yet they struggle with inconsistent reasoning, particularly in novel domains and complex logical sequences. This research introduces Proof of Thought, a framework that enhances the reliability and transparency of LLM outputs. Our approach bridges LLM-generated ideas with formal logic verification, employing a custom interpreter to convert LLM outputs into First Order Logic constructs for theorem prover scrutiny. Central to our method is an intermediary JSON-based Domain-Specific Language, which by design balances precise logical structures with intuitive human concepts. This hybrid representation enables both rigorous validation and accessible human comprehension of LLM reasoning processes. Key contributions include a robust type system with sort management for enhanced logical integrity, explicit representation of rules for clear distinction between factual and inferential knowledge, and a flexible architecture that allows for easy extension to various domain-specific applications. We demonstrate Proof of Thought's effectiveness through benchmarking on StrategyQA and a novel multimodal reasoning task, showing improved performance in open-ended scenarios. By providing verifiable and interpretable results, our technique addresses critical needs for AI system accountability and sets a foundation for human-in-the-loop oversight in high-stakes domains.
comment: 38th Conference on Neural Information Processing Systems (NeurIPS 2024) System 2 Reasoning At Scale Workshop
♻ ☆ AlleNoise: large-scale text classification benchmark dataset with real-world label noise
Label noise remains a challenge for training robust classification models. Most methods for mitigating label noise have been benchmarked using primarily datasets with synthetic noise. While the need for datasets with realistic noise distribution has partially been addressed by web-scraped benchmarks such as WebVision and Clothing1M, those benchmarks are restricted to the computer vision domain. With the growing importance of Transformer-based models, it is crucial to establish text classification benchmarks for learning with noisy labels. In this paper, we present AlleNoise, a new curated text classification benchmark dataset with real-world instance-dependent label noise, containing over 500,000 examples across approximately 5,600 classes, complemented with a meaningful, hierarchical taxonomy of categories. The noise distribution comes from actual users of a major e-commerce marketplace, so it realistically reflects the semantics of human mistakes. In addition to the noisy labels, we provide human-verified clean labels, which help to get a deeper insight into the noise distribution, unlike web-scraped datasets typically used in the field. We demonstrate that a representative selection of established methods for learning with noisy labels is inadequate to handle such real-world noise. In addition, we show evidence that these algorithms do not alleviate excessive memorization. As such, with AlleNoise, we set the bar high for the development of label noise methods that can handle real-world label noise in text classification tasks. The code and dataset are available for download at https://github.com/allegro/AlleNoise.
♻ ☆ CondTSF: One-line Plugin of Dataset Condensation for Time Series Forecasting NeurIPS 2024
Dataset condensation is a newborn technique that generates a small dataset that can be used in training deep neural networks to lower training costs. The objective of dataset condensation is to ensure that the model trained with the synthetic dataset can perform comparably to the model trained with full datasets. However, existing methods predominantly concentrate on classification tasks, posing challenges in their adaptation to time series forecasting (TS-forecasting). This challenge arises from disparities in the evaluation of synthetic data. In classification, the synthetic data is considered well-distilled if the model trained with the full dataset and the model trained with the synthetic dataset yield identical labels for the same input, regardless of variations in output logits distribution. Conversely, in TS-forecasting, the effectiveness of synthetic data distillation is determined by the distance between predictions of the two models. The synthetic data is deemed well-distilled only when all data points within the predictions are similar. Consequently, TS-forecasting has a more rigorous evaluation methodology compared to classification. To mitigate this gap, we theoretically analyze the optimization objective of dataset condensation for TS-forecasting and propose a new one-line plugin of dataset condensation designated as Dataset Condensation for Time Series Forecasting (CondTSF) based on our analysis. Plugging CondTSF into previous dataset condensation methods facilitates a reduction in the distance between the predictions of the model trained with the full dataset and the model trained with the synthetic dataset, thereby enhancing performance. We conduct extensive experiments on eight commonly used time series datasets. CondTSF consistently improves the performance of all previous dataset condensation methods across all datasets, particularly at low condensing ratios.
comment: Accepted by NeurIPS 2024, the project can be found at https://github.com/RafaDD/CondTSF
♻ ☆ Learning a quantum computer's capability
Accurately predicting a quantum computer's capability -- which circuits it can run and how well it can run them -- is a foundational goal of quantum characterization and benchmarking. As modern quantum computers become increasingly hard to simulate, we must develop accurate and scalable predictive capability models to help researchers and stakeholders decide which quantum computers to build and use. In this work, we propose a hardware-agnostic method to efficiently construct scalable predictive models of a quantum computer's capability for almost any class of circuits, and demonstrate our method using convolutional neural networks (CNNs). Our CNN-based approach works by efficiently representing a circuit as a three-dimensional tensor and then using a CNN to predict its success rate. Our CNN capability models obtain approximately a $1\%$ average absolute prediction error when modeling processors experiencing both Markovian and non-Markovian stochastic Pauli errors. We also apply our CNNs to model the capabilities of cloud-access quantum computing systems, obtaining moderate prediction accuracy (average absolute error around $2-5\%$), and we highlight the challenges to building better neural network capability models.
comment: 20 pages, 11 figures, plus appendices
♻ ☆ Federated Class-Incremental Learning with Hierarchical Generative Prototypes
Federated Learning (FL) aims at unburdening the training of deep models by distributing computation across multiple devices (clients) while safeguarding data privacy. On top of that, Federated Continual Learning (FCL) also accounts for data distribution evolving over time, mirroring the dynamic nature of real-world environments. While previous studies have identified Catastrophic Forgetting and Client Drift as primary causes of performance degradation in FCL, we shed light on the importance of Incremental Bias and Federated Bias, which cause models to prioritize classes that are recently introduced or locally predominant, respectively. Our proposal constrains both biases in the last layer by efficiently finetuning a pre-trained backbone using learnable prompts, resulting in clients that produce less biased representations and more biased classifiers. Therefore, instead of solely relying on parameter aggregation, we leverage generative prototypes to effectively balance the predictions of the global model. Our method significantly improves the current State Of The Art, providing an average increase of +7.8% in accuracy. Code to reproduce the results is provided in the suppl. material.
♻ ☆ Do Large Language Models Truly Grasp Mathematics? An Empirical Exploration
Despite their proficiency in math tasks, the mechanisms underlying LLMs' mathematical reasoning abilities remain a subject of debate. Recent studies suggest that chain-of-thought (CoT) prompts can bolster mathematical reasoning by encouraging LLMs to employ human-like logical reasoning (System 2), enabling them to excel on the Cognitive Reflection Test (CRT). To assess whether LLMs genuinely possess System 2-like logical reasoning, we introduced targeted modifications to CRT problems. Our findings reveal that, despite the use of CoT prompts, mainstream LLMs, including the latest o1-preview model, continue to exhibit a significant error rate. Further analysis indicates that they predominantly rely on System 1-like intuitive reasoning and pattern matching derived from training data, rather than demonstrating mastery of mathematical thinking. This discovery challenges the prevailing notion that LLMs possess genuine logical reasoning abilities and that CoT can enhance them. Consequently, this work may temper overly optimistic projections regarding LLMs' advancement toward artificial general intelligence.
♻ ☆ On the potential of Optimal Transport in Geospatial Data Science
Prediction problems in geographic information science and transportation are often motivated by the possibility to enhance operational efficiency and thereby reduce emissions. Examples range from predicting car sharing demand for relocation planning to forecasting traffic congestion for navigation purposes. However, conventional accuracy metrics ignore the spatial distribution of the errors, despite its relevance for operations. Here, we put forward a spatially aware evaluation metric and loss function based on Optimal Transport (OT). Our framework leverages partial OT and can minimize relocation costs in any spatial prediction problem. We showcase the advantages of OT-based evaluation over conventional metrics and further demonstrate the application of an OT loss function for improving forecasts of bike sharing demand and charging station occupancy. Thus, our framework not only aligns with operational considerations, but also signifies a step forward in refining predictions within geospatial applications. All code is available at https://github.com/mie-lab/geospatialOT.
♻ ☆ Quantum Architecture Search with Unsupervised Representation Learning
Unsupervised representation learning presents new opportunities for advancing Quantum Architecture Search (QAS) on Noisy Intermediate-Scale Quantum (NISQ) devices. QAS is designed to optimize quantum circuits for Variational Quantum Algorithms (VQAs). Most QAS algorithms tightly couple the search space and search algorithm, typically requiring the evaluation of numerous quantum circuits, resulting in high computational costs and limiting scalability to larger quantum circuits. Predictor-based QAS algorithms mitigate this issue by estimating circuit performance based on structure or embedding. However, these methods often demand time-intensive labeling to optimize gate parameters across many circuits, which is crucial for training accurate predictors. Inspired by the classical neural architecture search algorithm Arch2vec, we investigate the potential of unsupervised representation learning for QAS without relying on predictors. Our framework decouples unsupervised architecture representation learning from the search process, enabling the learned representations to be applied across various downstream tasks. Additionally, it integrates an improved quantum circuit graph encoding scheme, addressing the limitations of existing representations and enhancing search efficiency. This predictor-free approach removes the need for large labeled datasets. During the search, we employ REINFORCE and Bayesian Optimization to explore the latent representation space and compare their performance against baseline methods. Our results demonstrate that the framework efficiently identifies high-performing quantum circuits with fewer search iterations.
comment: 9 Pages, quantum architecture search, unsupervised representation learning
♻ ☆ MOTIVE: A Drug-Target Interaction Graph For Inductive Link Prediction
Drug-target interaction (DTI) prediction is crucial for identifying new therapeutics and detecting mechanisms of action. While structure-based methods accurately model physical interactions between a drug and its protein target, cell-based assays such as Cell Painting can better capture complex DTI interactions. This paper introduces MOTIVE, a Morphological cOmpound Target Interaction Graph dataset comprising Cell Painting features for 11,000 genes and 3,600 compounds, along with their relationships extracted from seven publicly available databases. We provide random, cold-source (new drugs), and cold-target (new genes) data splits to enable rigorous evaluation under realistic use cases. Our benchmark results show that graph neural networks that use Cell Painting features consistently outperform those that learn from graph structure alone, feature-based models, and topological heuristics. MOTIVE accelerates both graph ML research and drug discovery by promoting the development of more reliable DTI prediction models. MOTIVE resources are available at https://github.com/carpenter-singh-lab/motive.
♻ ☆ Bounded KRnet and its applications to density estimation and approximation
In this paper, we develop an invertible mapping, called B-KRnet, on a bounded domain and apply it to density estimation/approximation for data or the solutions of PDEs such as the Fokker-Planck equation and the Keller-Segel equation. Similar to KRnet, the structure of B-KRnet adapts the pseudo-triangular structure into a normalizing flow model. The main difference between B-KRnet and KRnet is that B-KRnet is defined on a hypercube while KRnet is defined on the whole space, in other words, a new mechanism is introduced in B-KRnet to maintain the exact invertibility. Using B-KRnet as a transport map, we obtain an explicit probability density function (PDF) model that corresponds to the pushforward of a prior (uniform) distribution on the hypercube. It can be directly applied to density estimation when only data are available. By coupling KRnet and B-KRnet, we define a deep generative model on a high-dimensional domain where some dimensions are bounded and other dimensions are unbounded. A typical case is the solution of the stationary kinetic Fokker-Planck equation, which is a PDF of position and momentum. Based on B-KRnet, we develop an adaptive learning approach to approximate partial differential equations whose solutions are PDFs or can be treated as PDFs. A variety of numerical experiments is presented to demonstrate the effectiveness of B-KRnet.
comment: 26 pages, 13 figures
♻ ☆ Linear Adversarial Concept Erasure ICML 2022
Modern neural models trained on textual data rely on pre-trained representations that emerge without direct supervision. As these representations are increasingly being used in real-world applications, the inability to \emph{control} their content becomes an increasingly important problem. We formulate the problem of identifying and erasing a linear subspace that corresponds to a given concept, in order to prevent linear predictors from recovering the concept. We model this problem as a constrained, linear maximin game, and show that existing solutions are generally not optimal for this task. We derive a closed-form solution for certain objectives, and propose a convex relaxation, \method, that works well for others. When evaluated in the context of binary gender removal, the method recovers a low-dimensional subspace whose removal mitigates bias by intrinsic and extrinsic evaluation. We show that the method is highly expressive, effectively mitigating bias in deep nonlinear classifiers while maintaining tractability and interpretability.
comment: Accepted in ICML 2022; a revised version
♻ ☆ Accessible, At-Home Detection of Parkinson's Disease via Multi-task Video Analysis
Limited accessibility to neurological care leads to underdiagnosed Parkinson's Disease (PD), preventing early intervention. Existing AI-based PD detection methods primarily focus on unimodal analysis of motor or speech tasks, overlooking the multifaceted nature of the disease. To address this, we introduce a large-scale, multi-task video dataset consisting of 1102 sessions (each containing videos of finger tapping, facial expression, and speech tasks captured via webcam) from 845 participants (272 with PD). We propose a novel Uncertainty-calibrated Fusion Network (UFNet) that leverages this multimodal data to enhance diagnostic accuracy. UFNet employs independent task-specific networks, trained with Monte Carlo Dropout for uncertainty quantification, followed by self-attended fusion of features, with attention weights dynamically adjusted based on task-specific uncertainties. To ensure patient-centered evaluation, the participants were randomly split into three sets: 60% for training, 20% for model selection, and 20% for final performance evaluation. UFNet significantly outperformed single-task models in terms of accuracy, area under the ROC curve (AUROC), and sensitivity while maintaining non-inferior specificity. Withholding uncertain predictions further boosted the performance, achieving 88.0+-0.3%$ accuracy, 93.0+-0.2% AUROC, 79.3+-0.9% sensitivity, and 92.6+-0.3% specificity, at the expense of not being able to predict for 2.3+-0.3% data (+- denotes 95% confidence interval). Further analysis suggests that the trained model does not exhibit any detectable bias across sex and ethnic subgroups and is most effective for individuals aged between 50 and 80. Requiring only a webcam and microphone, our approach facilitates accessible home-based PD screening, especially in regions with limited healthcare resources.
♻ ☆ Certifiably Robust Policies for Uncertain Parametric Environments
We present a data-driven approach for producing policies that are provably robust across unknown stochastic environments. Existing approaches can learn models of a single environment as an interval Markov decision processes (IMDP) and produce a robust policy with a probably approximately correct (PAC) guarantee on its performance. However these are unable to reason about the impact of environmental parameters underlying the uncertainty. We propose a framework based on parametric Markov decision processes (MDPs) with unknown distributions over parameters. We learn and analyse IMDPs for a set of unknown sample environments induced by parameters. The key challenge is then to produce meaningful performance guarantees that combine the two layers of uncertainty: (1) multiple environments induced by parameters with an unknown distribution; (2) unknown induced environments which are approximated by IMDPs. We present a novel approach based on scenario optimisation that yields a single PAC guarantee quantifying the risk level for which a specified performance level can be assured in unseen environments, plus a means to trade-off risk and performance. We implement and evaluate our framework using multiple robust policy generation methods on a range of benchmarks. We show that our approach produces tight bounds on a policy's performance with high confidence.
♻ ☆ On provable privacy vulnerabilities of graph representations
Graph representation learning (GRL) is critical for extracting insights from complex network structures, but it also raises security concerns due to potential privacy vulnerabilities in these representations. This paper investigates the structural vulnerabilities in graph neural models where sensitive topological information can be inferred through edge reconstruction attacks. Our research primarily addresses the theoretical underpinnings of similarity-based edge reconstruction attacks (SERA), furnishing a non-asymptotic analysis of their reconstruction capacities. Moreover, we present empirical corroboration indicating that such attacks can perfectly reconstruct sparse graphs as graph size increases. Conversely, we establish that sparsity is a critical factor for SERA's effectiveness, as demonstrated through analysis and experiments on (dense) stochastic block models. Finally, we explore the resilience of private graph representations produced via noisy aggregation (NAG) mechanism against SERA. Through theoretical analysis and empirical assessments, we affirm the mitigation of SERA using NAG . In parallel, we also empirically delineate instances wherein SERA demonstrates both efficacy and deficiency in its capacity to function as an instrument for elucidating the trade-off between privacy and utility.
♻ ☆ Adaptive Variance Reduction for Stochastic Optimization under Weaker Assumptions
This paper explores adaptive variance reduction methods for stochastic optimization based on the STORM technique. Existing adaptive extensions of STORM rely on strong assumptions like bounded gradients and bounded function values, or suffer an additional $\mathcal{O}(\log T)$ term in the convergence rate. To address these limitations, we introduce a novel adaptive STORM method that achieves an optimal convergence rate of $\mathcal{O}(T^{-1/3})$ for non-convex functions with our newly designed learning rate strategy. Compared with existing approaches, our method requires weaker assumptions and attains the optimal convergence rate without the additional $\mathcal{O}(\log T)$ term. We also extend the proposed technique to stochastic compositional optimization, obtaining the same optimal rate of $\mathcal{O}(T^{-1/3})$. Furthermore, we investigate the non-convex finite-sum problem and develop another innovative adaptive variance reduction method that achieves an optimal convergence rate of $\mathcal{O}(n^{1/4} T^{-1/2} )$, where $n$ represents the number of component functions. Numerical experiments across various tasks validate the effectiveness of our method.
comment: arXiv admin note: substantial text overlap with arXiv:2406.00489
♻ ☆ Posterior Sampling-based Online Learning for Episodic POMDPs
Learning in POMDPs is known to be significantly harder than in MDPs. In this paper, we consider the online learning problem for episodic POMDPs with unknown transition and observation models. We propose a Posterior Sampling-based reinforcement learning algorithm for POMDPs (PS4POMDPs), which is much simpler and more implementable compared to state-of-the-art optimism-based online learning algorithms for POMDPs. We show that the Bayesian regret of the proposed algorithm scales as the square root of the number of episodes and is polynomial in the other parameters. In a general setting, the regret scales exponentially in the horizon length $H$, and we show that this is inevitable by providing a lower bound. However, when the POMDP is undercomplete and weakly revealing (a common assumption in the recent literature), we establish a polynomial Bayesian regret bound. We finally propose a posterior sampling algorithm for multi-agent POMDPs, and show it too has sublinear regret.
comment: 41 pages, 9 figures
♻ ☆ Efficient Sign-Based Optimization: Accelerating Convergence via Variance Reduction
Sign stochastic gradient descent (signSGD) is a communication-efficient method that transmits only the sign of stochastic gradients for parameter updating. Existing literature has demonstrated that signSGD can achieve a convergence rate of $\mathcal{O}(d^{1/2}T^{-1/4})$, where $d$ represents the dimension and $T$ is the iteration number. In this paper, we improve this convergence rate to $\mathcal{O}(d^{1/2}T^{-1/3})$ by introducing the Sign-based Stochastic Variance Reduction (SSVR) method, which employs variance reduction estimators to track gradients and leverages their signs to update. For finite-sum problems, our method can be further enhanced to achieve a convergence rate of $\mathcal{O}(m^{1/4}d^{1/2}T^{-1/2})$, where $m$ denotes the number of component functions. Furthermore, we investigate the heterogeneous majority vote in distributed settings and introduce two novel algorithms that attain improved convergence rates of $\mathcal{O}(d^{1/2}T^{-1/2} + dn^{-1/2})$ and $\mathcal{O}(d^{1/4}T^{-1/4})$ respectively, outperforming the previous results of $\mathcal{O}(dT^{-1/4} + dn^{-1/2})$ and $\mathcal{O}(d^{3/8}T^{-1/8})$, where $n$ represents the number of nodes. Numerical experiments across different tasks validate the effectiveness of our proposed methods.
♻ ☆ Anomaly Prediction: A Novel Approach with Explicit Delay and Horizon
Anomaly detection in time series data is a critical challenge across various domains. Traditional methods typically focus on identifying anomalies in immediate subsequent steps, often underestimating the significance of temporal dynamics such as delay time and horizons of anomalies, which generally require extensive post-analysis. This paper introduces a novel approach for time series anomaly prediction, incorporating temporal information directly into the prediction results. We propose a new dataset specifically designed to evaluate this approach and conduct comprehensive experiments using several state-of-the-art methods. Our results demonstrate the efficacy of our approach in providing timely and accurate anomaly predictions, setting a new benchmark for future research in this field.
♻ ☆ Acquiring Better Load Estimates by Combining Anomaly and Change Point Detection in Power Grid Time-series Measurements
In this paper we present novel methodology for automatic anomaly and switch event filtering to improve load estimation in power grid systems. By leveraging unsupervised methods with supervised optimization, our approach prioritizes interpretability while ensuring robust and generalizable performance on unseen data. Through experimentation, a combination of binary segmentation for change point detection and statistical process control for anomaly detection emerges as the most effective strategy, specifically when ensembled in a novel sequential manner. Results indicate the clear wasted potential when filtering is not applied. The automatic load estimation is also fairly accurate, with approximately 90% of estimates falling within a 10% error margin, with only a single significant failure in both the minimum and maximum load estimates across 60 measurements in the test set. Our methodology's interpretability makes it particularly suitable for critical infrastructure planning, thereby enhancing decision-making processes.
comment: All code can be found at: https://github.com/RoelBouman/StormPhase2
♻ ☆ A spring-block theory of feature learning in deep neural networks
Feature-learning deep nets progressively collapse data to a regular low-dimensional geometry. How this phenomenon emerges from collective action of nonlinearity, noise, learning rate, and other choices that shape the dynamics, has eluded first-principles theories built from microscopic neuronal dynamics. We exhibit a noise-nonlinearity phase diagram that identifies regimes where shallow or deep layers learn more effectively. We then propose a macroscopic mechanical theory that reproduces the diagram, explaining why some DNNs are lazy and some active, and linking feature learning across layers to generalization.
♻ ☆ CPE-Pro: A Structure-Sensitive Deep Learning Method for Protein Representation and Origin Evaluation
Protein structures are important for understanding their functions and interactions. Currently, many protein structure prediction methods are enriching the structure database. Discriminating the origin of structures is crucial for distinguishing between experimentally resolved and computationally predicted structures, evaluating the reliability of prediction methods, and guiding downstream biological studies. Building on works in structure prediction, We developed a structure-sensitive supervised deep learning model, Crystal vs Predicted Evaluator for Protein Structure (CPE-Pro), to represent and discriminate the origin of protein structures. CPE-Pro learns the structural information of proteins and captures inter-structural differences to achieve accurate traceability on four data classes, and is expected to be extended to more. Simultaneously, we utilized Foldseek to encode protein structures into "structure-sequences" and trained a protein Structural Sequence Language Model, SSLM. Preliminary experiments demonstrated that, compared to large-scale protein language models pre-trained on vast amounts of amino acid sequences, the "structure-sequence" enables the language model to learn more informative protein features, enhancing and optimizing structural representations. We have provided the code, model weights, and all related materials on https://github.com/GouWenrui/CPE-Pro-main.git.
♻ ☆ Towards Croppable Implicit Neural Representations NeurIPS 2024
Implicit Neural Representations (INRs) have peaked interest in recent years due to their ability to encode natural signals using neural networks. While INRs allow for useful applications such as interpolating new coordinates and signal compression, their black-box nature makes it difficult to modify them post-training. In this paper we explore the idea of editable INRs, and specifically focus on the widely used cropping operation. To this end, we present Local-Global SIRENs -- a novel INR architecture that supports cropping by design. Local-Global SIRENs are based on combining local and global feature extraction for signal encoding. What makes their design unique is the ability to effortlessly remove specific portions of an encoded signal, with a proportional weight decrease. This is achieved by eliminating the corresponding weights from the network, without the need for retraining. We further show how this architecture can be used to support the straightforward extension of previously encoded signals. Beyond signal editing, we examine how the Local-Global approach can accelerate training, enhance encoding of various signals, improve downstream performance, and be applied to modern INRs such as INCODE, highlighting its potential and flexibility. Code is available at https://github.com/maorash/Local-Global-INRs.
comment: Accepted to NeurIPS 2024
♻ ☆ Breaking Class Barriers: Efficient Dataset Distillation via Inter-Class Feature Compensator
Dataset distillation has emerged as a technique aiming to condense informative features from large, natural datasets into a compact and synthetic form. While recent advancements have refined this technique, its performance is bottlenecked by the prevailing class-specific synthesis paradigm. Under this paradigm, synthetic data is optimized exclusively for a pre-assigned one-hot label, creating an implicit class barrier in feature condensation. This leads to inefficient utilization of the distillation budget and oversight of inter-class feature distributions, which ultimately limits the effectiveness and efficiency, as demonstrated in our analysis. To overcome these constraints, this paper presents the Inter-class Feature Compensator (INFER), an innovative distillation approach that transcends the class-specific data-label framework widely utilized in current dataset distillation methods. Specifically, INFER leverages a Universal Feature Compensator (UFC) to enhance feature integration across classes, enabling the generation of multiple additional synthetic instances from a single UFC input. This significantly improves the efficiency of the distillation budget. Moreover, INFER enriches inter-class interactions during the distillation, thereby enhancing the effectiveness and generalizability of the distilled data. By allowing for the linear interpolation of labels similar to those in the original dataset, INFER meticulously optimizes the synthetic data and dramatically reduces the size of soft labels in the synthetic dataset to almost zero, establishing a new benchmark for efficiency and effectiveness in dataset distillation.
♻ ☆ I've Got 99 Problems But FLOPS Ain't One
Hyperscalers dominate the landscape of large network deployments, yet they rarely share data or insights about the challenges they face. In light of this supremacy, what problems can we find to solve in this space? We take an unconventional approach to find relevant research directions, starting from public plans to build a $100 billion datacenter for machine learning applications. Leveraging the language models scaling laws, we discover what workloads such a datacenter might carry and explore the challenges one may encounter in doing so, with a focus on networking research. We conclude that building the datacenter and training such models is technically possible, but this requires novel wide-area transports for inter-DC communication, a multipath transport and novel datacenter topologies for intra-datacenter communication, high speed scale-up networks and transports, outlining a rich research agenda for the networking community.
♻ ☆ TargetCall: Eliminating the Wasted Computation in Basecalling via Pre-Basecalling Filtering
Basecalling is an essential step in nanopore sequencing analysis where the raw signals of nanopore sequencers are converted into nucleotide sequences, i.e., reads. State-of-the-art basecallers employ complex deep learning models to achieve high basecalling accuracy. This makes basecalling computationally inefficient and memory-hungry, bottlenecking the entire genome analysis pipeline. However, for many applications, the majority of reads do no match the reference genome of interest (i.e., target reference) and thus are discarded in later steps in the genomics pipeline, wasting the basecalling computation. To overcome this issue, we propose TargetCall, the first pre-basecalling filter to eliminate the wasted computation in basecalling. TargetCall's key idea is to discard reads that will not match the target reference (i.e., off-target reads) prior to basecalling. TargetCall consists of two main components: (1) LightCall, a lightweight neural network basecaller that produces noisy reads; and (2) Similarity Check, which labels each of these noisy reads as on-target or off-target by matching them to the target reference. Our thorough experimental evaluations show that TargetCall 1) improves the end-to-end basecalling runtime performance of the state-of-the-art basecaller by 3.31x while maintaining high (98.88%) recall in keeping on-target reads, 2) maintains high accuracy in downstream analysis, and 3) achieves better runtime performance, throughput, recall, precision, and generality compared to prior works. TargetCall is available at https://github.com/CMU-SAFARI/TargetCall.
♻ ☆ Interpreting Context Look-ups in Transformers: Investigating Attention-MLP Interactions EMNLP 2024
Understanding the inner workings of large language models (LLMs) is crucial for advancing their theoretical foundations and real-world applications. While the attention mechanism and multi-layer perceptrons (MLPs) have been studied independently, their interactions remain largely unexplored. This study investigates how attention heads and next-token neurons interact in LLMs to predict new words. We propose a methodology to identify next-token neurons, find prompts that highly activate them, and determine the upstream attention heads responsible. We then generate and evaluate explanations for the activity of these attention heads in an automated manner. Our findings reveal that some attention heads recognize specific contexts relevant to predicting a token and activate a downstream token-predicting neuron accordingly. This mechanism provides a deeper understanding of how attention heads work with MLP neurons to perform next-token prediction. Our approach offers a foundation for further research into the intricate workings of LLMs and their impact on text generation and understanding.
comment: Accepted to EMNLP 2024 Main Conference
♻ ☆ Ornstein-Uhlenbeck Adaptation as a Mechanism for Learning in Brains and Machines
Learning is a fundamental property of intelligent systems, observed across biological organisms and engineered systems. While modern intelligent systems typically rely on gradient descent for learning, the need for exact gradients and complex information flow makes its implementation in biological and neuromorphic systems challenging. This has motivated the exploration of alternative learning mechanisms that can operate locally and do not rely on exact gradients. In this work, we introduce a novel approach that leverages noise in the parameters of the system and global reinforcement signals. Using an Ornstein-Uhlenbeck process with adaptive dynamics, our method balances exploration and exploitation during learning, driven by deviations from error predictions, akin to reward prediction error. Operating in continuous time, Orstein-Uhlenbeck adaptation (OUA) is proposed as a general mechanism for learning dynamic, time-evolving environments. We validate our approach across diverse tasks, including supervised learning and reinforcement learning in feedforward and recurrent systems. Additionally, we demonstrate that it can perform meta-learning, adjusting hyper-parameters autonomously. Our results indicate that OUA provides a viable alternative to traditional gradient-based methods, with potential applications in neuromorphic computing. It also hints at a possible mechanism for noise-driven learning in the brain, where stochastic neurotransmitter release may guide synaptic adjustments.
♻ ☆ FOOGD: Federated Collaboration for Both Out-of-distribution Generalization and Detection NeurIPS 2024
Federated learning (FL) is a promising machine learning paradigm that collaborates with client models to capture global knowledge. However, deploying FL models in real-world scenarios remains unreliable due to the coexistence of in-distribution data and unexpected out-of-distribution (OOD) data, such as covariate-shift and semantic-shift data. Current FL researches typically address either covariate-shift data through OOD generalization or semantic-shift data via OOD detection, overlooking the simultaneous occurrence of various OOD shifts. In this work, we propose FOOGD, a method that estimates the probability density of each client and obtains reliable global distribution as guidance for the subsequent FL process. Firstly, SM3D in FOOGD estimates score model for arbitrary distributions without prior constraints, and detects semantic-shift data powerfully. Then SAG in FOOGD provides invariant yet diverse knowledge for both local covariate-shift generalization and client performance generalization. In empirical validations, FOOGD significantly enjoys three main advantages: (1) reliably estimating non-normalized decentralized distributions, (2) detecting semantic shift data via score values, and (3) generalizing to covariate-shift data by regularizing feature extractor. The prejoct is open in https://github.com/XeniaLLL/FOOGD-main.git.
comment: NeurIPS 2024
♻ ☆ Few-Shot Adversarial Prompt Learning on Vision-Language Models NeurIPS 2024
The vulnerability of deep neural networks to imperceptible adversarial perturbations has attracted widespread attention. Inspired by the success of vision-language foundation models, previous efforts achieved zero-shot adversarial robustness by aligning adversarial visual features with text supervision. However, in practice, they are still unsatisfactory due to several issues, including heavy adaptation cost, suboptimal text supervision, and uncontrolled natural generalization capacity. In this paper, to address these issues, we propose a few-shot adversarial prompt framework where adapting input sequences with limited data makes significant adversarial robustness improvement. Specifically, we achieve this by providing adversarially correlated text supervision that is end-to-end learned from adversarial examples. We also propose a novel training objective that enhances the consistency of multi-modal features while encourages differentiated uni-modal features between natural and adversarial examples. The proposed framework gives access to learn adversarial text supervision, which provides superior cross-modal adversarial alignment and matches state-of-the-art zero-shot adversarial robustness with only 1% training data. Code is available at: https://github.com/lionel-w2/FAP.
comment: NeurIPS 2024
♻ ☆ Optimal Design for Reward Modeling in RLHF
Reinforcement Learning from Human Feedback (RLHF) has become a popular approach to align language models (LMs) with human preferences. This method involves collecting a large dataset of human pairwise preferences across various text generations and using it to infer (implicitly or explicitly) a reward model. Numerous methods have been proposed to learn the reward model and align a LM with it. However, the costly process of collecting human preferences has received little attention and could benefit from theoretical insights. This paper addresses this issue and aims to formalize the reward training model in RLHF. We frame the selection of an effective dataset as a simple regret minimization task, using a linear contextual dueling bandit method. Given the potentially large number of arms, this approach is more coherent than the best-arm identification setting. We then propose an offline framework for solving this problem. Under appropriate assumptions - linearity of the reward model in the embedding space, and boundedness of the reward parameter - we derive bounds on the simple regret. Finally, we provide a lower bound that matches our upper bound up to constant and logarithmic terms. To our knowledge, this is the first theoretical contribution in this area to provide an offline approach as well as worst-case guarantees.
♻ ☆ Reducing Variance in Meta-Learning via Laplace Approximation for Regression Tasks
Given a finite set of sample points, meta-learning algorithms aim to learn an optimal adaptation strategy for new, unseen tasks. Often, this data can be ambiguous as it might belong to different tasks concurrently. This is particularly the case in meta-regression tasks. In such cases, the estimated adaptation strategy is subject to high variance due to the limited amount of support data for each task, which often leads to sub-optimal generalization performance. In this work, we address the problem of variance reduction in gradient-based meta-learning and formalize the class of problems prone to this, a condition we refer to as \emph{task overlap}. Specifically, we propose a novel approach that reduces the variance of the gradient estimate by weighing each support point individually by the variance of its posterior over the parameters. To estimate the posterior, we utilize the Laplace approximation, which allows us to express the variance in terms of the curvature of the loss landscape of our meta-learner. Experimental results demonstrate the effectiveness of the proposed method and highlight the importance of variance reduction in meta-learning.
♻ ☆ Conquering the Communication Constraints to Enable Large Pre-Trained Models in Federated Learning
Federated learning (FL) has emerged as a promising paradigm for enabling the collaborative training of models without centralized access to the raw data on local devices. In the typical FL paradigm (e.g., FedAvg), model weights are sent to and from the server each round to participating clients. Recently, the use of small pre-trained models has been shown effective in federated learning optimization and improving convergence. However, recent state-of-the-art pre-trained models are getting more capable but also have more parameters. In conventional FL, sharing the enormous model weights can quickly put a massive communication burden on the system, especially if more capable models are employed. Can we find a solution to enable those strong and readily-available pre-trained models in FL to achieve excellent performance while simultaneously reducing the communication burden? To this end, we investigate the use of parameter-efficient fine-tuning in federated learning and thus introduce a new framework: FedPEFT. Specifically, we systemically evaluate the performance of FedPEFT across a variety of client stability, data distribution, and differential privacy settings. By only locally tuning and globally sharing a small portion of the model weights, significant reductions in the total communication overhead can be achieved while maintaining competitive or even better performance in a wide range of federated learning scenarios, providing insight into a new paradigm for practical and effective federated systems.
♻ ☆ Generative AI Models for Different Steps in Architectural Design: A Literature Review
Recent advances in generative artificial intelligence (AI) technologies have been significantly driven by models such as generative adversarial networks (GANs), variational autoencoders (VAEs), and denoising diffusion probabilistic models (DDPMs). Although architects recognize the potential of generative AI in design, personal barriers often restrict their access to the latest technological developments, thereby causing the application of generative AI in architectural design to lag behind. Therefore, it is essential to comprehend the principles and advancements of generative AI models and analyze their relevance in architecture applications. This paper first provides an overview of generative AI technologies, with a focus on probabilistic diffusion models (DDPMs), 3D generative models, and foundation models, highlighting their recent developments and main application scenarios. Then, the paper explains how the abovementioned models could be utilized in architecture. We subdivide the architectural design process into six steps and review related research projects in each step from 2020 to the present. Lastly, this paper discusses potential future directions for applying generative AI in the architectural design steps. This research can help architects quickly understand the development and latest progress of generative AI and contribute to the further development of intelligent architecture.
comment: 34 pages, 14 figures, accepted by Frontiers of Architectural Research
♻ ☆ Simplifying Deep Temporal Difference Learning
Q-learning played a foundational role in the field reinforcement learning (RL). However, TD algorithms with off-policy data, such as Q-learning, or nonlinear function approximation like deep neural networks require several additional tricks to stabilise training, primarily a replay buffer and target networks. Unfortunately, the delayed updating of frozen network parameters in the target network harms the sample efficiency and, similarly, the replay buffer introduces memory and implementation overheads. In this paper, we investigate whether it is possible to accelerate and simplify TD training while maintaining its stability. Our key theoretical result demonstrates for the first time that regularisation techniques such as LayerNorm can yield provably convergent TD algorithms without the need for a target network, even with off-policy data. Empirically, we find that online, parallelised sampling enabled by vectorised environments stabilises training without the need of a replay buffer. Motivated by these findings, we propose PQN, our simplified deep online Q-Learning algorithm. Surprisingly, this simple algorithm is competitive with more complex methods like: Rainbow in Atari, R2D2 in Hanabi, QMix in Smax, PPO-RNN in Craftax, and can be up to 50x faster than traditional DQN without sacrificing sample efficiency. In an era where PPO has become the go-to RL algorithm, PQN reestablishes Q-learning as a viable alternative.
♻ ☆ Stable generative modeling using Schrödinger bridges
We consider the problem of sampling from an unknown distribution for which only a sufficiently large number of training samples are available. Such settings have recently drawn considerable interest in the context of generative modelling and Bayesian inference. In this paper, we propose a generative model combining Schr\"odinger bridges and Langevin dynamics. Schr\"odinger bridges over an appropriate reversible reference process are used to approximate the conditional transition probability from the available training samples, which is then implemented in a discrete-time reversible Langevin sampler to generate new samples. By setting the kernel bandwidth in the reference process to match the time step size used in the unadjusted Langevin algorithm, our method effectively circumvents any stability issues typically associated with the time-stepping of stiff stochastic differential equations. Moreover, we introduce a novel split-step scheme, ensuring that the generated samples remain within the convex hull of the training samples. Our framework can be naturally extended to generate conditional samples and to Bayesian inference problems. We demonstrate the performance of our proposed scheme through experiments on synthetic datasets with increasing dimensions and on a stochastic subgrid-scale parametrization conditional sampling problem as well as generating sample trajectories of a dynamical system using conditional sampling.
♻ ☆ Causality-Aware Spatiotemporal Graph Neural Networks for Spatiotemporal Time Series Imputation CIKM'2024
Spatiotemporal time series are usually collected via monitoring sensors placed at different locations, which usually contain missing values due to various failures, such as mechanical damages and Internet outages. Imputing the missing values is crucial for analyzing time series. When recovering a specific data point, most existing methods consider all the information relevant to that point regardless of the cause-and-effect relationship. During data collection, it is inevitable that some unknown confounders are included, e.g., background noise in time series and non-causal shortcut edges in the constructed sensor network. These confounders could open backdoor paths and establish non-causal correlations between the input and output. Over-exploiting these non-causal correlations could cause overfitting. In this paper, we first revisit spatiotemporal time series imputation from a causal perspective and show how to block the confounders via the frontdoor adjustment. Based on the results of frontdoor adjustment, we introduce a novel Causality-Aware Spatiotemporal Graph Neural Network (Casper), which contains a novel Prompt Based Decoder (PBD) and a Spatiotemporal Causal Attention (SCA). PBD could reduce the impact of confounders and SCA could discover the sparse causal relationships among embeddings. Theoretical analysis reveals that SCA discovers causal relationships based on the values of gradients. We evaluate Casper on three real-world datasets, and the experimental results show that Casper could outperform the baselines and could effectively discover causal relationships.
comment: Accepted by CIKM'2024. Fixed typos
♻ ☆ MoC-System: Efficient Fault Tolerance for Sparse Mixture-of-Experts Model Training
As large language models continue to scale up, distributed training systems have expanded beyond 10k nodes, intensifying the importance of fault tolerance. Checkpoint has emerged as the predominant fault tolerance strategy, with extensive studies dedicated to optimizing its efficiency. However, the advent of the sparse Mixture-of-Experts (MoE) model presents new challenges due to the substantial increase in model size, despite comparable computational demands to dense models. In this work, we propose the Mixture-of-Checkpoint System (MoC-System) to orchestrate the vast array of checkpoint shards produced in distributed training systems. MoC-System features a novel Partial Experts Checkpointing (PEC) mechanism, an algorithm-system co-design that strategically saves a selected subset of experts, effectively reducing the MoE checkpoint size to levels comparable with dense models. Incorporating hybrid parallel strategies, MoC-System involves fully sharded checkpointing strategies to evenly distribute the workload across distributed ranks. Furthermore, MoC-System introduces a two-level checkpointing management method that asynchronously handles in-memory snapshots and persistence processes. We build MoC-System upon the Megatron-DeepSpeed framework, achieving up to a 98.9% reduction in overhead for each checkpointing process compared to the original method, during MoE model training with ZeRO-2 data parallelism and expert parallelism. Additionally, extensive empirical analyses substantiate that our methods enhance efficiency while maintaining comparable model accuracy, even achieving an average accuracy increase of 1.08% on downstream tasks.
♻ ☆ Automated Contrastive Learning Strategy Search for Time Series CIKM'2024
In recent years, Contrastive Learning (CL) has become a predominant representation learning paradigm for time series. Most existing methods manually build specific CL Strategies (CLS) by human heuristics for certain datasets and tasks. However, manually developing CLS usually requires excessive prior knowledge about the data, and massive experiments to determine the detailed CL configurations. In this paper, we present an Automated Machine Learning (AutoML) practice at Microsoft, which automatically learns CLS for time series datasets and tasks, namely Automated Contrastive Learning (AutoCL). We first construct a principled search space of size over $3\times10^{12}$, covering data augmentation, embedding transformation, contrastive pair construction, and contrastive losses. Further, we introduce an efficient reinforcement learning algorithm, which optimizes CLS from the performance on the validation tasks, to obtain effective CLS within the space. Experimental results on various real-world datasets demonstrate that AutoCL could automatically find the suitable CLS for the given dataset and task. From the candidate CLS found by AutoCL on several public datasets/tasks, we compose a transferable Generally Good Strategy (GGS), which has a strong performance for other datasets. We also provide empirical analysis as a guide for the future design of CLS.
comment: Accepted by CIKM'2024. Fixed typos
♻ ☆ Bayesian Analysis of Combinatorial Gaussian Process Bandits
We consider the combinatorial volatile Gaussian process (GP) semi-bandit problem. Each round, an agent is provided a set of available base arms and must select a subset of them to maximize the long-term cumulative reward. We study the Bayesian setting and provide novel Bayesian cumulative regret bounds for three GP-based algorithms: GP-UCB, GP-BayesUCB and GP-TS. Our bounds extend previous results for GP-UCB and GP-TS to the infinite, volatile and combinatorial setting, and to the best of our knowledge, we provide the first regret bound for GP-BayesUCB. Volatile arms encompass other widely considered bandit problems such as contextual bandits. Furthermore, we employ our framework to address the challenging real-world problem of online energy-efficient navigation, where we demonstrate its effectiveness compared to the alternatives.
comment: 32 pages, 10 figures
♻ ☆ On the explainability of quantum neural networks based on variational quantum circuits
Ridge functions are used to describe and study the lower bound of the approximation done by the neural networks which can be written as a linear combination of activation functions. If the activation functions are also ridge functions, these networks are called explainable neural networks. In this brief paper, we first show that quantum neural networks which are based on variational quantum circuits can be written as a linear combination of ridge functions by following matrix notations. Consequently, we show that the interpretability and explainability of such quantum neural networks can be directly considered and studied as an approximation with the linear combination of ridge functions.
comment: a brief paper,a few missing references have been added
♻ ☆ Conformal Prediction for Causal Effects of Continuous Treatments
Uncertainty quantification of causal effects is crucial for safety-critical applications such as personalized medicine. A powerful approach for this is conformal prediction, which has several practical benefits due to model-agnostic finite-sample guarantees. Yet, existing methods for conformal prediction of causal effects are limited to binary/discrete treatments and make highly restrictive assumptions such as known propensity scores. In this work, we provide a novel conformal prediction method for potential outcomes of continuous treatments. We account for the additional uncertainty introduced through propensity estimation so that our conformal prediction intervals are valid even if the propensity score is unknown. Our contributions are three-fold: (1) We derive finite-sample prediction intervals for potential outcomes of continuous treatments. (2) We provide an algorithm for calculating the derived intervals. (3) We demonstrate the effectiveness of the conformal prediction intervals in experiments on synthetic and real-world datasets. To the best of our knowledge, we are the first to propose conformal prediction for continuous treatments when the propensity score is unknown and must be estimated from data.
♻ ☆ Over-the-Air Federated Learning in Cell-Free MIMO with Long-term Power Constraint
Wireless networks supporting artificial intelligence have gained significant attention, with Over-the-Air Federated Learning emerging as a key application due to its unique transmission and distributed computing characteristics. This paper derives error bounds for Over-the-Air Federated Learning in a Cell-free MIMO system and formulates an optimization problem to minimize optimality gap via joint optimization of power control and beamforming. We introduce the MOP-LOFPC algorithm, which employs Lyapunov optimization to decouple long-term constraints across rounds while requiring only causal channel state information. Experimental results demonstrate that MOP-LOFPC achieves a better and more flexible trade-off between the model's training loss and adherence to long-term power constraints compared to existing baselines.
♻ ☆ Improve Value Estimation of Q Function and Reshape Reward with Monte Carlo Tree Search
Reinforcement learning has achieved remarkable success in perfect information games such as Go and Atari, enabling agents to compete at the highest levels against human players. However, research in reinforcement learning for imperfect information games has been relatively limited due to the more complex game structures and randomness. Traditional methods face challenges in training and improving performance in imperfect information games due to issues like inaccurate Q value estimation and reward sparsity. In this paper, we focus on Uno, an imperfect information game, and aim to address these problems by reducing Q value overestimation and reshaping reward function. We propose a novel algorithm that utilizes Monte Carlo Tree Search to average the value estimations in Q function. Even though we choose Double Deep Q Learning as the foundational framework in this paper, our method can be generalized and used in any algorithm which needs Q value estimation, such as the Actor-Critic. Additionally, we employ Monte Carlo Tree Search to reshape the reward structure in the game environment. We compare our algorithm with several traditional methods applied to games such as Double Deep Q Learning, Deep Monte Carlo and Neural Fictitious Self Play, and the experiments demonstrate that our algorithm consistently outperforms these approaches, especially as the number of players in Uno increases, indicating a higher level of difficulty.
♻ ☆ Using Stochastic Gradient Descent to Smooth Nonconvex Functions: Analysis of Implicit Graduated Optimization
The graduated optimization approach is a heuristic method for finding global optimal solutions for nonconvex functions by using a function smoothing operation with stochastic noise. We show that stochastic noise in stochastic gradient descent (SGD) has the effect of smoothing the objective function, the degree of which is determined by the learning rate, batch size, and variance of the stochastic gradient. Using this finding, we propose and analyze a new graduated optimization algorithm that varies the degree of smoothing by varying the learning rate and batch size, and provide experimental results on image classification tasks with ResNets that support our theoretical findings. We further show that there is an interesting correlation between the degree of smoothing by SGD's stochastic noise, the well-studied ``sharpness'' indicator, and the generalization performance of the model.
comment: The latest version was updated in October 2024. Under review
♻ ☆ Do causal predictors generalize better to new domains? NeurIPS'24
We study how well machine learning models trained on causal features generalize across domains. We consider 16 prediction tasks on tabular datasets covering applications in health, employment, education, social benefits, and politics. Each dataset comes with multiple domains, allowing us to test how well a model trained in one domain performs in another. For each prediction task, we select features that have a causal influence on the target of prediction. Our goal is to test the hypothesis that models trained on causal features generalize better across domains. Without exception, we find that predictors using all available features, regardless of causality, have better in-domain and out-of-domain accuracy than predictors using causal features. Moreover, even the absolute drop in accuracy from one domain to the other is no better for causal predictors than for models that use all features. In addition, we show that recent causal machine learning methods for domain generalization do not perform better in our evaluation than standard predictors trained on the set of causal features. Likewise, causal discovery algorithms either fail to run or select causal variables that perform no better than our selection. Extensive robustness checks confirm that our findings are stable under variable misclassification.
comment: 118 pages, 55 figures, accepted at NeurIPS'24
♻ ☆ Explainable Hierarchical Urban Representation Learning for Commuting Flow Prediction
Commuting flow prediction is an essential task for municipal operations in the real world. Previous studies have revealed that it is feasible to estimate the commuting origin-destination (OD) demand within a city using multiple auxiliary data. However, most existing methods are not suitable to deal with a similar task at a large scale, namely within a prefecture or the whole nation, owing to the increased number of geographical units that need to be maintained. In addition, region representation learning is a universal approach for gaining urban knowledge for diverse metropolitan downstream tasks. Although many researchers have developed comprehensive frameworks to describe urban units from multi-source data, they have not clarified the relationship between the selected geographical elements. Furthermore, metropolitan areas naturally preserve ranked structures, like cities and their inclusive districts, which makes elucidating relations between cross-level urban units necessary. Therefore, we develop a heterogeneous graph-based model to generate meaningful region embeddings at multiple spatial resolutions for predicting different types of inter-level OD flows. To demonstrate the effectiveness of the proposed method, extensive experiments were conducted using real-world aggregated mobile phone datasets collected from Shizuoka Prefecture, Japan. The results indicate that our proposed model outperforms existing models in terms of a uniform urban structure. We extend the understanding of predicted results using reasonable explanations to enhance the credibility of the model.
♻ ☆ Generative Forests NeurIPS'24
We focus on generative AI for a type of data that still represent one of the most prevalent form of data: tabular data. Our paper introduces two key contributions: a new powerful class of forest-based models fit for such tasks and a simple training algorithm with strong convergence guarantees in a boosting model that parallels that of the original weak / strong supervised learning setting. This algorithm can be implemented by a few tweaks to the most popular induction scheme for decision tree induction (i.e. supervised learning) with two classes. Experiments on the quality of generated data display substantial improvements compared to the state of the art. The losses our algorithm minimize and the structure of our models make them practical for related tasks that require fast estimation of a density given a generative model and an observation (even partially specified): such tasks include missing data imputation and density estimation. Additional experiments on these tasks reveal that our models can be notably good contenders to diverse state of the art methods, relying on models as diverse as (or mixing elements of) trees, neural nets, kernels or graphical models.
comment: NeurIPS'24
♻ ☆ Probabilistic ML Verification via Weighted Model Integration
In machine learning (ML) verification, the majority of procedures are non-quantitative and therefore cannot be used for verifying probabilistic models, or be applied in domains where hard guarantees are practically unachievable. The probabilistic formal verification (PFV) of ML models is in its infancy, with the existing approaches limited to specific ML models, properties, or both. This contrasts with standard formal methods techniques, whose successful adoption in real-world scenarios is also due to their support for a wide range of properties and diverse systems. We propose a unifying framework for the PFV of ML systems based on Weighted Model Integration (WMI), a relatively recent formalism for probabilistic inference with algebraic and logical constraints. Crucially, reducing the PFV of ML models to WMI enables the verification of many properties of interest over a wide range of systems, addressing multiple limitations of deterministic verification and ad-hoc algorithms. We substantiate the generality of the approach on prototypical tasks involving the verification of group fairness, monotonicity, robustness to noise, probabilistic local robustness and equivalence among predictors. We characterize the challenges related to the scalability of the approach and, through our WMI-based perspective, we show how successful scaling techniques in the ML verification literature can be generalized beyond their original scope.
♻ ☆ Multi-Excitation Projective Simulation with a Many-Body Physics Inspired Inductive Bias
With the impressive progress of deep learning, applications relying on machine learning are increasingly being integrated into daily life. However, most deep learning models have an opaque, oracle-like nature making it difficult to interpret and understand their decisions. This problem led to the development of the field known as eXplainable Artificial Intelligence (XAI). One method in this field known as Projective Simulation (PS) models a chain-of-thought as a random walk of a particle on a graph with vertices that have concepts attached to them. While this description has various benefits, including the possibility of quantization, it cannot be naturally used to model thoughts that combine several concepts simultaneously. To overcome this limitation, we introduce Multi-Excitation Projective Simulation (mePS), a generalization that considers a chain-of-thought to be a random walk of several particles on a hypergraph. A definition for a dynamic hypergraph is put forward to describe the agent's training history along with applications to AI and hypergraph visualization. An inductive bias inspired by the remarkably successful few-body interaction models used in quantum many-body physics is formalized for our classical mePS framework and employed to tackle the exponential complexity associated with naive implementations of hypergraphs. We prove that our inductive bias reduces the complexity from exponential to polynomial, with the exponent representing the cutoff on how many particles can interact. We numerically apply our method to two toy environments and a more complex scenario modelling the diagnosis of a broken computer. These environments demonstrate the resource savings provided by an appropriate choice of inductive bias, as well as showcasing aspects of interpretability. A quantum model for mePS is also briefly outlined and some future directions for it are discussed.
comment: 26 pages, 8 figures; Code repository at https://github.com/MariusKrumm/ManyBodyMEPS. Reorganized main text for better readability
♻ ☆ Uncertainty Estimation and Quantification for LLMs: A Simple Supervised Approach
In this paper, we study the problem of uncertainty estimation and calibration for LLMs. We begin by formulating the uncertainty estimation problem, a relevant yet underexplored area in existing literature. We then propose a supervised approach that leverages labeled datasets to estimate the uncertainty in LLMs' responses. Based on the formulation, we illustrate the difference between the uncertainty estimation for LLMs and that for standard ML models and explain why the hidden neurons of the LLMs may contain uncertainty information. Our designed approach demonstrates the benefits of utilizing hidden activations to enhance uncertainty estimation across various tasks and shows robust transferability in out-of-distribution settings. We distinguish the uncertainty estimation task from the uncertainty calibration task and show that better uncertainty estimation leads to better calibration performance. Furthermore, our method is easy to implement and adaptable to different levels of model accessibility including black box, grey box, and white box.
comment: 29 pages, 14 figures
♻ ☆ Towards Foundation Model for Chemical Reactor Modeling: Meta-Learning with Physics-Informed Adaptation
In this work, we present a novel application of foundation models for chemical reactor modeling. Accurate modeling of real-world chemical reactors through first-principles is often challenging, and the process of rebuilding and retraining models for each new chemical process is inefficient. This raises a critical question: can we develop a single, universal neural network (i.e., a foundation model) that can rapidly adapt to any new chemical process in a reactor? To address this, we propose a foundation model for chemical reactor modeling that employs a meta-learning approach, followed by physics-informed fine-tuning on new tasks with only a few data samples. Our model is designed to generalize across three classic reactor types: continuous stirred tank reactors, batch reactors, and plug flow reactors. Compared to conventional methods such as data-driven learning, physics-informed learning, transfer learning, and meta-learning, our approach demonstrates superior performance in few-shot scenarios. Specifically, it shows rapid adaptation to unseen reactions with varying integer orders across different reactor set-ups, requiring minimal data for fine-tuning. Source code is available at https://github.com/killingbear999/chemical-reactor-foundation-model.
♻ ☆ xLSTM-Mixer: Multivariate Time Series Forecasting by Mixing via Scalar Memories
Time series data is prevalent across numerous fields, necessitating the development of robust and accurate forecasting models. Capturing patterns both within and between temporal and multivariate components is crucial for reliable predictions. We introduce xLSTM-Mixer, a model designed to effectively integrate temporal sequences, joint time-variate information, and multiple perspectives for robust forecasting. Our approach begins with a linear forecast shared across variates, which is then refined by xLSTM blocks. These blocks serve as key elements for modeling the complex dynamics of challenging time series data. xLSTM-Mixer ultimately reconciles two distinct views to produce the final forecast. Our extensive evaluations demonstrate xLSTM-Mixer's superior long-term forecasting performance compared to recent state-of-the-art methods. A thorough model analysis provides further insights into its key components and confirms its robustness and effectiveness. This work contributes to the resurgence of recurrent models in time series forecasting.
♻ ☆ The Art of Imitation: Learning Long-Horizon Manipulation Tasks from Few Demonstrations
Task Parametrized Gaussian Mixture Models (TP-GMM) are a sample-efficient method for learning object-centric robot manipulation tasks. However, there are several open challenges to applying TP-GMMs in the wild. In this work, we tackle three crucial challenges synergistically. First, end-effector velocities are non-Euclidean and thus hard to model using standard GMMs. We thus propose to factorize the robot's end-effector velocity into its direction and magnitude, and model them using Riemannian GMMs. Second, we leverage the factorized velocities to segment and sequence skills from complex demonstration trajectories. Through the segmentation, we further align skill trajectories and hence leverage time as a powerful inductive bias. Third, we present a method to automatically detect relevant task parameters per skill from visual observations. Our approach enables learning complex manipulation tasks from just five demonstrations while using only RGB-D observations. Extensive experimental evaluations on RLBench demonstrate that our approach achieves state-of-the-art performance with 20-fold improved sample efficiency. Our policies generalize across different environments, object instances, and object positions, while the learned skills are reusable.
♻ ☆ Integral Operator Approaches for Scattered Data Fitting on Spheres
This paper focuses on scattered data fitting problems on spheres. We study the approximation performance of a class of weighted spectral filter algorithms, including Tikhonov regularization, Landaweber iteration, spectral cut-off, and iterated Tikhonov, in fitting noisy data with possibly unbounded random noise. For the analysis, we develop an integral operator approach that can be regarded as an extension of the widely used sampling inequality approach and norming set method in the community of scattered data fitting. After providing an equivalence between the operator differences and quadrature rules, we succeed in deriving optimal Sobolev-type error estimates of weighted spectral filter algorithms. Our derived error estimates do not suffer from the saturation phenomenon for Tikhonov regularization in the literature, native-space-barrier for existing error analysis and adapts to different embedding spaces. We also propose a divide-and-conquer scheme to equip weighted spectral filter algorithms to reduce their computational burden and present the optimal approximation error bounds.
♻ ☆ Hadamard Representations: Augmenting Hyperbolic Tangents in RL
Activation functions are one of the key components of a deep neural network. The most commonly used activation functions can be classed into the category of continuously differentiable (e.g. tanh) and linear-unit functions (e.g. ReLU), both having their own strengths and drawbacks with respect to downstream performance and representation capacity through learning (e.g. measured by the number of dead neurons and the effective rank). In reinforcement learning, the performance of continuously differentiable activations often falls short as compared to linear-unit functions. We provide insights into the vanishing gradients associated with the former, and show that the dying neuron problem is not exclusive to ReLU's. To alleviate vanishing gradients and the resulting dying neuron problem occurring with continuously differentiable activations, we propose a Hadamard representation. Using deep Q-networks and proximal policy optimization in the Atari domain, we show faster learning, a reduction in dead neurons and increased effective rank.
comment: 24 pages, 19 figures, 3 tables
♻ ☆ Binarized Simplicial Convolutional Neural Networks
Graph Neural Networks have a limitation of solely processing features on graph nodes, neglecting data on high-dimensional structures such as edges and triangles. Simplicial Convolutional Neural Networks (SCNN) represent higher-order structures using simplicial complexes to break this limitation albeit still lacking time efficiency. In this paper, we propose a novel neural network architecture on simplicial complexes named Binarized Simplicial Convolutional Neural Networks (Bi-SCNN) based on the combination of simplicial convolution with a binary-sign forward propagation strategy. The usage of the Hodge Laplacian on a binary-sign forward propagation enables Bi-SCNN to efficiently and effectively represent simplicial features that have higher-order structures than traditional graph node representations. Compared to the previous Simplicial Convolutional Neural Networks, the reduced model complexity of Bi-SCNN shortens the execution time without sacrificing the prediction performance and is less prone to the over-smoothing effect. Experimenting with real-world citation and ocean-drifter data confirmed that our proposed Bi-SCNN is efficient and accurate.
♻ ☆ Understanding Gradient Boosting Classifier: Training, Prediction, and the Role of $γ_j$
The Gradient Boosting Classifier (GBC) is a widely used machine learning algorithm for binary classification, which builds decision trees iteratively to minimize prediction errors. This document explains the GBC's training and prediction processes, focusing on the computation of terminal node values $\gamma_j$, which are crucial to optimizing the logistic loss function. We derive $\gamma_j$ through a Taylor series approximation and provide a step-by-step pseudocode for the algorithm's implementation. The guide explains the theory of GBC and its practical application, demonstrating its effectiveness in binary classification tasks. We provide a step-by-step example in the appendix to help readers understand.
♻ ☆ Statistical Efficiency of Distributional Temporal Difference Learning NeurIPS 2024
Distributional reinforcement learning (DRL) has achieved empirical success in various domains. One core task in the field of DRL is distributional policy evaluation, which involves estimating the return distribution $\eta^\pi$ for a given policy $\pi$. The distributional temporal difference learning has been accordingly proposed, which is an extension of the temporal difference learning (TD) in the classic RL area. In the tabular case, \citet{rowland2018analysis} and \citet{rowland2023analysis} proved the asymptotic convergence of two instances of distributional TD, namely categorical temporal difference learning (CTD) and quantile temporal difference learning (QTD), respectively. In this paper, we go a step further and analyze the finite-sample performance of distributional TD. To facilitate theoretical analysis, we propose non-parametric distributional TD learning (NTD). For a $\gamma$-discounted infinite-horizon tabular Markov decision process, we show that for NTD we need $\tilde{O}\left(\frac{1}{\varepsilon^{2p}(1-\gamma)^{2p+1}}\right)$ iterations to achieve an $\varepsilon$-optimal estimator with high probability, when the estimation error is measured by the $p$-Wasserstein distance. This sample complexity bound is minimax optimal up to logarithmic factors in the case of the $1$-Wasserstein distance. To achieve this, we establish a novel Freedman's inequality in Hilbert spaces, which would be of independent interest. In addition, we revisit CTD, showing that the same non-asymptotic convergence bounds hold for CTD in the case of the $p$-Wasserstein distance for $p\geq 1$.
comment: NeurIPS 2024 (oral)
♻ ☆ On the Design and Performance of Machine Learning Based Error Correcting Decoders
This paper analyzes the design and competitiveness of four neural network (NN) architectures recently proposed as decoders for forward error correction (FEC) codes. We first consider the so-called single-label neural network (SLNN) and the multi-label neural network (MLNN) decoders which have been reported to achieve near maximum likelihood (ML) performance. Here, we show analytically that SLNN and MLNN decoders can always achieve ML performance, regardless of the code dimensions -- although at the cost of computational complexity -- and no training is in fact required. We then turn our attention to two transformer-based decoders: the error correction code transformer (ECCT) and the cross-attention message passing transformer (CrossMPT). We compare their performance against traditional decoders, and show that ordered statistics decoding outperforms these transformer-based decoders. The results in this paper cast serious doubts on the application of NN-based FEC decoders in the short and medium block length regime.
comment: 6 pages, 4 figures, submitted for possible presentation in a conference (v2: Pre-FEC BER curves are corrected)
♻ ☆ Adaptive Spatio-temporal Estimation on the Graph Edges via Line Graph Transformation
Spatio-temporal estimation of signals on graph edges is challenging because most conventional Graph Signal Processing techniques are defined on the graph nodes. Leveraging the Line Graph transform, the Line Graph Least Mean Square (LGLMS) algorithm is proposed to conduct adaptive estimation of time-varying edge signals by projecting the edge signals from edge space to node space. LGLMS is an adaptive algorithm analogous to the classical LMS algorithm but applied to graph edges. Unlike edge-specific methods, LGLMS retains all GSP concepts and techniques originally designed for graph nodes, without the need for redefinition on the edges. Experimenting with transportation graphs and meteorological graphs, with the signal observations having noisy and missing values, we confirmed that LGLMS is suitable for the online prediction of time-varying edge signals.
♻ ☆ Understanding Transfer Learning via Mean-field Analysis
We propose a novel framework for exploring generalization errors of transfer learning through the lens of differential calculus on the space of probability measures. In particular, we consider two main transfer learning scenarios, $\alpha$-ERM and fine-tuning with the KL-regularized empirical risk minimization and establish generic conditions under which the generalization error and the population risk convergence rates for these scenarios are studied. Based on our theoretical results, we show the benefits of transfer learning with a one-hidden-layer neural network in the mean-field regime under some suitable integrability and regularity assumptions on the loss and activation functions.
comment: Under review
♻ ☆ Universal approximation results for neural networks with non-polynomial activation function over non-compact domains
In this paper, we generalize the universal approximation property of single-hidden-layer feed-forward neural networks beyond the classical formulation over compact domains. More precisely, by assuming that the activation function is non-polynomial, we derive universal approximation results for neural networks within function spaces over non-compact subsets of a Euclidean space, e.g., weighted spaces, $L^p$-spaces, and (weighted) Sobolev spaces over unbounded domains, where the latter includes the approximation of the (weak) derivatives. Furthermore, we provide some dimension-independent rates for approximating a function with sufficiently regular and integrable Fourier transform by neural networks with non-polynomial activation function.
comment: arXiv admin note: text overlap with arXiv:2312.08410
♻ ☆ Generative AI Security: Challenges and Countermeasures
Generative AI's expanding footprint across numerous industries has led to both excitement and increased scrutiny. This paper delves into the unique security challenges posed by Generative AI, and outlines potential research directions for managing these risks.
♻ ☆ Empirical investigation of multi-source cross-validation in clinical ECG classification
Traditionally, machine learning-based clinical prediction models have been trained and evaluated on patient data from a single source, such as a hospital. Cross-validation methods can be used to estimate the accuracy of such models on new patients originating from the same source, by repeated random splitting of the data. However, such estimates tend to be highly overoptimistic when compared to accuracy obtained from deploying models to sources not represented in the dataset, such as a new hospital. The increasing availability of multi-source medical datasets provides new opportunities for obtaining more comprehensive and realistic evaluations of expected accuracy through source-level cross-validation designs. In this study, we present a systematic empirical evaluation of standard K-fold cross-validation and leave-source-out cross-validation methods in a multi-source setting. We consider the task of electrocardiogram based cardiovascular disease classification, combining and harmonizing the openly available PhysioNet CinC Challenge 2021 and the Shandong Provincial Hospital datasets for our study. Our results show that K-fold cross-validation, both on single-source and multi-source data, systemically overestimates prediction performance when the end goal is to generalize to new sources. Leave-source-out cross-validation provides more reliable performance estimates, having close to zero bias though larger variability. The evaluation highlights the dangers of obtaining misleading cross-validation results on medical data and demonstrates how these issues can be mitigated when having access to multi-source data.
comment: 19 pages, 5 figures
♻ ☆ P1-KAN an effective Kolmogorov Arnold Network for function approximation
A new Kolmogorov-Arnold network (KAN) is proposed to approximate potentially irregular functions in high dimension. We show that it outperforms multilayer perceptrons in terms of accuracy and converges faster. We also compare it with several proposed KAN networks: the original spline-based KAN network appears to be more effective for smooth functions, while the P1-KAN network is more effective for irregular functions.
♻ ☆ Masked Clinical Modelling: A Framework for Synthetic and Augmented Survival Data Generation
Access to real clinical data is often restricted due to privacy obligations, creating significant barriers for healthcare research. Synthetic datasets provide a promising solution, enabling secure data sharing and model development. However, most existing approaches focus on data realism rather than utility -- ensuring that models trained on synthetic data yield clinically meaningful insights comparable to those trained on real data. In this paper, we present Masked Clinical Modelling (MCM), a framework inspired by masked language modelling, designed for both data synthesis and conditional data augmentation. We evaluate this prototype on the WHAS500 dataset using Cox Proportional Hazards models, focusing on the preservation of hazard ratios as key clinical metrics. Our results show that data generated using the MCM framework improves both discrimination and calibration in survival analysis, outperforming existing methods. MCM demonstrates strong potential to support survival data analysis and broader healthcare applications.
comment: Re-archived due to incorrect ORCiD. Last edited: 2024-10-23
♻ ☆ Advancing Open-Set Domain Generalization Using Evidential Bi-Level Hardest Domain Scheduler NeurIPS 2024
In Open-Set Domain Generalization (OSDG), the model is exposed to both new variations of data appearance (domains) and open-set conditions, where both known and novel categories are present at test time. The challenges of this task arise from the dual need to generalize across diverse domains and accurately quantify category novelty, which is critical for applications in dynamic environments. Recently, meta-learning techniques have demonstrated superior results in OSDG, effectively orchestrating the meta-train and -test tasks by employing varied random categories and predefined domain partition strategies. These approaches prioritize a well-designed training schedule over traditional methods that focus primarily on data augmentation and the enhancement of discriminative feature learning. The prevailing meta-learning models in OSDG typically utilize a predefined sequential domain scheduler to structure data partitions. However, a crucial aspect that remains inadequately explored is the influence brought by strategies of domain schedulers during training. In this paper, we observe that an adaptive domain scheduler benefits more in OSDG compared with prefixed sequential and random domain schedulers. We propose the Evidential Bi-Level Hardest Domain Scheduler (EBiL-HaDS) to achieve an adaptive domain scheduler. This method strategically sequences domains by assessing their reliabilities in utilizing a follower network, trained with confidence scores learned in an evidential manner, regularized by max rebiasing discrepancy, and optimized in a bi-level manner. The results show that our method substantially improves OSDG performance and achieves more discriminative embeddings for both the seen and unseen categories. The source code is publicly available at https://github.com/KPeng9510/EBiL-HaDS.
comment: Accepted to NeurIPS 2024. The source code is publicly available at https://github.com/KPeng9510/EBiL-HaDS
♻ ☆ Diffusion-Reward Adversarial Imitation Learning
Imitation learning aims to learn a policy from observing expert demonstrations without access to reward signals from environments. Generative adversarial imitation learning (GAIL) formulates imitation learning as adversarial learning, employing a generator policy learning to imitate expert behaviors and discriminator learning to distinguish the expert demonstrations from agent trajectories. Despite its encouraging results, GAIL training is often brittle and unstable. Inspired by the recent dominance of diffusion models in generative modeling, we propose Diffusion-Reward Adversarial Imitation Learning (DRAIL), which integrates a diffusion model into GAIL, aiming to yield more robust and smoother rewards for policy learning. Specifically, we propose a diffusion discriminative classifier to construct an enhanced discriminator, and design diffusion rewards based on the classifier's output for policy learning. Extensive experiments are conducted in navigation, manipulation, and locomotion, verifying DRAIL's effectiveness compared to prior imitation learning methods. Moreover, additional experimental results demonstrate the generalizability and data efficiency of DRAIL. Visualized learned reward functions of GAIL and DRAIL suggest that DRAIL can produce more robust and smoother rewards. Project page: https://nturobotlearninglab.github.io/DRAIL/
♻ ☆ UCB Exploration for Fixed-Budget Bayesian Best Arm Identification
We study best-arm identification (BAI) in the fixed-budget setting. Adaptive allocations based on upper confidence bounds (UCBs), such as UCBE, are known to work well in BAI. However, it is well-known that its optimal regret is theoretically dependent on instances, which we show to be an artifact in many fixed-budget BAI problems. In this paper we propose an UCB exploration algorithm that is both theoretically and empirically efficient for the fixed budget BAI problem under a Bayesian setting. The key idea is to learn prior information, which can enhance the performance of UCB-based BAI algorithm as it has done in the cumulative regret minimization problem. We establish bounds on the failure probability and the simple regret for the Bayesian BAI problem, providing upper bounds of order $\tilde{O}(\sqrt{K/n})$, up to logarithmic factors, where $n$ represents the budget and $K$ denotes the number of arms. Furthermore, we demonstrate through empirical results that our approach consistently outperforms state-of-the-art baselines.
♻ ☆ Diffusion Models are Certifiably Robust Classifiers NeurIPS 2024
Generative learning, recognized for its effective modeling of data distributions, offers inherent advantages in handling out-of-distribution instances, especially for enhancing robustness to adversarial attacks. Among these, diffusion classifiers, utilizing powerful diffusion models, have demonstrated superior empirical robustness. However, a comprehensive theoretical understanding of their robustness is still lacking, raising concerns about their vulnerability to stronger future attacks. In this study, we prove that diffusion classifiers possess $O(1)$ Lipschitzness, and establish their certified robustness, demonstrating their inherent resilience. To achieve non-constant Lipschitzness, thereby obtaining much tighter certified robustness, we generalize diffusion classifiers to classify Gaussian-corrupted data. This involves deriving the evidence lower bounds (ELBOs) for these distributions, approximating the likelihood using the ELBO, and calculating classification probabilities via Bayes' theorem. Experimental results show the superior certified robustness of these Noised Diffusion Classifiers (NDCs). Notably, we achieve over 80% and 70% certified robustness on CIFAR-10 under adversarial perturbations with \(\ell_2\) norms less than 0.25 and 0.5, respectively, using a single off-the-shelf diffusion model without any additional data.
comment: Accepted by NeurIPS 2024
♻ ☆ A Simple Baseline for Predicting Events with Auto-Regressive Tabular Transformers
Many real-world applications of tabular data involve using historic events to predict properties of new ones, for example whether a credit card transaction is fraudulent or what rating a customer will assign a product on a retail platform. Existing approaches to event prediction include costly, brittle, and application-dependent techniques such as time-aware positional embeddings, learned row and field encodings, and oversampling methods for addressing class imbalance. Moreover, these approaches often assume specific use-cases, for example that we know the labels of all historic events or that we only predict a pre-specified label and not the data's features themselves. In this work, we propose a simple but flexible baseline using standard autoregressive LLM-style transformers with elementary positional embeddings and a causal language modeling objective. Our baseline outperforms existing approaches across popular datasets and can be employed for various use-cases. We demonstrate that the same model can predict labels, impute missing values, or model event sequences.
comment: 10 pages, 6 pages of references+appendix
♻ ☆ Deep Learning for Detecting and Early Predicting Chronic Obstructive Pulmonary Disease from Spirogram Time Series
Chronic Obstructive Pulmonary Disease (COPD) is a chronic lung disease that causes airflow obstruction. Current methods can only detect COPD from prominent features in spirogram (Volume-Flow time series) but cannot predict future COPD risk from subtle data patterns. We propose a deep learning-based method, DeepSpiro, for early prediction of future COPD risk. DeepSpiro consists of four key components: SpiroSmoother for stabilizing the Volume-Flow curve, SpiroEncoder for capturing volume evolution through key patches of varying lengths, SpiroExplainer for integrating heterogeneous data and explaining predictions through volume attention, and SpiroPredictor for predicting the disease risk of undiagnosed high-risk patients based on key patch concavity, with prediction horizons of 1, 2, 3, 4, 5 years, or even longer. Evaluated on the UK Biobank dataset, DeepSpiro achieved an AUC of 0.8328 for COPD detection and demonstrated strong predictive performance for future COPD risk (p-value < 0.001). DeepSpiro effectively predicts the long-term progression of the disease.
♻ ☆ Utility Theory of Synthetic Data Generation
Synthetic data algorithms are widely employed in industries to generate artificial data for downstream learning tasks. While existing research primarily focuses on empirically evaluating utility of synthetic data, its theoretical understanding is largely lacking. This paper bridges the practice-theory gap by establishing relevant utility theory in a statistical learning framework. It considers two utility metrics: generalization and ranking of models trained on synthetic data. The former is defined as the generalization difference between models trained on synthetic and on real data. By deriving analytical bounds for this utility metric, we demonstrate that the synthetic feature distribution does not need to be similar as that of real data for ensuring comparable generalization of synthetic models, provided proper model specifications in downstream learning tasks. The latter utility metric studies the relative performance of models trained on synthetic data. In particular, we discover that the distribution of synthetic data is not necessarily similar as the real one to ensure consistent model comparison. Interestingly, consistent model comparison is still achievable even when synthetic responses are not well generated, as long as downstream models are separable by a generalization gap. Finally, extensive experiments on non-parametric models and deep neural networks have been conducted to validate these theoretical findings.
♻ ☆ Spectraformer: A Unified Random Feature Framework for Transformer
Linearization of attention using various kernel approximation and kernel learning techniques has shown promise. Past methods use a subset of combinations of component functions and weight matrices within the random features paradigm. We identify the need for a systematic comparison of different combinations of weight matrices and component functions for attention learning in Transformer. In this work, we introduce Spectraformer, a unified framework for approximating and learning the kernel function in linearized attention of the Transformer. We experiment with broad classes of component functions and weight matrices for three textual tasks in the LRA benchmark. Our empirical findings indicate that different kernels are good at different tasks and that kernel choice is fundamental to performant models. Our code is available at: https://github.com/dukenguyenxyz/spectraformer .
♻ ☆ Set-based Meta-Interpolation for Few-Task Meta-Learning
Meta-learning approaches enable machine learning systems to adapt to new tasks given few examples by leveraging knowledge from related tasks. However, a large number of meta-training tasks are still required for generalization to unseen tasks during meta-testing, which introduces a critical bottleneck for real-world problems that come with only few tasks, due to various reasons including the difficulty and cost of constructing tasks. Recently, several task augmentation methods have been proposed to tackle this issue using domain-specific knowledge to design augmentation techniques to densify the meta-training task distribution. However, such reliance on domain-specific knowledge renders these methods inapplicable to other domains. While Manifold Mixup based task augmentation methods are domain-agnostic, we empirically find them ineffective on non-image domains. To tackle these limitations, we propose a novel domain-agnostic task augmentation method, Meta-Interpolation, which utilizes expressive neural set functions to densify the meta-training task distribution using bilevel optimization. We empirically validate the efficacy of Meta-Interpolation on eight datasets spanning across various domains such as image classification, molecule property prediction, text classification and speech recognition. Experimentally, we show that Meta-Interpolation consistently outperforms all the relevant baselines. Theoretically, we prove that task interpolation with the set function regularizes the meta-learner to improve generalization.
comment: First two authors contributed equally. Name order decided by a coin toss
♻ ☆ Quantifying the Gain in Weak-to-Strong Generalization NeurIPS 2024
Recent advances in large language models have shown capabilities that are extraordinary and near-superhuman. These models operate with such complexity that reliably evaluating and aligning them proves challenging for humans. This leads to the natural question: can guidance from weak models (like humans) adequately direct the capabilities of strong models? In a recent and somewhat surprising work, Burns et al. (2023) empirically demonstrated that when strong models (like GPT-4) are finetuned using labels generated by weak supervisors (like GPT-2), the strong models outperform their weaker counterparts -- a phenomenon they term weak-to-strong generalization. In this work, we present a theoretical framework for understanding weak-to-strong generalization. Specifically, we show that the improvement in performance achieved by strong models over their weaker counterparts is quantified by the misfit error incurred by the strong model on labels generated by the weaker model. Our theory reveals several curious algorithmic insights. For instance, we can predict the amount by which the strong model will improve over the weak model, and also choose among different weak models to train the strong model, based on its misfit error. We validate our theoretical findings through various empirical assessments.
comment: 19 pages; NeurIPS 2024 camera-ready version with additional experiments, references and discussion
♻ ☆ LLMScan: Causal Scan for LLM Misbehavior Detection
Despite the success of Large Language Models (LLMs) across various fields, their potential to generate untruthful, biased and harmful responses poses significant risks, particularly in critical applications. This highlights the urgent need for systematic methods to detect and prevent such misbehavior. While existing approaches target specific issues such as harmful responses, this work introduces LLMScan, an innovative LLM monitoring technique based on causality analysis, offering a comprehensive solution. LLMScan systematically monitors the inner workings of an LLM through the lens of causal inference, operating on the premise that the LLM's `brain' behaves differently when misbehaving. By analyzing the causal contributions of the LLM's input tokens and transformer layers, LLMScan effectively detects misbehavior. Extensive experiments across various tasks and models reveal clear distinctions in the causal distributions between normal behavior and misbehavior, enabling the development of accurate, lightweight detectors for a variety of misbehavior detection tasks.
♻ ☆ Real-World Robot Applications of Foundation Models: A Review
Recent developments in foundation models, like Large Language Models (LLMs) and Vision-Language Models (VLMs), trained on extensive data, facilitate flexible application across different tasks and modalities. Their impact spans various fields, including healthcare, education, and robotics. This paper provides an overview of the practical application of foundation models in real-world robotics, with a primary emphasis on the replacement of specific components within existing robot systems. The summary encompasses the perspective of input-output relationships in foundation models, as well as their role in perception, motion planning, and control within the field of robotics. This paper concludes with a discussion of future challenges and implications for practical robot applications.
♻ ☆ Audio-visual cross-modality knowledge transfer for machine learning-based in-situ monitoring in laser additive manufacturing
Various machine learning (ML)-based in-situ monitoring systems have been developed to detect anomalies and defects in laser additive manufacturing (LAM) processes. While multimodal fusion, which integrates data from visual, audio, and other modalities, can improve monitoring performance, it also increases hardware, computational, and operational costs due to the use of multiple sensor types. This paper introduces a cross-modality knowledge transfer (CMKT) methodology for LAM in-situ monitoring, which transfers knowledge from a source modality to a target modality. CMKT enhances the representativeness of the features extracted from the target modality, allowing the removal of source modality sensors during prediction. This paper proposes three CMKT methods: semantic alignment, fully supervised mapping, and semi-supervised mapping. The semantic alignment method establishes a shared encoded space between modalities to facilitate knowledge transfer. It employs a semantic alignment loss to align the distributions of identical groups (e.g., visual and audio defective groups) and a separation loss to distinguish different groups (e.g., visual defective and audio defect-free groups). The two mapping methods transfer knowledge by deriving features from one modality to another using fully supervised and semi-supervised learning approaches. In a case study for LAM in-situ defect detection, the proposed CMKT methods were compared with multimodal audio-visual fusion. The semantic alignment method achieved an accuracy of 98.7% while removing the audio modality during the prediction phase, which is comparable to the 98.2% accuracy obtained through multimodal fusion. Using explainable artificial intelligence, we discovered that semantic alignment CMKT can extract more representative features while reducing noise by leveraging the inherent correlations between modalities.
comment: 45 pages, 17 figures, 6 tables
♻ ☆ The Dark Side of Rich Rewards: Understanding and Mitigating Noise in VLM Rewards
While Vision-Language Models (VLMs) are increasingly used to generate reward signals for training embodied agents to follow instructions, our research reveals that agents guided by VLM rewards often underperform compared to those employing only intrinsic (exploration-driven) rewards, contradicting expectations set by recent work. We hypothesize that false positive rewards -- instances where unintended trajectories are incorrectly rewarded -- are more detrimental than false negatives. Our analysis confirms this hypothesis, revealing that the widely used cosine similarity metric is prone to false positive reward estimates. To address this, we introduce BiMI ({Bi}nary {M}utual {I}nformation), a novel reward function designed to mitigate noise. BiMI significantly enhances learning efficiency across diverse and challenging embodied navigation environments. Our findings offer a nuanced understanding of how different types of reward noise impact agent learning and highlight the importance of addressing multimodal reward signal noise when training embodied agents
comment: 10 main body pages, 11 appendix pages
♻ ☆ BrainTransformers: SNN-LLM
This study introduces BrainTransformers, an innovative Large Language Model (LLM) implemented using Spiking Neural Networks (SNN). Our key contributions include: (1) designing SNN-compatible Transformer components such as SNNMatmul, SNNSoftmax, and SNNSiLU; (2) implementing an SNN approximation of the SiLU activation function; and (3) developing a Synapsis module to simulate synaptic plasticity. Our 3-billion parameter model, BrainTransformers-3B-Chat, demonstrates competitive performance across various benchmarks, including MMLU (63.2), BBH (54.1), ARC-C (54.3), and GSM8K (76.3), while potentially offering improved energy efficiency and biological plausibility. The model employs a three-stage training approach, including SNN-specific neuronal synaptic plasticity training. This research opens new avenues for brain-like AI systems in natural language processing and neuromorphic computing. Future work will focus on hardware optimization, developing specialized SNN fine-tuning tools, and exploring practical applications in energy-efficient computing environments.
♻ ☆ TSDS: Data Selection for Task-Specific Model Finetuning
Finetuning foundation models for specific tasks is an emerging paradigm in modern machine learning. The efficacy of task-specific finetuning largely depends on the selection of appropriate training data. We present TSDS (Task-Specific Data Selection), a framework to select data for task-specific model finetuning, guided by a small but representative set of examples from the target task. To do so, we formulate data selection for task-specific finetuning as an optimization problem with a distribution alignment loss based on optimal transport to capture the discrepancy between the selected data and the target distribution. In addition, we add a regularizer to encourage the diversity of the selected data and incorporate kernel density estimation into the regularizer to reduce the negative effects of near-duplicates among the candidate data. We connect our optimization problem to nearest neighbor search and design efficient algorithms to compute the optimal solution based on approximate nearest neighbor search techniques. We evaluate our method on data selection for both continued pretraining and instruction tuning of language models. We show that instruction tuning using data selected by our method with a 1% selection ratio often outperforms using the full dataset and beats the baseline selection methods by 1.5 points in F1 score on average.
comment: 31 pages, 1 figure
♻ ☆ Comparing Quantum Encoding Techniques
As quantum computers continue to become more capable, the possibilities of their applications increase. For example, quantum techniques are being integrated with classical neural networks to perform machine learning. In order to be used in this way, or for any other widespread use like quantum chemistry simulations or cryptographic applications, classical data must be converted into quantum states through quantum encoding. There are three fundamental encoding methods: basis, amplitude, and rotation, as well as several proposed combinations. This study explores the encoding methods, specifically in the context of hybrid quantum-classical machine learning. Using the QuClassi quantum neural network architecture to perform binary classification of the `3' and `6' digits from the MNIST datasets, this study obtains several metrics such as accuracy, entropy, loss, and resistance to noise, while considering resource usage and computational complexity to compare the three main encoding methods.
♻ ☆ No more hard prompts: SoftSRV prompting for synthetic data generation
We present a novel soft prompt based framework, SoftSRV, that leverages a frozen pre-trained large language model (LLM) to generate targeted synthetic text sequences. Given a sample from the target distribution, our proposed framework uses data-driven loss minimization to train a parameterized "contextual" soft prompt. This soft prompt is then used to steer the frozen LLM to generate synthetic sequences that are similar to the target distribution. We argue that SoftSRV provides a practical improvement over common hard-prompting approaches that rely on human-curated prompt-templates, which can be idiosyncratic, labor-intensive to craft, and may need to be specialized per domain. We empirically evaluate SoftSRV and hard-prompting baselines by generating synthetic data to fine-tune a small Gemma model on three different domains (coding, math, reasoning). To stress the generality of SoftSRV, we perform these evaluations without any particular specialization of the framework to each domain. We find that SoftSRV significantly improves upon hard-prompting baselines, generating data with superior fine-tuning performance and that better matches the target distribution according to the MAUVE similarity metric.
♻ ☆ RegExplainer: Generating Explanations for Graph Neural Networks in Regression Task NeurIPS 2024
Graph regression is a fundamental task and has received increasing attention in a wide range of graph learning tasks. However, the inference process is often not interpretable. Most existing explanation techniques are limited to understanding GNN behaviors in classification tasks. In this work, we seek an explanation to interpret the graph regression models (XAIG-R). We show that existing methods overlook the distribution shifting and continuously ordered decision boundary, which hinders them away from being applied in the regression tasks. To address these challenges, we propose a novel objective based on the information bottleneck theory and introduce a new mix-up framework, which could support various GNNs in a model-agnostic manner. We further present a contrastive learning strategy to tackle the continuously ordered labels in regression task. To empirically verify the effectiveness of the proposed method, we introduce three benchmark datasets and a real-life dataset for evaluation. Extensive experiments show the effectiveness of the proposed method in interpreting GNN models in regression tasks.
comment: Accepted by NeurIPS 2024
♻ ☆ RoPINN: Region Optimized Physics-Informed Neural Networks
Physics-informed neural networks (PINNs) have been widely applied to solve partial differential equations (PDEs) by enforcing outputs and gradients of deep models to satisfy target equations. Due to the limitation of numerical computation, PINNs are conventionally optimized on finite selected points. However, since PDEs are usually defined on continuous domains, solely optimizing models on scattered points may be insufficient to obtain an accurate solution for the whole domain. To mitigate this inherent deficiency of the default scatter-point optimization, this paper proposes and theoretically studies a new training paradigm as region optimization. Concretely, we propose to extend the optimization process of PINNs from isolated points to their continuous neighborhood regions, which can theoretically decrease the generalization error, especially for hidden high-order constraints of PDEs. A practical training algorithm, Region Optimized PINN (RoPINN), is seamlessly derived from this new paradigm, which is implemented by a straightforward but effective Monte Carlo sampling method. By calibrating the sampling process into trust regions, RoPINN finely balances optimization and generalization error. Experimentally, RoPINN consistently boosts the performance of diverse PINNs on a wide range of PDEs without extra backpropagation or gradient calculation. Code is available at this repository: https://github.com/thuml/RoPINN.
♻ ☆ Do LLMs Have Political Correctness? Analyzing Ethical Biases and Jailbreak Vulnerabilities in AI Systems
Although large language models (LLMs) demonstrate impressive proficiency in various tasks, they present potential safety risks, such as `jailbreaks', where malicious inputs can coerce LLMs into generating harmful content. To address these issues, many LLM developers have implemented various safety measures to align these models. This alignment involves several techniques, including data filtering during pre-training, supervised fine-tuning, reinforcement learning from human feedback, and red-teaming exercises. These methods often introduce deliberate and intentional biases similar to Political Correctness (PC) to ensure the ethical behavior of LLMs. In this paper, we delve into the intentional biases injected into LLMs for safety purposes and examine methods to circumvent these safety alignment techniques. Notably, these intentional biases result in a jailbreaking success rate in GPT-4o models that differs by 20% between non-binary and cisgender keywords and by 16% between white and black keywords, even when the other parts of the prompts are identical. We introduce the concept of PCJailbreak, highlighting the inherent risks posed by these safety-induced biases. Additionally, we propose an efficient defense method PCDefense, which prevents jailbreak attempts by injecting defense prompts prior to generation. PCDefense stands as an appealing alternative to Guard Models, such as Llama-Guard, that require additional inference cost after text generation. Our findings emphasize the urgent need for LLM developers to adopt a more responsible approach when designing and implementing safety measures.
♻ ☆ Online Differentially Private Synthetic Data Generation
We present a polynomial-time algorithm for online differentially private synthetic data generation. For a data stream within the hypercube $[0,1]^d$ and an infinite time horizon, we develop an online algorithm that generates a differentially private synthetic dataset at each time $t$. This algorithm achieves a near-optimal accuracy bound of $O(\log(t)t^{-1/d})$ for $d\geq 2$ and $O(\log^{4.5}(t)t^{-1})$ for $d=1$ in the 1-Wasserstein distance. This result extends the previous work on the continual release model for counting queries to Lipschitz queries. Compared to the offline case, where the entire dataset is available at once, our approach requires only an extra polylog factor in the accuracy bound.
comment: 21 pages
♻ ☆ Context-Aware Assistant Selection for Improved Inference Acceleration with Large Language Models EMNLP
Despite their widespread adoption, large language models (LLMs) remain prohibitive to use under resource constraints, with their ever growing sizes only increasing the barrier for use. One noted issue is the high latency associated with auto-regressive generation, rendering large LLMs use dependent on advanced computing infrastructure. Assisted decoding, where a smaller draft model guides a larger target model's generation, has helped alleviate this, but remains dependent on alignment between the two models. Thus if the draft model is insufficiently capable on some domain relative to the target model, performance can degrade. Alternatively, one can leverage multiple draft models to better cover the expertise of the target, but when multiple black-box draft models are available, selecting an assistant without details about its construction can be difficult. To better understand this decision making problem, we observe it as a contextual bandit, where a policy must choose a draft model based on a context. We show that even without prior knowledge of the draft models, creating an offline dataset from only outputs of independent draft/target models and training a policy over the alignment of these outputs can accelerate performance on multiple domains provided the candidates are effective. Further results show this to hold on various settings with multiple assisted decoding candidates, highlighting its flexibility and the advantageous role that such decision making can play.
comment: 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP); 14 pages (9 pages main content + references + appendix)
♻ ☆ Regularized Q-learning NeurIPS2024
Q-learning is widely used algorithm in reinforcement learning community. Under the lookup table setting, its convergence is well established. However, its behavior is known to be unstable with the linear function approximation case. This paper develops a new Q-learning algorithm that converges when linear function approximation is used. We prove that simply adding an appropriate regularization term ensures convergence of the algorithm. We prove its stability using a recent analysis tool based on switching system models. Moreover, we experimentally show that it converges in environments where Q-learning with linear function approximation has known to diverge. We also provide an error bound on the solution where the algorithm converges.
comment: NeurIPS2024
♻ ☆ Hybrid Spatial Representations for Species Distribution Modeling SDM
We address an important problem in ecology called Species Distribution Modeling (SDM), whose goal is to predict whether a species exists at a certain position on Earth. In particular, we tackle a challenging version of this task, where we learn from presence-only data in a community-sourced dataset, model a large number of species simultaneously, and do not use any additional environmental information. Previous work has used neural implicit representations to construct models that achieve promising results. However, implicit representations often generate predictions of limited spatial precision. We attribute this limitation to their inherently global formulation and inability to effectively capture local feature variations. This issue is especially pronounced with presence-only data and a large number of species. To address this, we propose a hybrid embedding scheme that combines both implicit and explicit embeddings. Specifically, the explicit embedding is implemented with a multiresolution hashgrid, enabling our models to better capture local information. Experiments demonstrate that our results exceed other works by a large margin on various standard benchmarks, and that the hybrid representation is better than both purely implicit and explicit ones. Qualitative visualizations and comprehensive ablation studies reveal that our hybrid representation successfully addresses the two main challenges. Our code is open-sourced at https://github.com/Shiran-Yuan/HSR-SDM.
comment: Project codebase https://github.com/Shiran-Yuan/HSR-SDM
♻ ☆ Harnessing Your DRAM and SSD for Sustainable and Accessible LLM Inference with Mixed-Precision and Multi-level Caching
Although Large Language Models (LLMs) have demonstrated remarkable capabilities, their massive parameter counts and associated extensive computing make LLMs' deployment the main part of carbon emission from nowadays AI applications. Compared to modern GPUs like H$100$, it would be significantly carbon-sustainable if we could leverage old-fashioned GPUs such as M$40$ (as shown in Figure 1, M$40$ only has one third carbon emission of H$100$'s) for LLM servings. However, the limited High Bandwidth Memory (HBM) available on such GPU often cannot support the loading of LLMs due to the gigantic model size and intermediate activation data, making their serving challenging. For instance, a LLaMA2 model with $70$B parameters typically requires $128$GB for inference, which substantially surpasses $24$GB HBM in a $3090$ GPU and remains infeasible even considering the additional $64$GB DRAM. To address this challenge, this paper proposes a mixed-precision with a model modularization algorithm to enable LLM inference on outdated hardware with resource constraints. (The precision denotes the numerical precision like FP16, INT8, INT4) and multi-level caching (M2Cache).) Specifically, our M2Cache first modulizes neurons in LLM and creates their importance ranking. Then, it adopts a dynamic sparse mixed-precision quantization mechanism in weight space to reduce computational demands and communication overhead at each decoding step. It collectively lowers the operational carbon emissions associated with LLM inference. Moreover, M2Cache introduces a three-level cache management system with HBM, DRAM, and SSDs that complements the dynamic sparse mixed-precision inference. To enhance communication efficiency, M2Cache maintains a neuron-level mixed-precision LRU cache in HBM, a larger layer-aware cache in DRAM, and a full model in SSD.
comment: 24 pages, 13 figures
♻ ☆ ODBAE: a high-performance model identifying complex phenotypes in high-dimensional biological datasets
Identifying complex phenotypes from high-dimensional biological data is challenging due to the intricate interdependencies among different physiological indicators. Traditional approaches often focus on detecting outliers in single variables, overlooking the broader network of interactions that contribute to phenotype emergence. Here, we introduce ODBAE (Outlier Detection using Balanced Autoencoders), a machine learning method designed to uncover both subtle and extreme outliers by capturing latent relationships among multiple physiological parameters. ODBAE's revised loss function enhances its ability to detect two key types of outliers: influential points (IP), which disrupt latent correlations between dimensions, and high leverage points (HLP), which deviate from the norm but go undetected by traditional autoencoder-based methods. Using data from the International Mouse Phenotyping Consortium (IMPC), we show that ODBAE can identify knockout mice with complex, multi-indicator phenotypes - normal in individual traits, but abnormal when considered together. In addition, this method reveals novel metabolism-related genes and uncovers coordinated abnormalities across metabolic indicators. Our results highlight the utility of ODBAE in detecting joint abnormalities and advancing our understanding of homeostatic perturbations in biological systems.
♻ ☆ On Catastrophic Inheritance of Large Foundation Models
Large foundation models (LFMs) are claiming incredible performances. Yet great concerns have been raised about their mythic and uninterpreted potentials not only in machine learning, but also in various other disciplines. In this position paper, we propose to identify a neglected issue deeply rooted in LFMs: Catastrophic Inheritance, describing the weaknesses and limitations inherited from biased large-scale pre-training data to behaviors of LFMs on the downstream tasks, including samples that are corrupted, long-tailed, noisy, out-of-distributed, to name a few. Such inheritance can potentially cause catastrophes to downstream applications, such as bias, lack of generalization, deteriorated performance, security vulnerability, privacy leakage, and value misalignment. We discuss the challenges behind this issue and propose UIM, a framework to Understand the catastrophic inheritance of LFMs from both pre-training and downstream adaptation, Interpret the implications of catastrophic inheritance on downstream tasks, and how to Mitigate it. UIM aims to unite both the machine learning and social sciences communities for more responsible and promising AI development and deployment.
comment: Accepted by DMLR
Multimedia 7
☆ Challenge on Sound Scene Synthesis: Evaluating Text-to-Audio Generation NeurIPS 2024
Despite significant advancements in neural text-to-audio generation, challenges persist in controllability and evaluation. This paper addresses these issues through the Sound Scene Synthesis challenge held as part of the Detection and Classification of Acoustic Scenes and Events 2024. We present an evaluation protocol combining objective metric, namely Fr\'echet Audio Distance, with perceptual assessments, utilizing a structured prompt format to enable diverse captions and effective evaluation. Our analysis reveals varying performance across sound categories and model architectures, with larger models generally excelling but innovative lightweight approaches also showing promise. The strong correlation between objective metrics and human ratings validates our evaluation approach. We discuss outcomes in terms of audio quality, controllability, and architectural considerations for text-to-audio synthesizers, providing direction for future research.
comment: accepted to NeurIPS 2024 Workshop: Audio Imagination
♻ ☆ LocoMotion: Learning Motion-Focused Video-Language Representations ACCV 2024
This paper strives for motion-focused video-language representations. Existing methods to learn video-language representations use spatial-focused data, where identifying the objects and scene is often enough to distinguish the relevant caption. We instead propose LocoMotion to learn from motion-focused captions that describe the movement and temporal progression of local object motions. We achieve this by adding synthetic motions to videos and using the parameters of these motions to generate corresponding captions. Furthermore, we propose verb-variation paraphrasing to increase the caption variety and learn the link between primitive motions and high-level verbs. With this, we are able to learn a motion-focused video-language representation. Experiments demonstrate our approach is effective for a variety of downstream tasks, particularly when limited data is available for fine-tuning. Code is available: https://hazeldoughty.github.io/Papers/LocoMotion/
comment: ACCV 2024 Oral
♻ ☆ Multi-Track MusicLDM: Towards Versatile Music Generation with Latent Diffusion Model
Diffusion models have shown promising results in cross-modal generation tasks involving audio and music, such as text-to-sound and text-to-music generation. These text-controlled music generation models typically focus on generating music by capturing global musical attributes like genre and mood. However, music composition is a complex, multilayered task that often involves musical arrangement as an integral part of the process. This process involves composing each instrument to align with existing ones in terms of beat, dynamics, harmony, and melody, requiring greater precision and control over tracks than text prompts usually provide. In this work, we address these challenges by extending the MusicLDM, a latent diffusion model for music, into a multi-track generative model. By learning the joint probability of tracks sharing a context, our model is capable of generating music across several tracks that correspond well to each other, either conditionally or unconditionally. Additionally, our model is capable of arrangement generation, where the model can generate any subset of tracks given the others (e.g., generating a piano track complementing given bass and drum tracks). We compared our model with an existing multi-track generative model and demonstrated that our model achieves considerable improvements across objective metrics for both total and arrangement generation tasks.
♻ ☆ Dual Inverse Degradation Network for Real-World SDRTV-to-HDRTV Conversion
In this study, we address the emerging necessity of converting Standard Dynamic Range Television (SDRTV) content into High Dynamic Range Television (HDRTV) in light of the limited number of native HDRTV content. A principal technical challenge in this conversion is the exacerbation of coding artifacts inherent in SDRTV, which detrimentally impacts the quality of the resulting HDRTV. To address this issue, our method introduces a novel approach that conceptualizes the SDRTV-to-HDRTV conversion as a composite task involving dual degradation restoration. This encompasses inverse tone mapping in conjunction with video restoration. We propose Dual Inversion Downgraded SDRTV to HDRTV Network (DIDNet), which can accurately perform inverse tone mapping while preventing encoding artifacts from being amplified, thereby significantly improving visual quality. DIDNet integrates an intermediate auxiliary loss function to effectively separate the dual degradation restoration tasks and efficient learning of both artifact reduction and inverse tone mapping during end-to-end training. Additionally, DIDNet introduces a spatio-temporal feature alignment module for video frame fusion, which augments texture quality and reduces artifacts. The architecture further includes a dual-modulation convolution mechanism for optimized inverse tone mapping. Recognizing the richer texture and high-frequency information in HDRTV compared to SDRTV, we further introduce a wavelet attention module to enhance frequency features. Our approach demonstrates marked superiority over existing state-of-the-art techniques in terms of quantitative performance and visual quality.
♻ ☆ OpenMU: Your Swiss Army Knife for Music Understanding
We present OpenMU-Bench, a large-scale benchmark suite for addressing the data scarcity issue in training multimodal language models to understand music. To construct OpenMU-Bench, we leveraged existing datasets and bootstrapped new annotations. OpenMU-Bench also broadens the scope of music understanding by including lyrics understanding and music tool usage. Using OpenMU-Bench, we trained our music understanding model, OpenMU, with extensive ablations, demonstrating that OpenMU outperforms baseline models such as MU-Llama. Both OpenMU and OpenMU-Bench are open-sourced to facilitate future research in music understanding and to enhance creative music production efficiency.
comment: Resources: https://github.com/mzhaojp22/openmu
♻ ☆ Exploring Self-Supervised Skeleton-Based Human Action Recognition under Occlusions
To integrate self-supervised skeleton-based action recognition methods into autonomous robotic systems, it is crucial to consider adverse situations involving target occlusions. Such a scenario, despite its practical relevance, is rarely addressed in existing self-supervised skeleton-based action recognition methods. To empower models with the capacity to address occlusion, we propose a simple and effective method. We first pre-train using occluded skeleton sequences, then use k-means clustering (KMeans) on sequence embeddings to group semantically similar samples. Next, we propose KNN-Imputation to fill in missing skeleton data based on the closest sample neighbors. Imputing incomplete skeleton sequences to create relatively complete sequences as input provides significant benefits to existing skeleton-based self-supervised methods. Meanwhile, building on the state-of-the-art Partial Spatio-Temporal Learning (PSTL), we introduce an Occluded Partial Spatio-Temporal Learning (OPSTL) framework. This enhancement utilizes Adaptive Spatial Masking (ASM) for better use of high-quality, intact skeletons. The new proposed method is verified on the challenging occluded versions of the NTURGB+D 60 and NTURGB+D 120. The source code is publicly available at https://github.com/cyfml/OPSTL.
comment: The source code is publicly available at https://github.com/cyfml/OPSTL
♻ ☆ MMBench-Video: A Long-Form Multi-Shot Benchmark for Holistic Video Understanding NeurIPS 2024
The advent of large vision-language models (LVLMs) has spurred research into their applications in multi-modal contexts, particularly in video understanding. Traditional VideoQA benchmarks, despite providing quantitative metrics, often fail to encompass the full spectrum of video content and inadequately assess models' temporal comprehension. To address these limitations, we introduce MMBench-Video, a quantitative benchmark designed to rigorously evaluate LVLMs' proficiency in video understanding. MMBench-Video incorporates lengthy videos from YouTube and employs free-form questions, mirroring practical use cases. The benchmark is meticulously crafted to probe the models' temporal reasoning skills, with all questions human-annotated according to a carefully constructed ability taxonomy. We employ GPT-4 for automated assessment, demonstrating superior accuracy and robustness over earlier LLM-based evaluations. Utilizing MMBench-Video, we have conducted comprehensive evaluations that include both proprietary and open-source LVLMs for images and videos. MMBench-Video stands as a valuable resource for the research community, facilitating improved evaluation of LVLMs and catalyzing progress in the field of video understanding. The evalutation code of MMBench-Video will be integrated into VLMEvalKit: https://github.com/open-compass/VLMEvalKit.
comment: Accepted in NeurIPS 2024 Datasets and Benchmarks Track
Computation and Language 145
☆ Altogether: Image Captioning via Re-aligning Alt-text EMNLP 2024
This paper focuses on creating synthetic data to improve the quality of image captions. Existing works typically have two shortcomings. First, they caption images from scratch, ignoring existing alt-text metadata, and second, lack transparency if the captioners' training data (e.g. GPT) is unknown. In this paper, we study a principled approach Altogether based on the key idea to edit and re-align existing alt-texts associated with the images. To generate training data, we perform human annotation where annotators start with the existing alt-text and re-align it to the image content in multiple rounds, consequently constructing captions with rich visual concepts. This differs from prior work that carries out human annotation as a one-time description task solely based on images and annotator knowledge. We train a captioner on this data that generalizes the process of re-aligning alt-texts at scale. Our results show our Altogether approach leads to richer image captions that also improve text-to-image generation and zero-shot image classification tasks.
comment: accepted by EMNLP 2024; MetaCLIPv2
☆ JMMMU: A Japanese Massive Multi-discipline Multimodal Understanding Benchmark for Culture-aware Evaluation
Accelerating research on Large Multimodal Models (LMMs) in non-English languages is crucial for enhancing user experiences across broader populations. In this paper, we introduce JMMMU (Japanese MMMU), the first large-scale Japanese benchmark designed to evaluate LMMs on expert-level tasks based on the Japanese cultural context. To facilitate comprehensive culture-aware evaluation, JMMMU features two complementary subsets: (i) culture-agnostic (CA) subset, where the culture-independent subjects (e.g., Math) are selected and translated into Japanese, enabling one-to-one comparison with its English counterpart MMMU; and (ii) culture-specific (CS) subset, comprising newly crafted subjects that reflect Japanese cultural context. Using the CA subset, we observe performance drop in many LMMs when evaluated in Japanese, which is purely attributable to language variation. Using the CS subset, we reveal their inadequate Japanese cultural understanding. Further, by combining both subsets, we identify that some LMMs perform well on the CA subset but not on the CS subset, exposing a shallow understanding of the Japanese language that lacks depth in cultural understanding. We hope this work will not only help advance LMM performance in Japanese but also serve as a guideline to create high-standard, culturally diverse benchmarks for multilingual LMM development. The project page is https://mmmu-japanese-benchmark.github.io/JMMMU/.
comment: Project page: https://mmmu-japanese-benchmark.github.io/JMMMU/
☆ PyramidDrop: Accelerating Your Large Vision-Language Models via Pyramid Visual Redundancy Reduction
In large vision-language models (LVLMs), images serve as inputs that carry a wealth of information. As the idiom "A picture is worth a thousand words" implies, representing a single image in current LVLMs can require hundreds or even thousands of tokens. This results in significant computational costs, which grow quadratically as input image resolution increases, thereby severely impacting the efficiency of both training and inference. Previous approaches have attempted to reduce the number of image tokens either before or within the early layers of LVLMs. However, these strategies inevitably result in the loss of crucial image information, ultimately diminishing model performance. To address this challenge, we conduct an empirical study revealing that all visual tokens are necessary for LVLMs in the shallow layers, and token redundancy progressively increases in the deeper layers of the model. To this end, we propose PyramidDrop, a visual redundancy reduction strategy for LVLMs to boost their efficiency in both training and inference with neglectable performance loss. Specifically, we partition the LVLM into several stages and drop part of the image tokens at the end of each stage with a pre-defined ratio, creating pyramid-like visual tokens across model layers. The dropping is based on a lightweight similarity calculation with a negligible time overhead. Extensive experiments demonstrate that PyramidDrop can achieve a 40% training time and 55% inference FLOPs acceleration of LLaVA-NeXT with comparable performance. Besides, the PyramidDrop could also serve as a plug-and-play strategy for inference acceleration without training, with better performance and lower inference cost than counterparts. We hope that the insights and approach introduced by PyramidDrop will inspire future research to further investigate the role of image tokens in LVLMs.
comment: 10 pages
☆ Towards Reliable Evaluation of Behavior Steering Interventions in LLMs NeurIPS 2024
Representation engineering methods have recently shown promise for enabling efficient steering of model behavior. However, evaluation pipelines for these methods have primarily relied on subjective demonstrations, instead of quantitative, objective metrics. We aim to take a step towards addressing this issue by advocating for four properties missing from current evaluations: (i) contexts sufficiently similar to downstream tasks should be used for assessing intervention quality; (ii) model likelihoods should be accounted for; (iii) evaluations should allow for standardized comparisons across different target behaviors; and (iv) baseline comparisons should be offered. We introduce an evaluation pipeline grounded in these criteria, offering both a quantitative and visual analysis of how effectively a given method works. We use this pipeline to evaluate two representation engineering methods on how effectively they can steer behaviors such as truthfulness and corrigibility, finding that some interventions are less effective than previously reported.
comment: Accepted to the NeurIPS 2024 - Workshop on Foundation Model Interventions
☆ SELA: Tree-Search Enhanced LLM Agents for Automated Machine Learning
Automated Machine Learning (AutoML) approaches encompass traditional methods that optimize fixed pipelines for model selection and ensembling, as well as newer LLM-based frameworks that autonomously build pipelines. While LLM-based agents have shown promise in automating machine learning tasks, they often generate low-diversity and suboptimal code, even after multiple iterations. To overcome these limitations, we introduce Tree-Search Enhanced LLM Agents (SELA), an innovative agent-based system that leverages Monte Carlo Tree Search (MCTS) to optimize the AutoML process. By representing pipeline configurations as trees, our framework enables agents to conduct experiments intelligently and iteratively refine their strategies, facilitating a more effective exploration of the machine learning solution space. This novel approach allows SELA to discover optimal pathways based on experimental feedback, improving the overall quality of the solutions. In an extensive evaluation across 20 machine learning datasets, we compare the performance of traditional and agent-based AutoML methods, demonstrating that SELA achieves a win rate of 65% to 80% against each baseline across all datasets. These results underscore the significant potential of agent-based strategies in AutoML, offering a fresh perspective on tackling complex machine learning challenges.
comment: The code is available at https://github.com/geekan/MetaGPT
☆ Large Language Models Empowered Personalized Web Agents
Web agents have emerged as a promising direction to automate Web task completion based on user instructions, significantly enhancing user experience. Recently, Web agents have evolved from traditional agents to Large Language Models (LLMs)-based Web agents. Despite their success, existing LLM-based Web agents overlook the importance of personalized data (e.g., user profiles and historical Web behaviors) in assisting the understanding of users' personalized instructions and executing customized actions. To overcome the limitation, we first formulate the task of LLM-empowered personalized Web agents, which integrate personalized data and user instructions to personalize instruction comprehension and action execution. To address the absence of a comprehensive evaluation benchmark, we construct a Personalized Web Agent Benchmark (PersonalWAB), featuring user instructions, personalized user data, Web functions, and two evaluation paradigms across three personalized Web tasks. Moreover, we propose a Personalized User Memory-enhanced Alignment (PUMA) framework to adapt LLMs to the personalized Web agent task. PUMA utilizes a memory bank with a task-specific retrieval strategy to filter relevant historical Web behaviors. Based on the behaviors, PUMA then aligns LLMs for personalized action execution through fine-tuning and direct preference optimization. Extensive experiments validate the superiority of PUMA over existing Web agents on PersonalWAB.
comment: The code and data are available on the project website https://hongrucai.github.io/PersonalWAB/
☆ Automated Spinal MRI Labelling from Reports Using a Large Language Model MICCAI 2024
We propose a general pipeline to automate the extraction of labels from radiology reports using large language models, which we validate on spinal MRI reports. The efficacy of our labelling method is measured on five distinct conditions: spinal cancer, stenosis, spondylolisthesis, cauda equina compression and herniation. Using open-source models, our method equals or surpasses GPT-4 on a held-out set of reports. Furthermore, we show that the extracted labels can be used to train imaging models to classify the identified conditions in the accompanying MR scans. All classifiers trained using automated labels achieve comparable performance to models trained using scans manually annotated by clinicians. Code can be found at https://github.com/robinyjpark/AutoLabelClassifier.
comment: Accepted to Medical Image Computing and Computer Assisted Intervention (MICCAI 2024, Spotlight). 11 pages plus appendix
☆ Fine-Tuning Large Language Models to Appropriately Abstain with Semantic Entropy NeurIPS
Large Language Models (LLMs) are known to hallucinate, whereby they generate plausible but inaccurate text. This phenomenon poses significant risks in critical applications, such as medicine or law, necessitating robust hallucination mitigation strategies. While recent works have proposed fine-tuning methods to teach LLMs to abstain from answering questions beyond their knowledge or capabilities, these methods rely on the existence of ground-truth labels or are limited to short-form responses. To address these limitations, we propose fine-tuning using semantic entropy, an uncertainty measure derived from introspection into the model which does not require external labels. We demonstrate that our approach matches or outperforms models fine-tuned using prior work and achieves strong performance for both short and long-form generations on a range of datasets.
comment: Accepted to NeurIPS Safe Generative AI Workshop 2024
☆ Dhoroni: Exploring Bengali Climate Change and Environmental Views with a Multi-Perspective News Dataset and Natural Language Processing
Climate change poses critical challenges globally, disproportionately affecting low-income countries that often lack resources and linguistic representation on the international stage. Despite Bangladesh's status as one of the most vulnerable nations to climate impacts, research gaps persist in Bengali-language studies related to climate change and NLP. To address this disparity, we introduce Dhoroni, a novel Bengali (Bangla) climate change and environmental news dataset, comprising a 2300 annotated Bangla news articles, offering multiple perspectives such as political influence, scientific/statistical data, authenticity, stance detection, and stakeholder involvement. Furthermore, we present an in-depth exploratory analysis of Dhoroni and introduce BanglaBERT-Dhoroni family, a novel baseline model family for climate and environmental opinion detection in Bangla, fine-tuned on our dataset. This research contributes significantly to enhancing accessibility and analysis of climate discourse in Bengali (Bangla), addressing crucial communication and research gaps in climate-impacted regions like Bangladesh with 180 million people.
comment: In Review
☆ Context-aware Prompt Tuning: Advancing In-Context Learning with Adversarial Methods
Fine-tuning Large Language Models (LLMs) typically involves updating at least a few billions of parameters. A more parameter-efficient approach is Prompt Tuning (PT), which updates only a few learnable tokens, and differently, In-Context Learning (ICL) adapts the model to a new task by simply including examples in the input without any training. When applying optimization-based methods, such as fine-tuning and PT for few-shot learning, the model is specifically adapted to the small set of training examples, whereas ICL leaves the model unchanged. This distinction makes traditional learning methods more prone to overfitting; in contrast, ICL is less sensitive to the few-shot scenario. While ICL is not prone to overfitting, it does not fully extract the information that exists in the training examples. This work introduces Context-aware Prompt Tuning (CPT), a method inspired by ICL, PT, and adversarial attacks. We build on the ICL strategy of concatenating examples before the input, but we extend this by PT-like learning, refining the context embedding through iterative optimization to extract deeper insights from the training examples. We carefully modify specific context tokens, considering the unique structure of input and output formats. Inspired by adversarial attacks, we adjust the input based on the labels present in the context, focusing on minimizing, rather than maximizing, the loss. Moreover, we apply a projected gradient descent algorithm to keep token embeddings close to their original values, under the assumption that the user-provided data is inherently valuable. Our method has been shown to achieve superior accuracy across multiple classification tasks using various LLM models.
☆ Creativity in AI: Progresses and Challenges
Creativity is the ability to produce novel, useful, and surprising ideas, and has been widely studied as a crucial aspect of human cognition. Machine creativity on the other hand has been a long-standing challenge. With the rise of advanced generative AI, there has been renewed interest and debate regarding AI's creative capabilities. Therefore, it is imperative to revisit the state of creativity in AI and identify key progresses and remaining challenges. In this work, we survey leading works studying the creative capabilities of AI systems, focusing on creative problem-solving, linguistic, artistic, and scientific creativity. Our review suggests that while the latest AI models are largely capable of producing linguistically and artistically creative outputs such as poems, images, and musical pieces, they struggle with tasks that require creative problem-solving, abstract thinking and compositionality and their generations suffer from a lack of diversity, originality, long-range incoherence and hallucinations. We also discuss key questions concerning copyright and authorship issues with generative models. Furthermore, we highlight the need for a comprehensive evaluation of creativity that is process-driven and considers several dimensions of creativity. Finally, we propose future research directions to improve the creativity of AI outputs, drawing inspiration from cognitive science and psychology.
comment: 44 pages
☆ MiniPLM: Knowledge Distillation for Pre-Training Language Models
Knowledge distillation (KD) is widely used to train small, high-performing student language models (LMs) using large teacher LMs. While effective in fine-tuning, KD during pre-training faces challenges in efficiency, flexibility, and effectiveness. Existing methods either incur high computational costs due to online teacher inference, require tokenization matching between teacher and student LMs, or risk losing the difficulty and diversity of the teacher-generated training data. To address these issues, we propose MiniPLM, a KD framework for pre-training LMs by refining the training data distribution with the teacher's knowledge. For efficiency, MiniPLM performs offline teacher LM inference, allowing KD for multiple student LMs without adding training-time costs. For flexibility, MiniPLM operates solely on the training corpus, enabling KD across model families. For effectiveness, MiniPLM leverages the differences between large and small LMs to enhance the difficulty and diversity of the training data, helping student LMs acquire versatile and sophisticated knowledge. Extensive experiments demonstrate that MiniPLM boosts the student LMs' performance on 9 widely used downstream tasks, improves the language modeling capabilities, and reduces pre-training computation. The benefit of MiniPLM extends to large pre-training scales, evidenced by the extrapolation of the scaling curves. Further analysis reveals that MiniPLM supports KD across model families and enhances the utilization of pre-training data. Our model, code, and data are available at https://github.com/thu-coai/MiniPLM.
☆ Exploring Possibilities of AI-Powered Legal Assistance in Bangladesh through Large Language Modeling
Purpose: Bangladesh's legal system struggles with major challenges like delays, complexity, high costs, and millions of unresolved cases, which deter many from pursuing legal action due to lack of knowledge or financial constraints. This research seeks to develop a specialized Large Language Model (LLM) to assist in the Bangladeshi legal system. Methods: We created UKIL-DB-EN, an English corpus of Bangladeshi legal documents, by collecting and scraping data on various legal acts. We fine-tuned the GPT-2 model on this dataset to develop GPT2-UKIL-EN, an LLM focused on providing legal assistance in English. Results: The model was rigorously evaluated using semantic assessments, including case studies supported by expert opinions. The evaluation provided promising results, demonstrating the potential for the model to assist in legal matters within Bangladesh. Conclusion: Our work represents the first structured effort toward building an AI-based legal assistant for Bangladesh. While the results are encouraging, further refinements are necessary to improve the model's accuracy, credibility, and safety. This is a significant step toward creating a legal AI capable of serving the needs of a population of 180 million.
comment: In Review
☆ Audio-to-Score Conversion Model Based on Whisper methodology
This thesis develops a Transformer model based on Whisper, which extracts melodies and chords from music audio and records them into ABC notation. A comprehensive data processing workflow is customized for ABC notation, including data cleansing, formatting, and conversion, and a mutation mechanism is implemented to increase the diversity and quality of training data. This thesis innovatively introduces the "Orpheus' Score", a custom notation system that converts music information into tokens, designs a custom vocabulary library, and trains a corresponding custom tokenizer. Experiments show that compared to traditional algorithms, the model has significantly improved accuracy and performance. While providing a convenient audio-to-score tool for music enthusiasts, this work also provides new ideas and tools for research in music information processing.
comment: 5 pages, 7 figures
☆ VoiceBench: Benchmarking LLM-Based Voice Assistants
Building on the success of large language models (LLMs), recent advancements such as GPT-4o have enabled real-time speech interactions through LLM-based voice assistants, offering a significantly improved user experience compared to traditional text-based interactions. However, the absence of benchmarks designed to evaluate these speech interaction capabilities has hindered progress of LLM-based voice assistants development. Current evaluations focus primarily on automatic speech recognition (ASR) or general knowledge evaluation with clean speeches, neglecting the more intricate, real-world scenarios that involve diverse speaker characteristics, environmental and content factors. To address this, we introduce VoiceBench, the first benchmark designed to provide a multi-faceted evaluation of LLM-based voice assistants. VoiceBench also includes both real and synthetic spoken instructions that incorporate the above three key real-world variations. Extensive experiments reveal the limitations of current LLM-based voice assistant models and offer valuable insights for future research and development in this field.
comment: Work in progress. Data is available at https://github.com/MatthewCYM/VoiceBench
☆ Language Model Non-myopic Generation for Reasoning and Planning
Large Language Models have demonstrated remarkable abilities in reasoning and planning by breaking down complex problems into sequential steps. Despite their success in various domains like mathematical problem-solving and coding, LLMs face challenges in ensuring reliable and optimal planning due to their inherent myopic nature of autoregressive decoding. This paper revisits LLM reasoning from an optimal-control perspective, proposing a novel method, Predictive-Decoding, that leverages Model Predictive Control to enhance planning accuracy. By re-weighting LLM distributions based on foresight trajectories, Predictive-Decoding aims to mitigate early errors and promote non-myopic planning. Our experiments show significant improvements in a wide range of tasks for math, coding, and agents. Furthermore, Predictive-Decoding demonstrates computational efficiency, outperforming search baselines with reduced computational resources. This study provides insights into optimizing LLM planning capabilities.
☆ From Attention to Activation: Unravelling the Enigmas of Large Language Models
We study two strange phenomena in auto-regressive Transformers: (1) the dominance of the first token in attention heads; (2) the occurrence of large outlier activations in the hidden states. We find that popular large language models, such as Llama attend maximally to the first token in 98% of attention heads, a behaviour we attribute to the softmax function. To mitigate this issue, we propose a reformulation of softmax to softmax-1. Furthermore, we identify adaptive optimisers, e.g. Adam, as the primary contributor to the large outlier activations and introduce OrthoAdam, a novel optimiser that utilises orthogonal matrices to transform gradients, to address this issue. Finally, not only do our methods prevent these phenomena from occurring, but additionally, they enable Transformers to sustain their performance when quantised using basic algorithms, something that standard methods are unable to do. In summary, our methods reduce the attention proportion on the first token from 65% to 3.3%, the activation kurtosis in the hidden states from 1657 to 3.1, and perplexity penalty under 4-bit weight quantisation from 3565 to 0.3.
comment: 10 pages
☆ Self-calibration for Language Model Quantization and Pruning
Quantization and pruning are fundamental approaches for model compression, enabling efficient inference for language models. In a post-training setting, state-of-the-art quantization and pruning methods require calibration data, a small set of unlabeled examples. Conventionally, randomly sampled web text is used, aiming to reflect the model training data. However, this poses two key problems: (1) unrepresentative calibration examples can harm model performance, and (2) organizations increasingly avoid releasing model training data. In this paper, we propose self-calibration as a solution. Our approach requires no external data, instead leveraging the model itself to generate synthetic calibration data as a better approximation of the pre-training data distribution. We extensively compare the performance of self-calibration with several baselines, across a variety of models, compression methods, and tasks. Our approach proves consistently competitive in maximizing downstream task performance, frequently outperforming even using real data.
comment: Work in progress
☆ Interchangeable Token Embeddings for Extendable Vocabulary and Alpha-Equivalence
We propose a novel approach for learning interchangeable tokens in language models to obtain an extendable vocabulary that can generalize to new tokens. Our method is designed to address alpha-equivalence, the principle that renaming bound variables in a syntactic expression preserves semantics. This property arises in many formal languages such as temporal logics, in which all proposition symbols represent the same concept but are distinguishable from each other. To handle such tokens, we develop a dual-part embedding approach. The first part is shared across all interchangeable tokens, thereby enforcing that they represent the same core concept. The second part is randomly generated for each token, which enables distinguishability. We evaluate our method in a Transformer encoder-decoder model on two tasks: solving linear temporal logic formulae and copying with extendable vocabulary. Our method demonstrates promising generalization capabilities in addition to introducing a favorable inductive bias for alpha-equivalence.
comment: 14 pages, 5 figures
☆ Improving Pinterest Search Relevance Using Large Language Models CIKM 2024
To improve relevance scoring on Pinterest Search, we integrate Large Language Models (LLMs) into our search relevance model, leveraging carefully designed text representations to predict the relevance of Pins effectively. Our approach uses search queries alongside content representations that include captions extracted from a generative visual language model. These are further enriched with link-based text data, historically high-quality engaged queries, user-curated boards, Pin titles and Pin descriptions, creating robust models for predicting search relevance. We use a semi-supervised learning approach to efficiently scale up the amount of training data, expanding beyond the expensive human labeled data available. By utilizing multilingual LLMs, our system extends training data to include unseen languages and domains, despite initial data and annotator expertise being confined to English. Furthermore, we distill from the LLM-based model into real-time servable model architectures and features. We provide comprehensive offline experimental validation for our proposed techniques and demonstrate the gains achieved through the final deployed system at scale.
comment: CIKM 2024 Workshop on Industrial Recommendation Systems
☆ Can General-Purpose Large Language Models Generalize to English-Thai Machine Translation ? EMNLP 2024
Large language models (LLMs) perform well on common tasks but struggle with generalization in low-resource and low-computation settings. We examine this limitation by testing various LLMs and specialized translation models on English-Thai machine translation and code-switching datasets. Our findings reveal that under more strict computational constraints, such as 4-bit quantization, LLMs fail to translate effectively. In contrast, specialized models, with comparable or lower computational requirements, consistently outperform LLMs. This underscores the importance of specialized models for maintaining performance under resource constraints.
comment: Accepted in GenBench EMNLP 2024
☆ Aligning Large Language Models via Self-Steering Optimization
Automated alignment develops alignment systems with minimal human intervention. The key to automated alignment lies in providing learnable and accurate preference signals for preference learning without human annotation. In this paper, we introduce Self-Steering Optimization ($SSO$), an algorithm that autonomously generates high-quality preference signals based on predefined principles during iterative training, eliminating the need for manual annotation. $SSO$ maintains the accuracy of signals by ensuring a consistent gap between chosen and rejected responses while keeping them both on-policy to suit the current policy model's learning capacity. $SSO$ can benefit the online and offline training of the policy model, as well as enhance the training of reward models. We validate the effectiveness of $SSO$ with two foundation models, Qwen2 and Llama3.1, indicating that it provides accurate, on-policy preference signals throughout iterative training. Without any manual annotation or external models, $SSO$ leads to significant performance improvements across six subjective or objective benchmarks. Besides, the preference data generated by $SSO$ significantly enhanced the performance of the reward model on Rewardbench. Our work presents a scalable approach to preference optimization, paving the way for more efficient and effective automated alignment.
☆ PAPILLON: PrivAcy Preservation from Internet-based and Local Language MOdel ENsembles
Users can divulge sensitive information to proprietary LLM providers, raising significant privacy concerns. While open-source models, hosted locally on the user's machine, alleviate some concerns, models that users can host locally are often less capable than proprietary frontier models. Toward preserving user privacy while retaining the best quality, we propose Privacy-Conscious Delegation, a novel task for chaining API-based and local models. We utilize recent public collections of user-LLM interactions to construct a natural benchmark called PUPA, which contains personally identifiable information (PII). To study potential approaches, we devise PAPILLON, a multi-stage LLM pipeline that uses prompt optimization to address a simpler version of our task. Our best pipeline maintains high response quality for 85.5% of user queries while restricting privacy leakage to only 7.5%. We still leave a large margin to the generation quality of proprietary LLMs for future work. Our data and code will be available at https://github.com/siyan-sylvia-li/PAPILLON.
☆ Exploring RL-based LLM Training for Formal Language Tasks with Programmed Rewards
Proximal Policy Optimization (PPO) is commonly used in Reinforcement Learning from Human Feedback to align large language models (LLMs) with downstream tasks. This paper investigates the feasibility of using PPO for direct reinforcement learning (RL) from explicitly programmed reward signals, as opposed to indirect learning from human feedback via an intermediary reward model. We focus on tasks expressed through formal languages, such as mathematics and programming, where explicit reward functions can be programmed to automatically assess the quality of generated outputs. We apply this approach to a sentiment alignment task, a simple arithmetic task, and a more complex game synthesis task. The sentiment alignment task replicates prior research and serves to validate our experimental setup. Our results show that pure RL-based training for the two formal language tasks is challenging, with success being limited even for the simple arithmetic task. We propose a novel batch-entropy regularization term to aid exploration, although training is not yet entirely stable. Our findings suggest that direct RL training of LLMs may be more suitable for relatively minor changes, such as alignment, than for learning new tasks altogether, even if an informative reward signal can be expressed programmatically.
comment: Accepted at BNAIC 2024
☆ Enhancing Answer Attribution for Faithful Text Generation with Large Language Models
The increasing popularity of Large Language Models (LLMs) in recent years has changed the way users interact with and pose questions to AI-based conversational systems. An essential aspect for increasing the trustworthiness of generated LLM answers is the ability to trace the individual claims from responses back to relevant sources that support them, the process known as answer attribution. While recent work has started exploring the task of answer attribution in LLMs, some challenges still remain. In this work, we first perform a case study analyzing the effectiveness of existing answer attribution methods, with a focus on subtasks of answer segmentation and evidence retrieval. Based on the observed shortcomings, we propose new methods for producing more independent and contextualized claims for better retrieval and attribution. The new methods are evaluated and shown to improve the performance of answer attribution components. We end with a discussion and outline of future directions for the task.
comment: Accepted to KDIR 2024 (part of IC3K 2024)
☆ Human-LLM Hybrid Text Answer Aggregation for Crowd Annotations EMNLP 2024
The quality is a crucial issue for crowd annotations. Answer aggregation is an important type of solution. The aggregated answers estimated from multiple crowd answers to the same instance are the eventually collected annotations, rather than the individual crowd answers themselves. Recently, the capability of Large Language Models (LLMs) on data annotation tasks has attracted interest from researchers. Most of the existing studies mainly focus on the average performance of individual crowd workers; several recent works studied the scenarios of aggregation on categorical labels and LLMs used as label creators. However, the scenario of aggregation on text answers and the role of LLMs as aggregators are not yet well-studied. In this paper, we investigate the capability of LLMs as aggregators in the scenario of close-ended crowd text answer aggregation. We propose a human-LLM hybrid text answer aggregation method with a Creator-Aggregator Multi-Stage (CAMS) crowdsourcing framework. We make the experiments based on public crowdsourcing datasets. The results show the effectiveness of our approach based on the collaboration of crowd workers and LLMs.
comment: Accepted in EMNLP 2024
☆ Science Out of Its Ivory Tower: Improving Accessibility with Reinforcement Learning
A vast amount of scholarly work is published daily, yet much of it remains inaccessible to the general public due to dense jargon and complex language. To address this challenge in science communication, we introduce a reinforcement learning framework that fine-tunes a language model to rewrite scholarly abstracts into more comprehensible versions. Guided by a carefully balanced combination of word- and sentence-level accessibility rewards, our language model effectively substitutes technical terms with more accessible alternatives, a task which models supervised fine-tuned or guided by conventional readability measures struggle to accomplish. Our best model adjusts the readability level of scholarly abstracts by approximately six U.S. grade levels -- in other words, from a postgraduate to a high school level. This translates to roughly a 90% relative boost over the supervised fine-tuning baseline, all while maintaining factual accuracy and high-quality language. An in-depth analysis of our approach shows that balanced rewards lead to systematic modifications in the base model, likely contributing to smoother optimization and superior performance. We envision this work as a step toward bridging the gap between scholarly research and the general public, particularly younger readers and those without a college degree.
☆ Continuous Speech Tokenizer in Text To Speech
The fusion of speech and language in the era of large language models has garnered significant attention. Discrete speech token is often utilized in text-to-speech tasks for speech compression and portability, which is convenient for joint training with text and have good compression efficiency. However, we found that the discrete speech tokenizer still suffers from information loss. Therefore, we propose a simple yet effective continuous speech tokenizer and a text-to-speech model based on continuous speech tokens. Our results show that the speech language model based on the continuous speech tokenizer has better continuity and higher estimated Mean Opinion Scores (MoS). This enhancement is attributed to better information preservation rate of the continuous speech tokenizer across both low and high frequencies in the frequency domain.
comment: 4 pages. Under review
☆ Data-driven Coreference-based Ontology Building
While coreference resolution is traditionally used as a component in individual document understanding, in this work we take a more global view and explore what can we learn about a domain from the set of all document-level coreference relations that are present in a large corpus. We derive coreference chains from a corpus of 30 million biomedical abstracts and construct a graph based on the string phrases within these chains, establishing connections between phrases if they co-occur within the same coreference chain. We then use the graph structure and the betweeness centrality measure to distinguish between edges denoting hierarchy, identity and noise, assign directionality to edges denoting hierarchy, and split nodes (strings) that correspond to multiple distinct concepts. The result is a rich, data-driven ontology over concepts in the biomedical domain, parts of which overlaps significantly with human-authored ontologies. We release the coreference chains and resulting ontology under a creative-commons license, along with the code.
☆ UnStar: Unlearning with Self-Taught Anti-Sample Reasoning for LLMs
The key components of machine learning are data samples for training, model for learning patterns, and loss function for optimizing accuracy. Analogously, unlearning can potentially be achieved through anti-data samples (or anti-samples), unlearning method, and reversed loss function. While prior research has explored unlearning methods and reversed loss functions, the potential of anti-samples remains largely untapped. In this paper, we introduce UnSTAR: Unlearning with Self-Taught Anti-Sample Reasoning for large language models (LLMs). Our contributions are threefold; first, we propose a novel concept of anti-sample-induced unlearning; second, we generate anti-samples by leveraging misleading rationales, which help reverse learned associations and accelerate the unlearning process; and third, we enable fine-grained targeted unlearning, allowing for the selective removal of specific associations without impacting related knowledge - something not achievable by previous works. Results demonstrate that anti-samples offer an efficient, targeted unlearning strategy for LLMs, opening new avenues for privacy-preserving machine learning and model modification.
☆ Arabic Dataset for LLM Safeguard Evaluation
The growing use of large language models (LLMs) has raised concerns regarding their safety. While many studies have focused on English, the safety of LLMs in Arabic, with its linguistic and cultural complexities, remains under-explored. Here, we aim to bridge this gap. In particular, we present an Arab-region-specific safety evaluation dataset consisting of 5,799 questions, including direct attacks, indirect attacks, and harmless requests with sensitive words, adapted to reflect the socio-cultural context of the Arab world. To uncover the impact of different stances in handling sensitive and controversial topics, we propose a dual-perspective evaluation framework. It assesses the LLM responses from both governmental and opposition viewpoints. Experiments over five leading Arabic-centric and multilingual LLMs reveal substantial disparities in their safety performance. This reinforces the need for culturally specific datasets to ensure the responsible deployment of LLMs.
comment: 17 pages, 6 figures, 10 tables
☆ DIRI: Adversarial Patient Reidentification with Large Language Models for Evaluating Clinical Text Anonymization
Sharing protected health information (PHI) is critical for furthering biomedical research. Before data can be distributed, practitioners often perform deidentification to remove any PHI contained in the text. Contemporary deidentification methods are evaluated on highly saturated datasets (tools achieve near-perfect accuracy) which may not reflect the full variability or complexity of real-world clinical text and annotating them is resource intensive, which is a barrier to real-world applications. To address this gap, we developed an adversarial approach using a large language model (LLM) to re-identify the patient corresponding to a redacted clinical note and evaluated the performance with a novel De-Identification/Re-Identification (DIRI) method. Our method uses a large language model to reidentify the patient corresponding to a redacted clinical note. We demonstrate our method on medical data from Weill Cornell Medicine anonymized with three deidentification tools: rule-based Philter and two deep-learning-based models, BiLSTM-CRF and ClinicalBERT. Although ClinicalBERT was the most effective, masking all identified PII, our tool still reidentified 9% of clinical notes Our study highlights significant weaknesses in current deidentification technologies while providing a tool for iterative development and improvement.
☆ Can a Machine Distinguish High and Low Amount of Social Creak in Speech?
Objectives: ncreased prevalence of social creak particularly among female speakers has been reported in several studies. The study of social creak has been previously conducted by combining perceptual evaluation of speech with conventional acoustical parameters such as the harmonic-to-noise ratio and cepstral peak prominence. In the current study, machine learning (ML) was used to automatically distinguish speech of low amount of social creak from speech of high amount of social creak. Methods: The amount of creak in continuous speech samples produced in Finnish by 90 female speakers was first perceptually assessed by two voice specialists. Based on their assessments, the speech samples were divided into two categories (low $vs$. high amount of creak). Using the speech signals and their creak labels, seven different ML models were trained. Three spectral representations were used as feature for each model. Results: The results show that the best performance (accuracy of 71.1\%) was obtained by the following two systems: an Adaboost classifier using the mel-spectrogram feature and a decision tree classifier using the mel-frequency cepstral coefficient feature. Conclusions: The study of social creak is becoming increasingly popular in sociolinguistic and vocological research. The conventional human perceptual assessment of the amount of creak is laborious and therefore ML technology could be used to assist researchers studying social creak. The classification systems reported in this study could be considered as baselines in future ML-based studies on social creak.
comment: Accepted in Journal of Voice
☆ SG-FSM: A Self-Guiding Zero-Shot Prompting Paradigm for Multi-Hop Question Answering Based on Finite State Machine
Large Language Models with chain-of-thought prompting, such as OpenAI-o1, have shown impressive capabilities in natural language inference tasks. However, Multi-hop Question Answering (MHQA) remains challenging for many existing models due to issues like hallucination, error propagation, and limited context length. To address these challenges and enhance LLMs' performance on MHQA, we propose the Self-Guiding prompting Finite State Machine (SG-FSM), designed to strengthen multi-hop reasoning abilities. Unlike traditional chain-of-thought methods, SG-FSM tackles MHQA by iteratively breaking down complex questions into sub-questions, correcting itself to improve accuracy. It processes one sub-question at a time, dynamically deciding the next step based on the current context and results, functioning much like an automaton. Experiments across various benchmarks demonstrate the effectiveness of our approach, outperforming strong baselines on challenging datasets such as Musique. SG-FSM reduces hallucination, enabling recovery of the correct final answer despite intermediate errors. It also improves adherence to specified output formats, simplifying evaluation significantly.
☆ Exploring Forgetting in Large Language Model Pre-Training
Catastrophic forgetting remains a formidable obstacle to building an omniscient model in large language models (LLMs). Despite the pioneering research on task-level forgetting in LLM fine-tuning, there is scant focus on forgetting during pre-training. We systematically explored the existence and measurement of forgetting in pre-training, questioning traditional metrics such as perplexity (PPL) and introducing new metrics to better detect entity memory retention. Based on our revised assessment of forgetting metrics, we explored low-cost, straightforward methods to mitigate forgetting during the pre-training phase. Further, we carefully analyzed the learning curves, offering insights into the dynamics of forgetting. Extensive evaluations and analyses on forgetting of pre-training could facilitate future research on LLMs.
☆ IPL: Leveraging Multimodal Large Language Models for Intelligent Product Listing
Unlike professional Business-to-Consumer (B2C) e-commerce platforms (e.g., Amazon), Consumer-to-Consumer (C2C) platforms (e.g., Facebook marketplace) are mainly targeting individual sellers who usually lack sufficient experience in e-commerce. Individual sellers often struggle to compose proper descriptions for selling products. With the recent advancement of Multimodal Large Language Models (MLLMs), we attempt to integrate such state-of-the-art generative AI technologies into the product listing process. To this end, we develop IPL, an Intelligent Product Listing tool tailored to generate descriptions using various product attributes such as category, brand, color, condition, etc. IPL enables users to compose product descriptions by merely uploading photos of the selling product. More importantly, it can imitate the content style of our C2C platform Xianyu. This is achieved by employing domain-specific instruction tuning on MLLMs and adopting the multi-modal Retrieval-Augmented Generation (RAG) process. A comprehensive empirical evaluation demonstrates that the underlying model of IPL significantly outperforms the base model in domain-specific tasks while producing less hallucination. IPL has been successfully deployed in our production system, where 72% of users have their published product listings based on the generated content, and those product listings are shown to have a quality score 5.6% higher than those without AI assistance.
☆ Learning Mathematical Rules with Large Language Models NeurIPS'24
In this paper, we study the ability of large language models to learn specific mathematical rules such as distributivity or simplifying equations. We present an empirical analysis of their ability to generalize these rules, as well as to reuse them in the context of word problems. For this purpose, we provide a rigorous methodology to build synthetic data incorporating such rules, and perform fine-tuning of large language models on such data. Our experiments show that our model can learn and generalize these rules to some extent, as well as suitably reuse them in the context of word problems.
comment: 4th MATH-AI Workshop at NeurIPS'24
☆ Math Neurosurgery: Isolating Language Models' Math Reasoning Abilities Using Only Forward Passes
Math reasoning is a highly active area of Large Language Model (LLM) research because it is a hallmark of artificial intelligence. However, few works have explored how math reasoning is encoded within LLM parameters and if it is a skill that can be isolated within a model. Doing so could allow targeted intervention to improve math performance without altering non-math behavior and foster understanding of how models encode math reasoning. We introduce Math Neurosurgery (MathNeuro), a method for isolating math-specific parameters in LLMs using only forward passes. MathNeuro builds on existing work by using weights and activations to calculate parameter importance, but isolates math-specific parameters by removing those important for general language tasks. Pruning parameters MathNeuro identifies deletes a LLM's math reasoning ability without destroying its general language ability. Scaling these parameters by a small constant improves a pretrained or instruction-tuned LLM's performance by 4-17% on GSM8K while leaving non-math behavior unaltered. MathNeuro is also data efficient: most of its effectiveness holds when identifying math-specific parameters using a single sample. MathNeuro highlights the potential for future work to intervene on math-specific parameters.
comment: 21 pages, 29 figures
☆ EnvBridge: Bridging Diverse Environments with Cross-Environment Knowledge Transfer for Embodied AI
In recent years, Large Language Models (LLMs) have demonstrated high reasoning capabilities, drawing attention for their applications as agents in various decision-making processes. One notably promising application of LLM agents is robotic manipulation. Recent research has shown that LLMs can generate text planning or control code for robots, providing substantial flexibility and interaction capabilities. However, these methods still face challenges in terms of flexibility and applicability across different environments, limiting their ability to adapt autonomously. Current approaches typically fall into two categories: those relying on environment-specific policy training, which restricts their transferability, and those generating code actions based on fixed prompts, which leads to diminished performance when confronted with new environments. These limitations significantly constrain the generalizability of agents in robotic manipulation. To address these limitations, we propose a novel method called EnvBridge. This approach involves the retention and transfer of successful robot control codes from source environments to target environments. EnvBridge enhances the agent's adaptability and performance across diverse settings by leveraging insights from multiple environments. Notably, our approach alleviates environmental constraints, offering a more flexible and generalizable solution for robotic manipulation tasks. We validated the effectiveness of our method using robotic manipulation benchmarks: RLBench, MetaWorld, and CALVIN. Our experiments demonstrate that LLM agents can successfully leverage diverse knowledge sources to solve complex tasks. Consequently, our approach significantly enhances the adaptability and robustness of robotic manipulation agents in planning across diverse environments.
☆ Tracing the Development of the Virtual Particle Concept Using Semantic Change Detection
Virtual particles are peculiar objects. They figure prominently in much of theoretical and experimental research in elementary particle physics. But exactly what they are is far from obvious. In particular, to what extent they should be considered "real" remains a matter of controversy in philosophy of science. Also their origin and development has only recently come into focus of scholarship in the history of science. In this study, we propose using the intriguing case of virtual particles to discuss the efficacy of Semantic Change Detection (SCD) based on contextualized word embeddings from a domain-adapted BERT model in studying specific scientific concepts. We find that the SCD metrics align well with qualitative research insights in the history and philosophy of science, as well as with the results obtained from Dependency Parsing to determine the frequency and connotations of the term "virtual." Still, the metrics of SCD provide additional insights over and above the qualitative research and the Dependency Parsing. Among other things, the metrics suggest that the concept of the virtual particle became more stable after 1950 but at the same time also more polysemous.
comment: CHR 2024: Computational Humanities Research Conference
☆ ETHIC: Evaluating Large Language Models on Long-Context Tasks with High Information Coverage
Recent advancements in large language models (LLM) capable of processing extremely long texts highlight the need for a dedicated evaluation benchmark to assess their long-context capabilities. However, existing methods, like the needle-in-a-haystack test, do not effectively assess whether these models fully utilize contextual information, raising concerns about the reliability of current evaluation techniques. To thoroughly examine the effectiveness of existing benchmarks, we introduce a new metric called information coverage (IC), which quantifies the proportion of the input context necessary for answering queries. Our findings indicate that current benchmarks exhibit low IC; although the input context may be extensive, the actual usable context is often limited. To address this, we present ETHIC, a novel benchmark designed to assess LLMs' ability to leverage the entire context. Our benchmark comprises 2,648 test instances spanning four long-context tasks with high IC scores in the domains of books, debates, medicine, and law. Our evaluations reveal significant performance drops in contemporary LLMs, highlighting a critical challenge in managing long contexts. Our benchmark is available at https://github.com/dmis-lab/ETHIC.
comment: 15 pages, 5 figures
☆ Trustworthy Alignment of Retrieval-Augmented Large Language Models via Reinforcement Learning ICML 2024
Trustworthiness is an essential prerequisite for the real-world application of large language models. In this paper, we focus on the trustworthiness of language models with respect to retrieval augmentation. Despite being supported with external evidence, retrieval-augmented generation still suffers from hallucinations, one primary cause of which is the conflict between contextual and parametric knowledge. We deem that retrieval-augmented language models have the inherent capabilities of supplying response according to both contextual and parametric knowledge. Inspired by aligning language models with human preference, we take the first step towards aligning retrieval-augmented language models to a status where it responds relying merely on the external evidence and disregards the interference of parametric knowledge. Specifically, we propose a reinforcement learning based algorithm Trustworthy-Alignment, theoretically and experimentally demonstrating large language models' capability of reaching a trustworthy status without explicit supervision on how to respond. Our work highlights the potential of large language models on exploring its intrinsic abilities by its own and expands the application scenarios of alignment from fulfilling human preference to creating trustworthy agents.
comment: ICML 2024
☆ Assessment of Transformer-Based Encoder-Decoder Model for Human-Like Summarization
In recent times, extracting valuable information from large text is making significant progress. Especially in the current era of social media, people expect quick bites of information. Automatic text summarization seeks to tackle this by slimming large texts down into more manageable summaries. This important research area can aid in decision-making by digging out salient content from large text. With the progress in deep learning models, significant work in language models has emerged. The encoder-decoder framework in deep learning has become the central approach for automatic text summarization. This work leverages transformer-based BART model for human-like summarization which is an open-ended problem with many challenges. On training and fine-tuning the encoder-decoder model, it is tested with diverse sample articles and the quality of summaries of diverse samples is assessed based on human evaluation parameters. Further, the finetuned model performance is compared with the baseline pretrained model based on evaluation metrics like ROUGE score and BERTScore. Additionally, domain adaptation of the model is required for improved performance of abstractive summarization of dialogues between interlocutors. On investigating, the above popular evaluation metrics are found to be insensitive to factual errors. Further investigation of the summaries generated by finetuned model is done using the contemporary evaluation metrics of factual consistency like WeCheck and SummaC. Empirical results on BBC News articles highlight that the gold standard summaries written by humans are more factually consistent by 17% than the abstractive summaries generated by finetuned model.
comment: Pre-print
☆ Analyzing and Evaluating Correlation Measures in NLG Meta-Evaluation
The correlation between NLG automatic evaluation metrics and human evaluation is often regarded as a critical criterion for assessing the capability of an evaluation metric. However, different grouping methods and correlation coefficients result in various types of correlation measures used in meta-evaluation. In specific evaluation scenarios, prior work often directly follows conventional measure settings, but the characteristics and differences between these measures have not gotten sufficient attention. Therefore, this paper analyzes 12 common correlation measures using a large amount of real-world data from six widely-used NLG evaluation datasets and 32 evaluation metrics, revealing that different measures indeed impact the meta-evaluation results. Furthermore, we propose three perspectives that reflect the capability of meta-evaluation and find that the measure using global grouping and Pearson correlation exhibits the best overall performance, involving the discriminative power, ranking consistency, and sensitivity to score granularity.
☆ Optimizing Chain-of-Thought Reasoning: Tackling Arranging Bottleneck via Plan Augmentation
Multi-step reasoning ability of large language models is crucial in tasks such as math and tool utilization. Current researches predominantly focus on enhancing model performance in these multi-step reasoning tasks through fine-tuning with Chain-of-Thought (CoT) steps, yet these methods tend to be heuristic, without exploring nor resolving the bottleneck. In this study, we subdivide CoT reasoning into two parts: arranging and executing, and identify that the bottleneck of models mainly lies in arranging rather than executing. Based on this finding, we propose a plan-based training and reasoning method that guides models to generate arranging steps through abstract plans. We experiment on both math (GSM8k) and tool utilization (ToolBench) benchmarks. Results show that compared to fine-tuning directly with CoT data, our approach achieves a better performance on alleviating arranging bottleneck, particularly excelling in long-distance reasoning generalization.
☆ Context-aware Inductive Knowledge Graph Completion with Latent Type Constraints and Subgraph Reasoning
Inductive knowledge graph completion (KGC) aims to predict missing triples with unseen entities. Recent works focus on modeling reasoning paths between the head and tail entity as direct supporting evidence. However, these methods depend heavily on the existence and quality of reasoning paths, which limits their general applicability in different scenarios. In addition, we observe that latent type constraints and neighboring facts inherent in KGs are also vital in inferring missing triples. To effectively utilize all useful information in KGs, we introduce CATS, a novel context-aware inductive KGC solution. With sufficient guidance from proper prompts and supervised fine-tuning, CATS activates the strong semantic understanding and reasoning capabilities of large language models to assess the existence of query triples, which consist of two modules. First, the type-aware reasoning module evaluates whether the candidate entity matches the latent entity type as required by the query relation. Then, the subgraph reasoning module selects relevant reasoning paths and neighboring facts, and evaluates their correlation to the query triple. Experiment results on three widely used datasets demonstrate that CATS significantly outperforms state-of-the-art methods in 16 out of 18 transductive, inductive, and few-shot settings with an average absolute MRR improvement of 7.2%.
☆ Controlled Low-Rank Adaptation with Subspace Regularization for Continued Training on Large Language Models
Large language models (LLMs) exhibit remarkable capabilities in natural language processing but face catastrophic forgetting when learning new tasks, where adaptation to a new domain leads to a substantial decline in performance on previous tasks. In this paper, we propose Controlled LoRA (CLoRA), a subspace regularization method on LoRA structure. Aiming to reduce the scale of output change while introduce minimal constraint on model capacity, CLoRA imposes constraint on the direction of updating matrix null space. Experimental results on commonly used LLM finetuning tasks reveal that CLoRA significantly outperforms existing LoRA subsequent methods on both in-domain and outdomain evaluations, highlighting the superority of CLoRA as a effective parameter-efficient finetuning method with catastrophic forgetting mitigating. Further investigation for model parameters indicates that CLoRA effectively balances the trade-off between model capacity and degree of forgetting.
☆ Correct after Answer: Enhancing Multi-Span Question Answering with Post-Processing Method EMNLP 2024
Multi-Span Question Answering (MSQA) requires models to extract one or multiple answer spans from a given context to answer a question. Prior work mainly focuses on designing specific methods or applying heuristic strategies to encourage models to predict more correct predictions. However, these models are trained on gold answers and fail to consider the incorrect predictions. Through a statistical analysis, we observe that models with stronger abilities do not predict less incorrect predictions compared with other models. In this work, we propose Answering-Classifying-Correcting (ACC) framework, which employs a post-processing strategy to handle incorrect predictions. Specifically, the ACC framework first introduces a classifier to classify the predictions into three types and exclude "wrong predictions", then introduces a corrector to modify "partially correct predictions". Experiments on several MSQA datasets show that ACC framework significantly improves the Exact Match (EM) scores, and further analysis demostrates that ACC framework efficiently reduces the number of incorrect predictions, improving the quality of predictions.
comment: Accepted by EMNLP 2024 Findings
☆ Beyond Retrieval: Generating Narratives in Conversational Recommender Systems
The recent advances in Large Language Model's generation and reasoning capabilities present an opportunity to develop truly conversational recommendation systems. However, effectively integrating recommender system knowledge into LLMs for natural language generation which is tailored towards recommendation tasks remains a challenge. This paper addresses this challenge by making two key contributions. First, we introduce a new dataset (REGEN) for natural language generation tasks in conversational recommendations. REGEN (Reviews Enhanced with GEnerative Narratives) extends the Amazon Product Reviews dataset with rich user narratives, including personalized explanations of product preferences, product endorsements for recommended items, and summaries of user purchase history. REGEN is made publicly available to facilitate further research. Furthermore, we establish benchmarks using well-known generative metrics, and perform an automated evaluation of the new dataset using a rater LLM. Second, the paper introduces a fusion architecture (CF model with an LLM) which serves as a baseline for REGEN. And to the best of our knowledge, represents the first attempt to analyze the capabilities of LLMs in understanding recommender signals and generating rich narratives. We demonstrate that LLMs can effectively learn from simple fusion architectures utilizing interaction-based CF embeddings, and this can be further enhanced using the metadata and personalization data associated with items. Our experiments show that combining CF and content embeddings leads to improvements of 4-12% in key language metrics compared to using either type of embedding individually. We also provide an analysis to interpret how CF and content embeddings contribute to this new generative task.
☆ Context-Aware LLM Translation System Using Conversation Summarization and Dialogue History
Translating conversational text, particularly in customer support contexts, presents unique challenges due to its informal and unstructured nature. We propose a context-aware LLM translation system that leverages conversation summarization and dialogue history to enhance translation quality for the English-Korean language pair. Our approach incorporates the two most recent dialogues as raw data and a summary of earlier conversations to manage context length effectively. We demonstrate that this method significantly improves translation accuracy, maintaining coherence and consistency across conversations. This system offers a practical solution for customer support translation tasks, addressing the complexities of conversational text.
comment: Accepted to WMT 2024
☆ Forewarned is Forearmed: Leveraging LLMs for Data Synthesis through Failure-Inducing Exploration
Large language models (LLMs) have significantly benefited from training on diverse, high-quality task-specific data, leading to impressive performance across a range of downstream applications. Current methods often rely on human-annotated data or predefined task templates to direct powerful LLMs in synthesizing task-relevant data for effective model training. However, this dependence on manually designed components may constrain the scope of generated data, potentially overlooking critical edge cases or novel scenarios that could challenge the model. In this paper, we present a novel approach, ReverseGen, designed to automatically generate effective training samples that expose the weaknesses of LLMs. Specifically, we introduce a dedicated proposer trained to produce queries that lead target models to generate unsatisfactory responses. These failure-inducing queries are then used to construct training data, helping to address the models' shortcomings and improve overall performance. Our approach is flexible and can be applied to models of various scales (3B, 7B, and 8B). We evaluate ReverseGen on three key applications (safety, honesty, and math), demonstrating that our generated data is both highly effective and diverse. Models fine-tuned with ReverseGen-generated data consistently outperform those trained on human-annotated or general model-generated data, offering a new perspective on data synthesis for task-specific LLM enhancement.
☆ Enhancing Low-Resource ASR through Versatile TTS: Bridging the Data Gap
While automatic speech recognition (ASR) systems have achieved remarkable performance with large-scale datasets, their efficacy remains inadequate in low-resource settings, encompassing dialects, accents, minority languages, and long-tail hotwords, domains with significant practical relevance. With the advent of versatile and powerful text-to-speech (TTS) models, capable of generating speech with human-level naturalness, expressiveness, and diverse speaker profiles, leveraging TTS for ASR data augmentation provides a cost-effective and practical approach to enhancing ASR performance. Comprehensive experiments on an unprecedentedly rich variety of low-resource datasets demonstrate consistent and substantial performance improvements, proving that the proposed method of enhancing low-resource ASR through a versatile TTS model is highly effective and has broad application prospects. Furthermore, we delve deeper into key characteristics of synthesized speech data that contribute to ASR improvement, examining factors such as text diversity, speaker diversity, and the volume of synthesized data, with text diversity being studied for the first time in this work. We hope our findings provide helpful guidance and reference for the practical application of TTS-based data augmentation and push the advancement of low-resource ASR one step further.
☆ Magnetic Preference Optimization: Achieving Last-iterate Convergence for Language Models Alignment
Self-play methods have demonstrated remarkable success in enhancing model capabilities across various domains. In the context of Reinforcement Learning from Human Feedback (RLHF), self-play not only boosts Large Language Model (LLM) performance but also overcomes the limitations of traditional Bradley-Terry (BT) model assumptions by finding the Nash equilibrium (NE) of a preference-based, two-player constant-sum game. However, existing methods either guarantee only average-iterate convergence, incurring high storage and inference costs, or converge to the NE of a regularized game, failing to accurately reflect true human preferences. In this paper, we introduce Magnetic Preference Optimization (MPO), a novel approach capable of achieving last-iterate convergence to the NE of the original game, effectively overcoming the limitations of existing methods. Building upon Magnetic Mirror Descent (MMD), MPO attains a linear convergence rate, making it particularly suitable for fine-tuning LLMs. To ensure our algorithm is both theoretically sound and practically viable, we present a simple yet effective implementation that adapts the theoretical insights to the RLHF setting. Empirical results demonstrate that MPO can significantly enhance the performance of LLMs, highlighting the potential of self-play methods in alignment.
comment: Under review
☆ DENOASR: Debiasing ASRs through Selective Denoising
Automatic Speech Recognition (ASR) systems have been examined and shown to exhibit biases toward particular groups of individuals, influenced by factors such as demographic traits, accents, and speech styles. Noise can disproportionately impact speakers with certain accents, dialects, or speaking styles, leading to biased error rates. In this work, we introduce a novel framework DENOASR, which is a selective denoising technique to reduce the disparity in the word error rates between the two gender groups, male and female. We find that a combination of two popular speech denoising techniques, viz. DEMUCS and LE, can be effectively used to mitigate ASR disparity without compromising their overall performance. Experiments using two state-of-the-art open-source ASRs - OpenAI WHISPER and NVIDIA NEMO - on multiple benchmark datasets, including TIE, VOX-POPULI, TEDLIUM, and FLEURS, show that there is a promising reduction in the average word error rate gap across the two gender groups. For a given dataset, the denoising is selectively applied on speech samples having speech intelligibility below a certain threshold, estimated using a small validation sample, thus ameliorating the need for large-scale human-written ground-truth transcripts. Our findings suggest that selective denoising can be an elegant approach to mitigate biases in present-day ASR systems.
comment: Paper accepted at IEEE ICKG 2024
☆ Influential Language Data Selection via Gradient Trajectory Pursuit
Curating a desirable dataset for training has been the core of building highly capable large language models (Touvron et al., 2023; Achiam et al., 2023; Team et al.,2024). Gradient influence scores (Pruthi et al., 2020; Xia et al., 2024) are shown to be correlated with model performance and are commonly used as the criterion for data selection. However, existing methods are built upon either individual sample rankings or inefficient matching process, leading to suboptimal performance or scaling up issues.In this paper, we propose Gradient Trajectory Pursuit (GTP), an algorithm that performs pursuit of gradient trajectories via jointly selecting data points under an L0-norm regularized objective. The proposed algorithm highlights: (1) joint selection instead of independent top-k selection, which automatically de-duplicates samples; (2) higher efficiency with compressive sampling processes, which can be further sped up using a distributed framework. In the experiments, we demonstrate the algorithm in both in-domain and target-domain selection benchmarks and show that it outperforms top-k selection and competitive algorithms consistently, for example, our algorithm chooses as low as 0.5% data to achieve full performance on the targeted instruction tuning tasks
☆ Atomic Fact Decomposition Helps Attributed Question Answering
Attributed Question Answering (AQA) aims to provide both a trustworthy answer and a reliable attribution report for a given question. Retrieval is a widely adopted approach, including two general paradigms: Retrieval-Then-Read (RTR) and post-hoc retrieval. Recently, Large Language Models (LLMs) have shown remarkable proficiency, prompting growing interest in AQA among researchers. However, RTR-based AQA often suffers from irrelevant knowledge and rapidly changing information, even when LLMs are adopted, while post-hoc retrieval-based AQA struggles with comprehending long-form answers with complex logic, and precisely identifying the content needing revision and preserving the original intent. To tackle these problems, this paper proposes an Atomic fact decomposition-based Retrieval and Editing (ARE) framework, which decomposes the generated long-form answers into molecular clauses and atomic facts by the instruction-tuned LLMs. Notably, the instruction-tuned LLMs are fine-tuned using a well-constructed dataset, generated from large scale Knowledge Graphs (KGs). This process involves extracting one-hop neighbors from a given set of entities and transforming the result into coherent long-form text. Subsequently, ARE leverages a search engine to retrieve evidences related to atomic facts, inputting these evidences into an LLM-based verifier to determine whether the facts require expansion for re-retrieval or editing. Furthermore, the edited facts are backtracked into the original answer, with evidence aggregated based on the relationship between molecular clauses and atomic facts. Extensive evaluations demonstrate the superior performance of our proposed method over the state-of-the-arts on several datasets, with an additionally proposed new metric $Attr_{p}$ for evaluating the precision of evidence attribution.
☆ PLDR-LLM: Large Language Model from Power Law Decoder Representations
We present the Large Language Model from Power Law Decoder Representations (PLDR-LLM), a language model that leverages non-linear and linear transformations through Power Law Graph Attention mechanism to generate well-defined deductive and inductive outputs. We pretrain the PLDR-LLMs of varying layer sizes with a small batch size of 32 and $\sim$8B tokens from the RefinedWeb dataset, and show that they achieve competitive performance in zero-shot and few-shot settings compared to scaled dot-product LLMs of similar model size reported in the literature. We show that deductive outputs of PLDR-LLMs can be used to compare model characteristics or improve the performance by introducing the Directed Acyclic Graph (DAG) loss as a metric and regularizer. Our results indicate that the initial maximum learning rate and warm-up steps have a lasting impact on deductive outputs throughout the pretraining. We provide a detailed description of PLDR-LLM architecture, its implementation and the pretraining procedure.
comment: 22 pages, 4 figures, 10 tables
☆ Methods of improving LLM training stability
Training stability of large language models(LLMs) is an important research topic. Reproducing training instabilities can be costly, so we use a small language model with 830M parameters and experiment with higher learning rates to force models to diverge. One of the sources of training instability is the growth of logits in attention layers. We extend the focus of the previous work and look not only at the magnitude of the logits but at all outputs of linear layers in the Transformer block. We observe that with a high learning rate the L2 norm of all linear layer outputs can grow with each training step and the model diverges. Specifically we observe that QKV, Proj and FC2 layers have the largest growth of the output magnitude. This prompts us to explore several options: 1) apply layer normalization not only after QK layers but also after Proj and FC2 layers too; 2) apply layer normalization after the QKV layer (and remove pre normalization). 3) apply QK layer normalization together with softmax capping. We show that with the last two methods we can increase learning rate by 1.5x (without model divergence) in comparison to an approach based on QK layer normalization only. Also we observe significant perplexity improvements for all three methods in comparison to the baseline model.
☆ Improving Causal Reasoning in Large Language Models: A Survey
Causal reasoning (CR) is a crucial aspect of intelligence, essential for problem-solving, decision-making, and understanding the world. While large language models (LLMs) can generate rationales for their outputs, their ability to reliably perform causal reasoning remains uncertain, often falling short in tasks requiring a deep understanding of causality. In this survey, we provide a comprehensive review of research aimed at enhancing LLMs for causal reasoning. We categorize existing methods based on the role of LLMs: either as reasoning engines or as helpers providing knowledge or data to traditional CR methods, followed by a detailed discussion of the methodologies in each category. We then evaluate the performance of LLMs on various causal reasoning tasks, providing key findings and in-depth analysis. Finally, we provide insights from current studies and highlight promising directions for future research. We aim for this work to serve as a comprehensive resource, fostering further advancements in causal reasoning with LLMs. Resources are available at https://github.com/chendl02/Awesome-LLM-causal-reasoning.
☆ SafetyAnalyst: Interpretable, transparent, and steerable LLM safety moderation
The ideal LLM content moderation system would be both structurally interpretable (so its decisions can be explained to users) and steerable (to reflect a community's values or align to safety standards). However, current systems fall short on both of these dimensions. To address this gap, we present SafetyAnalyst, a novel LLM safety moderation framework. Given a prompt, SafetyAnalyst creates a structured "harm-benefit tree," which identifies 1) the actions that could be taken if a compliant response were provided, 2) the harmful and beneficial effects of those actions (along with their likelihood, severity, and immediacy), and 3) the stakeholders that would be impacted by those effects. It then aggregates this structured representation into a harmfulness score based on a parameterized set of safety preferences, which can be transparently aligned to particular values. Using extensive harm-benefit features generated by SOTA LLMs on 19k prompts, we fine-tuned an open-weight LM to specialize in generating harm-benefit trees through symbolic knowledge distillation. On a comprehensive set of prompt safety benchmarks, we show that our system (average F1=0.75) outperforms existing LLM safety moderation systems (average F1$<$0.72) on prompt harmfulness classification, while offering the additional advantages of interpretability and steerability.
☆ RKadiyala at SemEval-2024 Task 8: Black-Box Word-Level Text Boundary Detection in Partially Machine Generated Texts
With increasing usage of generative models for text generation and widespread use of machine generated texts in various domains, being able to distinguish between human written and machine generated texts is a significant challenge. While existing models and proprietary systems focus on identifying whether given text is entirely human written or entirely machine generated, only a few systems provide insights at sentence or paragraph level at likelihood of being machine generated at a non reliable accuracy level, working well only for a set of domains and generators. This paper introduces few reliable approaches for the novel task of identifying which part of a given text is machine generated at a word level while comparing results from different approaches and methods. We present a comparison with proprietary systems , performance of our model on unseen domains' and generators' texts. The findings reveal significant improvements in detection accuracy along with comparison on other aspects of detection capabilities. Finally we discuss potential avenues for improvement and implications of our work. The proposed model is also well suited for detecting which parts of a text are machine generated in outputs of Instruct variants of many LLMs.
comment: published at naacl 2024
☆ Adsorb-Agent: Autonomous Identification of Stable Adsorption Configurations via Large Language Model Agent
Adsorption energy is a key reactivity descriptor in catalysis, enabling the efficient screening of potential catalysts. However, determining adsorption energy involves comparing the energies of multiple adsorbate-catalyst configurations, which is computationally demanding due to a large number of possible configurations. Current algorithmic approaches typically enumerate adsorption sites and configurations without leveraging theoretical insights to guide the initial setup. In this work, we present Adsorb-Agent, a Large Language Model (LLM) agent designed to efficiently derive system-specific stable adsorption configurations with minimal human intervention. Adsorb-Agent leverages built-in knowledge and emergent reasoning capabilities, significantly reducing the number of initial configurations required while improving accuracy in predicting the minimum adsorption energy. We demonstrate its performance using two example systems, NNH-CuPd3 (111) and NNH-Mo3Pd (111), for the Nitrogen Reduction Reaction (NRR), a sustainable alternative to the Haber-Bosch process. Adsorb-Agent outperforms conventional "heuristic" and "random" algorithms by identifying lower-energy configurations with fewer initial setups, reducing computational costs while enhancing accuracy. This highlights its potential to accelerate catalyst discovery.
comment: 11 pages, 3 figures
☆ Chatting with Bots: AI, Speech Acts, and the Edge of Assertion
This paper addresses the question of whether large language model-powered chatbots are capable of assertion. According to what we call the Thesis of Chatbot Assertion (TCA), chatbots are the kinds of things that can assert, and at least some of the output produced by current-generation chatbots qualifies as assertion. We provide some motivation for TCA, arguing that it ought to be taken seriously and not simply dismissed. We also review recent objections to TCA, arguing that these objections are weighty. We thus confront the following dilemma: how can we do justice to both the considerations for and against TCA? We consider two influential responses to this dilemma - the first appeals to the notion of proxy-assertion; the second appeals to fictionalism - and argue that neither is satisfactory. Instead, reflecting on the ontogenesis of assertion, we argue that we need to make space for a category of proto-assertion. We then apply the category of proto-assertion to chatbots, arguing that treating chatbots as proto-assertors provides a satisfactory resolution to the dilemma of chatbot assertion.
☆ A Statistical Analysis of LLMs' Self-Evaluation Using Proverbs
Large language models (LLMs) such as ChatGPT, GPT-4, Claude-3, and Llama are being integrated across a variety of industries. Despite this rapid proliferation, experts are calling for caution in the interpretation and adoption of LLMs, owing to numerous associated ethical concerns. Research has also uncovered shortcomings in LLMs' reasoning and logical abilities, raising questions on the potential of LLMs as evaluation tools. In this paper, we investigate LLMs' self-evaluation capabilities on a novel proverb reasoning task. We introduce a novel proverb database consisting of 300 proverb pairs that are similar in intent but different in wordings, across topics spanning gender, wisdom, and society. We propose tests to evaluate textual consistencies as well as numerical consistencies across similar proverbs, and demonstrate the effectiveness of our method and dataset in identifying failures in LLMs' self-evaluation which in turn can highlight issues related to gender stereotypes and lack of cultural understanding in LLMs.
☆ LLMScan: Causal Scan for LLM Misbehavior Detection
Despite the success of Large Language Models (LLMs) across various fields, their potential to generate untruthful, biased and harmful responses poses significant risks, particularly in critical applications. This highlights the urgent need for systematic methods to detect and prevent such misbehavior. While existing approaches target specific issues such as harmful responses, this work introduces LLMScan, an innovative LLM monitoring technique based on causality analysis, offering a comprehensive solution. LLMScan systematically monitors the inner workings of an LLM through the lens of causal inference, operating on the premise that the LLM's `brain' behaves differently when misbehaving. By analyzing the causal contributions of the LLM's input tokens and transformer layers, LLMScan effectively detects misbehavior. Extensive experiments across various tasks and models reveal clear distinctions in the causal distributions between normal behavior and misbehavior, enabling the development of accurate, lightweight detectors for a variety of misbehavior detection tasks.
☆ Graph-Structured Trajectory Extraction from Travelogues
Previous studies on sequence-based extraction of human movement trajectories have an issue of inadequate trajectory representation. Specifically, a pair of locations may not be lined up in a sequence especially when one location includes the other geographically. In this study, we propose a graph representation that retains information on the geographic hierarchy as well as the temporal order of visited locations, and have constructed a benchmark dataset for graph-structured trajectory extraction. The experiments with our baselines have demonstrated that it is possible to accurately predict visited locations and the order among them, but it remains a challenge to predict the hierarchical relations.
☆ Distill-SynthKG: Distilling Knowledge Graph Synthesis Workflow for Improved Coverage and Efficiency
Knowledge graphs (KGs) generated by large language models (LLMs) are becoming increasingly valuable for Retrieval-Augmented Generation (RAG) applications that require knowledge-intensive reasoning. However, existing KG extraction methods predominantly rely on prompt-based approaches, which are inefficient for processing large-scale corpora. These approaches often suffer from information loss, particularly with long documents, due to the lack of specialized design for KG construction. Additionally, there is a gap in evaluation datasets and methodologies for ontology-free KG construction. To overcome these limitations, we propose SynthKG, a multi-step, document-level ontology-free KG synthesis workflow based on LLMs. By fine-tuning a smaller LLM on the synthesized document-KG pairs, we streamline the multi-step process into a single-step KG generation approach called Distill-SynthKG, substantially reducing the number of LLM inference calls. Furthermore, we re-purpose existing question-answering datasets to establish KG evaluation datasets and introduce new evaluation metrics. Using KGs produced by Distill-SynthKG, we also design a novel graph-based retrieval framework for RAG. Experimental results demonstrate that Distill-SynthKG not only surpasses all baseline models in KG quality -- including models up to eight times larger -- but also consistently excels in retrieval and question-answering tasks. Our proposed graph retrieval framework also outperforms all KG-retrieval methods across multiple benchmark datasets. We release the SynthKG dataset and Distill-SynthKG model publicly to support further research and development.
☆ ViMGuard: A Novel Multi-Modal System for Video Misinformation Guarding
The rise of social media and short-form video (SFV) has facilitated a breeding ground for misinformation. With the emergence of large language models, significant research has gone into curbing this misinformation problem with automatic false claim detection for text. Unfortunately, the automatic detection of misinformation in SFV is a more complex problem that remains largely unstudied. While text samples are monomodal (only containing words), SFVs comprise three different modalities: words, visuals, and non-linguistic audio. In this work, we introduce Video Masked Autoencoders for Misinformation Guarding (ViMGuard), the first deep-learning architecture capable of fact-checking an SFV through analysis of all three of its constituent modalities. ViMGuard leverages a dual-component system. First, Video and Audio Masked Autoencoders analyze the visual and non-linguistic audio elements of a video to discern its intention; specifically whether it intends to make an informative claim. If it is deemed that the SFV has informative intent, it is passed through our second component: a Retrieval Augmented Generation system that validates the factual accuracy of spoken words. In evaluation, ViMGuard outperformed three cutting-edge fact-checkers, thus setting a new standard for SFV fact-checking and marking a significant stride toward trustworthy news on social platforms. To promote further testing and iteration, VimGuard was deployed into a Chrome extension and all code was open-sourced on GitHub.
comment: 7 pages, 2 figures
☆ Dynamic Adaptive Rank Space Exploration for Efficient Sentiment Analysis with Large Language Models
Sentiment analysis has become increasingly important for assessing public opinion and informing decision-making. Large language models (LLMs) have revolutionized this field by capturing nuanced language patterns. However, adapting LLMs to domain-specific sentiment analysis tasks remains challenging due to computational constraints and the need for optimal fine-tuning. To address these challenges, we propose a novel Dynamic Adaptive Rank Space Exploration (DARSE) framework for efficient and effective sentiment analysis using LLMs. DARSE consists of a coarse-grained greedy algorithm to identify the optimal rank range, a fine-grained exploration algorithm to refine rank selection, and a dynamic rank allocation method to determine the optimal rank combination for each LLM layer. Extensive experiments demonstrate that DARSE significantly improves sentiment analysis accuracy, achieving a 15.1% improvement in MSE and a 4.3% improvement in accuracy compared to previous work. Our framework strikes a balance between computational efficiency and model performance, making it a promising approach for sentiment analysis with LLMs.
☆ Do Robot Snakes Dream like Electric Sheep? Investigating the Effects of Architectural Inductive Biases on Hallucination
The growth in prominence of large language models (LLMs) in everyday life can be largely attributed to their generative abilities, yet some of this is also owed to the risks and costs associated with their use. On one front is their tendency to \textit{hallucinate} false or misleading information, limiting their reliability. On another is the increasing focus on the computational limitations associated with traditional self-attention based LLMs, which has brought about new alternatives, in particular recurrent models, meant to overcome them. Yet it remains uncommon to consider these two concerns simultaneously. Do changes in architecture exacerbate/alleviate existing concerns about hallucinations? Do they affect how and where they occur? Through an extensive evaluation, we study how these architecture-based inductive biases affect the propensity to hallucinate. While hallucination remains a general phenomenon not limited to specific architectures, the situations in which they occur and the ease with which specific types of hallucinations can be induced can significantly differ based on the model architecture. These findings highlight the need for better understanding both these problems in conjunction with each other, as well as consider how to design more universal techniques for handling hallucinations.
☆ Decoding Time Series with LLMs: A Multi-Agent Framework for Cross-Domain Annotation
Time series data is ubiquitous across various domains, including manufacturing, finance, and healthcare. High-quality annotations are essential for effectively understanding time series and facilitating downstream tasks; however, obtaining such annotations is challenging, particularly in mission-critical domains. In this paper, we propose TESSA, a multi-agent system designed to automatically generate both general and domain-specific annotations for time series data. TESSA introduces two agents: a general annotation agent and a domain-specific annotation agent. The general agent captures common patterns and knowledge across multiple source domains, leveraging both time-series-wise and text-wise features to generate general annotations. Meanwhile, the domain-specific agent utilizes limited annotations from the target domain to learn domain-specific terminology and generate targeted annotations. Extensive experiments on multiple synthetic and real-world datasets demonstrate that TESSA effectively generates high-quality annotations, outperforming existing methods.
comment: 23 pages, 9 figures, 24 tables
☆ Interação entre robôs humanoides: desenvolvendo a colaboração e comunicação autônoma
This study investigates the interaction between humanoid robots NAO and Pepper, emphasizing their potential applications in educational settings. NAO, widely used in education, and Pepper, designed for social interactions, of er new opportunities for autonomous communication and collaboration. Through a series of programmed interactions, the robots demonstrated their ability to communicate and coordinate actions autonomously, highlighting their potential as tools for enhancing learning environments. The research also explores the integration of emerging technologies, such as artificial intelligence, into these systems, allowing robots to learn from each other and adapt their behavior. The findings suggest that NAO and Pepper can significantly contribute to both technical learning and the development of social and emotional skills in students, of ering innovative pedagogical approaches through the use of humanoid robotics.
comment: in Portuguese language
☆ In Context Learning and Reasoning for Symbolic Regression with Large Language Models
Large Language Models (LLMs) are transformer-based machine learning models that have shown remarkable performance in tasks for which they were not explicitly trained. Here, we explore the potential of LLMs to perform symbolic regression -- a machine-learning method for finding simple and accurate equations from datasets. We prompt GPT-4 to suggest expressions from data, which are then optimized and evaluated using external Python tools. These results are fed back to GPT-4, which proposes improved expressions while optimizing for complexity and loss. Using chain-of-thought prompting, we instruct GPT-4 to analyze the data, prior expressions, and the scientific context (expressed in natural language) for each problem before generating new expressions. We evaluated the workflow in rediscovery of five well-known scientific equations from experimental data, and on an additional dataset without a known equation. GPT-4 successfully rediscovered all five equations, and in general, performed better when prompted to use a scratchpad and consider scientific context. We also demonstrate how strategic prompting improves the model's performance and how the natural language interface simplifies integrating theory with data. Although this approach does not outperform established SR programs where target equations are more complex, LLMs can nonetheless iterate toward improved solutions while following instructions and incorporating scientific context in natural language.
☆ Evaluating AI-Generated Essays with GRE Analytical Writing Assessment
The recent revolutionary advance in generative AI enables the generation of realistic and coherent texts by large language models (LLMs). Despite many existing evaluation metrics on the quality of the generated texts, there is still a lack of rigorous assessment of how well LLMs perform in complex and demanding writing assessments. This study examines essays generated by ten leading LLMs for the analytical writing assessment of the Graduate Record Exam (GRE). We assessed these essays using both human raters and the e-rater automated scoring engine as used in the GRE scoring pipeline. Notably, the top-performing GPT-4o received an average score of 4.67, falling between "generally thoughtful, well-developed analysis of the issue and conveys meaning clearly" and "presents a competent analysis of the issue and conveys meaning with acceptable clarity" according to the GRE scoring guideline. We also evaluated the detection accuracy of these essays, with detectors trained on essays generated by the same and different LLMs.
comment: 20 pages, 6 figures
☆ Artificial Intelligence in Brazilian News: A Mixed-Methods Analysis
The current surge in Artificial Intelligence (AI) interest, reflected in heightened media coverage since 2009, has sparked significant debate on AI's implications for privacy, social justice, workers' rights, and democracy. The media plays a crucial role in shaping public perception and acceptance of AI technologies. However, research into how AI appears in media has primarily focused on anglophone contexts, leaving a gap in understanding how AI is represented globally. This study addresses this gap by analyzing 3,560 news articles from Brazilian media published between July 1, 2023, and February 29, 2024, from 13 popular online news outlets. Using Computational Grounded Theory (CGT), the study applies Latent Dirichlet Allocation (LDA), BERTopic, and Named-Entity Recognition to investigate the main topics in AI coverage and the entities represented. The findings reveal that Brazilian news coverage of AI is dominated by topics related to applications in the workplace and product launches, with limited space for societal concerns, which mostly focus on deepfakes and electoral integrity. The analysis also highlights a significant presence of industry-related entities, indicating a strong influence of corporate agendas in the country's news. This study underscores the need for a more critical and nuanced discussion of AI's societal impacts in Brazilian media.
comment: 18 pages, 8 figures, 3 tables
☆ Scalable Influence and Fact Tracing for Large Language Model Pretraining
Training data attribution (TDA) methods aim to attribute model outputs back to specific training examples, and the application of these methods to large language model (LLM) outputs could significantly advance model transparency and data curation. However, it has been challenging to date to apply these methods to the full scale of LLM pretraining. In this paper, we refine existing gradient-based methods to work effectively at scale, allowing us to retrieve influential examples for an 8B-parameter language model from a pretraining corpus of over 160B tokens with no need for subsampling or pre-filtering. Our method combines several techniques, including optimizer state correction, a task-specific Hessian approximation, and normalized encodings, which we find to be critical for performance at scale. In quantitative evaluations on a fact tracing task, our method performs best at identifying examples that influence model predictions, but classical, model-agnostic retrieval methods such as BM25 still perform better at finding passages which explicitly contain relevant facts. These results demonstrate a misalignment between factual attribution and causal influence. With increasing model size and training tokens, we find that influence more closely aligns with attribution. Finally, we examine different types of examples identified as influential by our method, finding that while many directly entail a particular fact, others support the same output by reinforcing priors on relation types, common entities, and names.
☆ AdvWeb: Controllable Black-box Attacks on VLM-powered Web Agents
Vision Language Models (VLMs) have revolutionized the creation of generalist web agents, empowering them to autonomously complete diverse tasks on real-world websites, thereby boosting human efficiency and productivity. However, despite their remarkable capabilities, the safety and security of these agents against malicious attacks remain critically underexplored, raising significant concerns about their safe deployment. To uncover and exploit such vulnerabilities in web agents, we provide AdvWeb, a novel black-box attack framework designed against web agents. AdvWeb trains an adversarial prompter model that generates and injects adversarial prompts into web pages, misleading web agents into executing targeted adversarial actions such as inappropriate stock purchases or incorrect bank transactions, actions that could lead to severe real-world consequences. With only black-box access to the web agent, we train and optimize the adversarial prompter model using DPO, leveraging both successful and failed attack strings against the target agent. Unlike prior approaches, our adversarial string injection maintains stealth and control: (1) the appearance of the website remains unchanged before and after the attack, making it nearly impossible for users to detect tampering, and (2) attackers can modify specific substrings within the generated adversarial string to seamlessly change the attack objective (e.g., purchasing stocks from a different company), enhancing attack flexibility and efficiency. We conduct extensive evaluations, demonstrating that AdvWeb achieves high success rates in attacking SOTA GPT-4V-based VLM agent across various web tasks. Our findings expose critical vulnerabilities in current LLM/VLM-based agents, emphasizing the urgent need for developing more reliable web agents and effective defenses. Our code and data are available at https://ai-secure.github.io/AdvWeb/ .
comment: 15 pages
☆ Do Vision-Language Models Represent Space and How? Evaluating Spatial Frame of Reference Under Ambiguities NeurIPS 2024
Spatial expressions in situated communication can be ambiguous, as their meanings vary depending on the frames of reference (FoR) adopted by speakers and listeners. While spatial language understanding and reasoning by vision-language models (VLMs) have gained increasing attention, potential ambiguities in these models are still under-explored. To address this issue, we present the COnsistent Multilingual Frame Of Reference Test (COMFORT), an evaluation protocol to systematically assess the spatial reasoning capabilities of VLMs. We evaluate nine state-of-the-art VLMs using COMFORT. Despite showing some alignment with English conventions in resolving ambiguities, our experiments reveal significant shortcomings of VLMs: notably, the models (1) exhibit poor robustness and consistency, (2) lack the flexibility to accommodate multiple FoRs, and (3) fail to adhere to language-specific or culture-specific conventions in cross-lingual tests, as English tends to dominate other languages. With a growing effort to align vision-language models with human cognitive intuitions, we call for more attention to the ambiguous nature and cross-cultural diversity of spatial reasoning.
comment: Accepted to Pluralistic Alignment @ NeurIPS 2024 | Project page: https://spatial-comfort.github.io/
☆ AMUSD: Asynchronous Multi-Device Speculative Decoding for LLM Acceleration
Large language models typically generate tokens autoregressively, using each token as input for the next. Recent work on Speculative Decoding has sought to accelerate this process by employing a smaller, faster draft model to more quickly generate candidate tokens. These candidates are then verified in parallel by the larger (original) verify model, resulting in overall speedup compared to using the larger model by itself in an autoregressive fashion. In this work, we introduce AMUSD (Asynchronous Multi-device Speculative Decoding), a system that further accelerates generation by decoupling the draft and verify phases into a continuous, asynchronous approach. Unlike conventional speculative decoding, where only one model (draft or verify) performs token generation at a time, AMUSD enables both models to perform predictions independently on separate devices (e.g., GPUs). We evaluate our approach over multiple datasets and show that AMUSD achieves an average 29% improvement over speculative decoding and up to 1.96$\times$ speedup over conventional autoregressive decoding, while achieving identical output quality. Our system is open-source and available at https://github.com/BradMcDanel/AMUSD/.
comment: 4 pages, 5 figures, 1 table, 1 algorithm
☆ All Entities are Not Created Equal: Examining the Long Tail for Fine-Grained Entity Typing
Pre-trained language models (PLMs) are trained on large amounts of data, which helps capture world knowledge alongside linguistic competence. Due to this, they are extensively used for ultra-fine entity typing tasks, where they provide the entity knowledge held in its parameter space. Given that PLMs learn from co-occurrence patterns, they likely contain more knowledge or less knowledge about entities depending on their how frequent they are in the pre-training data. In this work, we probe PLMs to elicit encoded entity probabilities and demonstrate that they highly correlate with their frequency in large-scale internet data. Then, we demonstrate that entity-typing approaches that rely on PLMs struggle with entities at the long tail on the distribution. Our findings suggests that we need to go beyond PLMs to produce solutions that perform well for rare, new or infrequent entities.
☆ Captions Speak Louder than Images (CASLIE): Generalizing Foundation Models for E-commerce from High-quality Multimodal Instruction Data
Leveraging multimodal data to drive breakthroughs in e-commerce applications through Multimodal Foundation Models (MFMs) is gaining increasing attention from the research community. However, there are significant challenges that hinder the optimal use of multimodal e-commerce data by foundation models: (1) the scarcity of large-scale, high-quality multimodal benchmark datasets; and (2) the lack of effective multimodal information integration methods. To address these challenges, in this paper, we introduce MMECInstruct, the first-ever, large-scale, and high-quality multimodal instruction dataset for e-commerce. We also develop CASLIE, a simple, lightweight, yet effective framework for integrating multimodal information for e-commerce. Leveraging MMECInstruct, we fine-tune a series of e-commerce MFMs within CASLIE, denoted as CASLIE models. Our comprehensive evaluation demonstrates that CASLIE models substantially outperform 5 categories of advanced baseline models in the in-domain evaluation. Moreover, CASLIE models show strong generalizability to out-of-domain settings. MMECInstruct and CASLIE models are publicly accessible through https://ninglab.github.io/CASLIE/.
comment: Xinyi Ling and Bo Peng contributed equally to this paper
☆ Are Large Language Models Ready for Travel Planning?
While large language models (LLMs) show promise in hospitality and tourism, their ability to provide unbiased service across demographic groups remains unclear. This paper explores gender and ethnic biases when LLMs are utilized as travel planning assistants. To investigate this issue, we apply machine learning techniques to analyze travel suggestions generated from three open-source LLMs. Our findings reveal that the performance of race and gender classifiers substantially exceeds random chance, indicating differences in how LLMs engage with varied subgroups. Specifically, outputs align with cultural expectations tied to certain races and genders. To minimize the effect of these stereotypes, we used a stop-word classification strategy, which decreased identifiable differences, with no disrespectful terms found. However, hallucinations related to African American and gender minority groups were noted. In conclusion, while LLMs can generate travel plans seemingly free from bias, it remains essential to verify the accuracy and appropriateness of their recommendations.
☆ Literature Meets Data: A Synergistic Approach to Hypothesis Generation
AI holds promise for transforming scientific processes, including hypothesis generation. Prior work on hypothesis generation can be broadly categorized into theory-driven and data-driven approaches. While both have proven effective in generating novel and plausible hypotheses, it remains an open question whether they can complement each other. To address this, we develop the first method that combines literature-based insights with data to perform LLM-powered hypothesis generation. We apply our method on five different datasets and demonstrate that integrating literature and data outperforms other baselines (8.97\% over few-shot, 15.75\% over literature-based alone, and 3.37\% over data-driven alone). Additionally, we conduct the first human evaluation to assess the utility of LLM-generated hypotheses in assisting human decision-making on two challenging tasks: deception detection and AI generated content detection. Our results show that human accuracy improves significantly by 7.44\% and 14.19\% on these tasks, respectively. These findings suggest that integrating literature-based and data-driven approaches provides a comprehensive and nuanced framework for hypothesis generation and could open new avenues for scientific inquiry.
comment: 30 pages, 7 figures, code link: https://github.com/ChicagoHAI/hypothesis-generation
♻ ☆ BRIGHT: A Realistic and Challenging Benchmark for Reasoning-Intensive Retrieval
Existing retrieval benchmarks primarily consist of information-seeking queries (e.g., aggregated questions from search engines) where keyword or semantic-based retrieval is usually sufficient. However, many complex real-world queries require in-depth reasoning to identify relevant documents that go beyond surface form matching. For example, finding documentation for a coding question requires understanding the logic and syntax of the functions involved. To better benchmark retrieval on such challenging queries, we introduce BRIGHT, the first text retrieval benchmark that requires intensive reasoning to retrieve relevant documents. Our dataset consists of 1,384 real-world queries spanning diverse domains, such as economics, psychology, mathematics, and coding. These queries are drawn from naturally occurring and carefully curated human data. Extensive evaluation reveals that even state-of-the-art retrieval models perform poorly on BRIGHT. The leading model on the MTEB leaderboard (Muennighoff et al., 2023), which achieves a score of 59.0 nDCG@10, produces a score of nDCG@10 of 18.3 on BRIGHT. We show that incorporating explicit reasoning about the query improves retrieval performance by up to 12.2 points. Moreover, incorporating retrieved documents from the top-performing retriever boosts question-answering performance by over 6.6 points. We believe that BRIGHT paves the way for future research on retrieval systems in more realistic and challenging settings.
comment: 48 pages
♻ ☆ Context-Parametric Inversion: Why Instruction Finetuning May Not Actually Improve Context Reliance
A standard practice when using large language models is for users to supplement their instruction with an input context containing new information for the model to process. However, models struggle to reliably follow the input context, especially when it conflicts with their parametric knowledge from pretraining. In-principle, one would expect models to adapt to the user context better after instruction finetuning, particularly when handling knowledge conflicts. However, we observe a surprising failure mode: during instruction tuning, the context reliance under knowledge conflicts initially increases as expected, but then gradually decreases as instruction finetuning progresses. This happens while the performance on standard benchmarks keeps on increasing far after this drop. We call this phenomenon context-parametric inversion and observe it across multiple general purpose instruction tuning datasets such as TULU, Alpaca and Ultrachat, across different model families like Llama, Mistral, and Pythia. We perform various controlled studies and theoretical analysis to show that context-parametric inversion occurs due to examples in the instruction finetuning data where the input context provides information that aligns with model's parametric knowledge. Our analysis suggests some natural mitigation strategies with limited but insightful gains, and serves as a useful starting point in addressing this deficiency in instruction finetuning.
comment: Under Review
♻ ☆ Can Large Language Models Identify Authorship? EMNLP 2024
The ability to accurately identify authorship is crucial for verifying content authenticity and mitigating misinformation. Large Language Models (LLMs) have demonstrated an exceptional capacity for reasoning and problem-solving. However, their potential in authorship analysis remains under-explored. Traditional studies have depended on hand-crafted stylistic features, whereas state-of-the-art approaches leverage text embeddings from pre-trained language models. These methods, which typically require fine-tuning on labeled data, often suffer from performance degradation in cross-domain applications and provide limited explainability. This work seeks to address three research questions: (1) Can LLMs perform zero-shot, end-to-end authorship verification effectively? (2) Are LLMs capable of accurately attributing authorship among multiple candidates authors (e.g., 10 and 20)? (3) Can LLMs provide explainability in authorship analysis, particularly through the role of linguistic features? Moreover, we investigate the integration of explicit linguistic features to guide LLMs in their reasoning processes. Our assessment demonstrates LLMs' proficiency in both tasks without the need for domain-specific fine-tuning, providing explanations into their decision making via a detailed analysis of linguistic features. This establishes a new benchmark for future research on LLM-based authorship analysis.
comment: Accepted to EMNLP 2024 Findings. The main paper is 9 pages long, with 16 pages total. The code, results, dataset, and additional resources are available on the project website: https://llm-authorship.github.io/
♻ ☆ The Impact of Large Language Models in Academia: from Writing to Speaking
Large language models (LLMs) are increasingly impacting human society, particularly in textual information. Based on more than 30,000 papers and 1,000 presentations from machine learning conferences, we examined and compared the words used in writing and speaking, representing the first large-scale study of how LLMs influence the two main modes of verbal communication and expression within the same group of people. Our empirical results show that LLM-style words such as "significant" have been used more frequently in abstracts and oral presentations. The impact on speaking is beginning to emerge and is likely to grow in the future, calling attention to the implicit influence and ripple effect of LLMs on human society.
comment: 23 pages
♻ ☆ Levels of AI Agents: from Rules to Large Language Models
AI agents are defined as artificial entities to perceive the environment, make decisions and take actions. Inspired by the 6 levels of autonomous driving by Society of Automotive Engineers, the AI agents are also categorized based on utilities and strongness, as the following levels: L0, no AI, with tools taking into account perception plus actions; L1, using rule-based AI; L2, making rule-based AI replaced by IL/RL-based AI, with additional reasoning & decision making; L3, applying LLM-based AI instead of IL/RL-based AI, additionally setting up memory & reflection; L4, based on L3, facilitating autonomous learning & generalization; L5, based on L4, appending personality of emotion and character and collaborative behavior with multi-agents.
♻ ☆ LLMs left, right, and center: Assessing GPT's capabilities to label political bias from web domains
This research investigates whether OpenAI's GPT-4, a state-of-the-art large language model, can accurately classify the political bias of news sources based solely on their URLs. Given the subjective nature of political labels, third-party bias ratings like those from Ad Fontes Media, AllSides, and Media Bias/Fact Check (MBFC) are often used in research to analyze news source diversity. This study aims to determine if GPT-4 can replicate these human ratings on a seven-degree scale ("far-left" to "far-right"). The analysis compares GPT-4's classifications against MBFC's, and controls for website popularity using Open PageRank scores. Findings reveal a high correlation ($\text{Spearman's } \rho = .89$, $n = 5,877$, $p < 0.001$) between GPT-4's and MBFC's ratings, indicating the model's potential reliability. However, GPT-4 abstained from classifying approximately $\frac{2}{3}$ of the dataset. It is more likely to abstain from rating unpopular websites, which also suffer from less accurate assessments. The LLM tends to avoid classifying sources that MBFC considers to be centrist, resulting in more polarized outputs. Finally, this analysis shows a slight leftward skew in GPT's classifications compared to MBFC's. Therefore, while this paper suggests that while GPT-4 can be a scalable, cost-effective tool for political bias classification of news websites, its use should be as a complement to human judgment to mitigate biases.
comment: 12 pages, 4 figures
♻ ☆ Toolshed: Scale Tool-Equipped Agents with Advanced RAG-Tool Fusion and Tool Knowledge Bases
Recent advancements in tool-equipped Agents (LLMs) have enabled complex tasks like secure database interactions and multi-agent code development. However, scaling tool capacity beyond agent reasoning or model limits remains a challenge. In this paper, we address these challenges by introducing Toolshed Knowledge Bases, a tool knowledge base (vector database) designed to store enhanced tool representations and optimize tool selection for large-scale tool-equipped Agents. Additionally, we propose Advanced RAG-Tool Fusion, a novel ensemble of tool-applied advanced retrieval-augmented generation (RAG) techniques across the pre-retrieval, intra-retrieval, and post-retrieval phases, without requiring model fine-tuning. During pre-retrieval, tool documents are enhanced with key information and stored in the Toolshed Knowledge Base. Intra-retrieval focuses on query planning and transformation to increase retrieval accuracy. Post-retrieval refines the retrieved tool documents and enables self-reflection. Furthermore, by varying both the total number of tools (tool-M) an Agent has access to and the tool selection threshold (top-k), we address trade-offs between retrieval accuracy, agent performance, and token cost. Our approach achieves 46%, 56%, and 47% absolute improvements on the ToolE single-tool, ToolE multi-tool and Seal-Tools benchmark datasets, respectively (Recall@5).
♻ ☆ NaturalBench: Evaluating Vision-Language Models on Natural Adversarial Samples NeurIPS 24
Vision-language models (VLMs) have made significant progress in recent visual-question-answering (VQA) benchmarks that evaluate complex visio-linguistic reasoning. However, are these models truly effective? In this work, we show that VLMs still struggle with natural images and questions that humans can easily answer, which we term natural adversarial samples. We also find it surprisingly easy to generate these VQA samples from natural image-text corpora using off-the-shelf models like CLIP and ChatGPT. We propose a semi-automated approach to collect a new benchmark, NaturalBench, for reliably evaluating VLMs with 10,000 human-verified VQA samples. Crucially, we adopt a $\textbf{vision-centric}$ design by pairing each question with two images that yield different answers, preventing blind solutions from answering without using the images. This makes NaturalBench more challenging than previous benchmarks that can be solved with commonsense priors. We evaluate 53 state-of-the-art VLMs on NaturalBench, showing that models like LLaVA-OneVision, Cambrian-1, Llama3.2-Vision, Molmo, Qwen2-VL, and even GPT-4o lag 50%-70% behind human performance (over 90%). We analyze why NaturalBench is hard from two angles: (1) Compositionality: Solving NaturalBench requires diverse visio-linguistic skills, including understanding attribute bindings, object relationships, and advanced reasoning like logic and counting. To this end, unlike prior work that uses a single tag per sample, we tag each NaturalBench sample with 1 to 8 skill tags for fine-grained evaluation. (2) Biases: NaturalBench exposes severe biases in VLMs, as models often choose the same answer regardless of the image. Lastly, we apply our benchmark curation method to diverse data sources, including long captions (over 100 words) and non-English languages like Chinese and Hindi, highlighting its potential for dynamic evaluations of VLMs.
comment: Accepted to NeurIPS 24; We open-source our dataset at: https://huggingface.co/datasets/BaiqiL/NaturalBench ; Project page at: https://linzhiqiu.github.io/papers/naturalbench/
♻ ☆ Do LLMs "know" internally when they follow instructions?
Instruction-following is crucial for building AI agents with large language models (LLMs), as these models must adhere strictly to user-provided constraints and guidelines. However, LLMs often fail to follow even simple and clear instructions. To improve instruction-following behavior and prevent undesirable outputs, a deeper understanding of how LLMs' internal states relate to these outcomes is required. Our analysis of LLM internal states reveal a dimension in the input embedding space linked to successful instruction-following. We demonstrate that modifying representations along this dimension improves instruction-following success rates compared to random changes, without compromising response quality. Further investigation reveals that this dimension is more closely related to the phrasing of prompts rather than the inherent difficulty of the task or instructions. This discovery also suggests explanations for why LLMs sometimes fail to follow clear instructions and why prompt engineering is often effective, even when the content remains largely unchanged. This work provides insight into the internal workings of LLMs' instruction-following, paving the way for reliable LLM agents.
♻ ☆ Lex2Sent: A bagging approach to unsupervised sentiment analysis
Unsupervised text classification, with its most common form being sentiment analysis, used to be performed by counting words in a text that were stored in a lexicon, which assigns each word to one class or as a neutral word. In recent years, these lexicon-based methods fell out of favor and were replaced by computationally demanding fine-tuning techniques for encoder-only models such as BERT and zero-shot classification using decoder-only models such as GPT-4. In this paper, we propose an alternative approach: Lex2Sent, which provides improvement over classic lexicon methods but does not require any GPU or external hardware. To classify texts, we train embedding models to determine the distances between document embeddings and the embeddings of the parts of a suitable lexicon. We employ resampling, which results in a bagging effect, boosting the performance of the classification. We show that our model outperforms lexica and provides a basis for a high performing few-shot fine-tuning approach in the task of binary sentiment analysis.
comment: 11 pages, 1 figure
♻ ☆ Do LLMs estimate uncertainty well in instruction-following?
Large language models (LLMs) could be valuable personal AI agents across various domains, provided they can precisely follow user instructions. However, recent studies have shown significant limitations in LLMs' instruction-following capabilities, raising concerns about their reliability in high-stakes applications. Accurately estimating LLMs' uncertainty in adhering to instructions is critical to mitigating deployment risks. We present, to our knowledge, the first systematic evaluation of the uncertainty estimation abilities of LLMs in the context of instruction-following. Our study identifies key challenges with existing instruction-following benchmarks, where multiple factors are entangled with uncertainty stems from instruction-following, complicating the isolation and comparison across methods and models. To address these issues, we introduce a controlled evaluation setup with two benchmark versions of data, enabling a comprehensive comparison of uncertainty estimation methods under various conditions. Our findings show that existing uncertainty methods struggle, particularly when models make subtle errors in instruction following. While internal model states provide some improvement, they remain inadequate in more complex scenarios. The insights from our controlled evaluation setups provide a crucial understanding of LLMs' limitations and potential for uncertainty estimation in instruction-following tasks, paving the way for more trustworthy AI agents.
♻ ☆ One Thousand and One Pairs: A "novel" challenge for long-context language models EMNLP 2024
Synthetic long-context LLM benchmarks (e.g., "needle-in-the-haystack") test only surface-level retrieval capabilities, but how well can long-context LLMs retrieve, synthesize, and reason over information across book-length inputs? We address this question by creating NoCha, a dataset of 1,001 minimally different pairs of true and false claims about 67 recently-published English fictional books, written by human readers of those books. In contrast to existing long-context benchmarks, our annotators confirm that the largest share of pairs in NoCha require global reasoning over the entire book to verify. Our experiments show that while human readers easily perform this task, it is enormously challenging for all ten long-context LLMs that we evaluate: no open-weight model performs above random chance (despite their strong performance on synthetic benchmarks), while GPT-4o achieves the highest accuracy at 55.8%. Further analysis reveals that (1) on average, models perform much better on pairs that require only sentence-level retrieval vs. global reasoning; (2) model-generated explanations for their decisions are often inaccurate even for correctly-labeled claims; and (3) models perform substantially worse on speculative fiction books that contain extensive world-building. The methodology proposed in NoCha allows for the evolution of the benchmark dataset and the easy analysis of future models.
comment: EMNLP 2024, camera ready
♻ ☆ SysBench: Can Large Language Models Follow System Messages?
Large Language Models (LLMs) have become instrumental across various applications, with the customization of these models to specific scenarios becoming increasingly critical. System message, a fundamental component of LLMs, is consist of carefully crafted instructions that guide the behavior of model to meet intended goals. Despite the recognized potential of system messages to optimize AI-driven solutions, there is a notable absence of a comprehensive benchmark for evaluating how well LLMs follow system messages. To fill this gap, we introduce SysBench, a benchmark that systematically analyzes system message following ability in terms of three limitations of existing LLMs: constraint violation, instruction misjudgement and multi-turn instability. Specifically, we manually construct evaluation dataset based on six prevalent types of constraints, including 500 tailor-designed system messages and multi-turn user conversations covering various interaction relationships. Additionally, we develop a comprehensive evaluation protocol to measure model performance. Finally, we conduct extensive evaluation across various existing LLMs, measuring their ability to follow specified constraints given in system messages. The results highlight both the strengths and weaknesses of existing models, offering key insights and directions for future research. The open source library SysBench is available at https://github.com/PKU-Baichuan-MLSystemLab/SysBench.
♻ ☆ Holmes: A Benchmark to Assess the Linguistic Competence of Language Models
We introduce Holmes, a new benchmark designed to assess language models (LMs) linguistic competence - their unconscious understanding of linguistic phenomena. Specifically, we use classifier-based probing to examine LMs' internal representations regarding distinct linguistic phenomena (e.g., part-of-speech tagging). As a result, we meet recent calls to disentangle LMs' linguistic competence from other cognitive abilities, such as following instructions in prompting-based evaluations. Composing Holmes, we review over 270 probing studies and include more than 200 datasets to assess syntax, morphology, semantics, reasoning, and discourse phenomena. Analyzing over 50 LMs reveals that, aligned with known trends, their linguistic competence correlates with model size. However, surprisingly, model architecture and instruction tuning also significantly influence performance, particularly in morphology and syntax. Finally, we propose FlashHolmes, a streamlined version that reduces the computation load while maintaining high-ranking precision.
♻ ☆ Typos that Broke the RAG's Back: Genetic Attack on RAG Pipeline by Simulating Documents in the Wild via Low-level Perturbations EMNLP
The robustness of recent Large Language Models (LLMs) has become increasingly crucial as their applicability expands across various domains and real-world applications. Retrieval-Augmented Generation (RAG) is a promising solution for addressing the limitations of LLMs, yet existing studies on the robustness of RAG often overlook the interconnected relationships between RAG components or the potential threats prevalent in real-world databases, such as minor textual errors. In this work, we investigate two underexplored aspects when assessing the robustness of RAG: 1) vulnerability to noisy documents through low-level perturbations and 2) a holistic evaluation of RAG robustness. Furthermore, we introduce a novel attack method, the Genetic Attack on RAG (\textit{GARAG}), which targets these aspects. Specifically, GARAG is designed to reveal vulnerabilities within each component and test the overall system functionality against noisy documents. We validate RAG robustness by applying our \textit{GARAG} to standard QA datasets, incorporating diverse retrievers and LLMs. The experimental results show that GARAG consistently achieves high attack success rates. Also, it significantly devastates the performance of each component and their synergy, highlighting the substantial risk that minor textual inaccuracies pose in disrupting RAG systems in the real world.
comment: Findings of EMNLP Camera-ready version
♻ ☆ Moonshine: Speech Recognition for Live Transcription and Voice Commands
This paper introduces Moonshine, a family of speech recognition models optimized for live transcription and voice command processing. Moonshine is based on an encoder-decoder transformer architecture and employs Rotary Position Embedding (RoPE) instead of traditional absolute position embeddings. The model is trained on speech segments of various lengths, but without using zero-padding, leading to greater efficiency for the encoder during inference time. When benchmarked against OpenAI's Whisper tiny-en, Moonshine Tiny demonstrates a 5x reduction in compute requirements for transcribing a 10-second speech segment while incurring no increase in word error rates across standard evaluation datasets. These results highlight Moonshine's potential for real-time and resource-constrained applications.
comment: 7 pages, 6 figures, 3 tables
♻ ☆ On-Device LLMs for SMEs: Challenges and Opportunities
This paper presents a systematic review of the infrastructure requirements for deploying Large Language Models (LLMs) on-device within the context of small and medium-sized enterprises (SMEs), focusing on both hardware and software perspectives. From the hardware viewpoint, we discuss the utilization of processing units like GPUs and TPUs, efficient memory and storage solutions, and strategies for effective deployment, addressing the challenges of limited computational resources typical in SME settings. From the software perspective, we explore framework compatibility, operating system optimization, and the use of specialized libraries tailored for resource-constrained environments. The review is structured to first identify the unique challenges faced by SMEs in deploying LLMs on-device, followed by an exploration of the opportunities that both hardware innovations and software adaptations offer to overcome these obstacles. Such a structured review provides practical insights, contributing significantly to the community by enhancing the technological resilience of SMEs in integrating LLMs.
comment: 9 pages, 1 figure. The work is supported by the SIT-NVIDIA Joint AI Centre
♻ ☆ LLM Gesticulator: Leveraging Large Language Models for Scalable and Controllable Co-Speech Gesture Synthesis
In this work, we present LLM Gesticulator, an LLM-based audio-driven co-speech gesture generation framework that synthesizes full-body animations that are rhythmically aligned with the input audio while exhibiting natural movements and editability. Compared to previous work, our model demonstrates substantial scalability. As the size of the backbone LLM model increases, our framework shows proportional improvements in evaluation metrics (a.k.a. scaling law). Our method also exhibits strong controllability where the content, style of the generated gestures can be controlled by text prompt. To the best of our knowledge, LLM gesticulator is the first work that use LLM on the co-speech generation task. Evaluation with existing objective metrics and user studies indicate that our framework outperforms prior works.
♻ ☆ System 2 thinking in OpenAI's o1-preview model: Near-perfect performance on a mathematics exam
The processes underlying human cognition are often divided into System 1, which involves fast, intuitive thinking, and System 2, which involves slow, deliberate reasoning. Previously, large language models were criticized for lacking the deeper, more analytical capabilities of System 2. In September 2024, OpenAI introduced the o1 model series, designed to handle System 2-like reasoning. While OpenAI's benchmarks are promising, independent validation is still needed. In this study, we tested the o1-preview model twice on the Dutch 'Mathematics B' final exam. It scored a near-perfect 76 and 74 out of 76 points. For context, only 24 out of 16,414 students in the Netherlands achieved a perfect score. By comparison, the GPT-4o model scored 66 and 62 out of 76, well above the Dutch average of 40.63 points. Neither model had access to the exam figures. Since there was a risk of model contamination (i.e., the knowledge cutoff of o1-preview and GPT-4o was after the exam was published online), we repeated the procedure with a new Mathematics B exam that was published after the cutoff date. The results again indicated that o1-preview performed strongly (97.8th percentile), which suggests that contamination was not a factor. We also show that there is some variability in the output of o1-preview, which means that sometimes there is 'luck' (the answer is correct) or 'bad luck' (the output has diverged into something that is incorrect). We demonstrate that a self-consistency approach, where repeated prompts are given and the most common answer is selected, is a useful strategy for identifying the correct answer. It is concluded that while OpenAI's new model series holds great potential, certain risks must be considered.
♻ ☆ GLBench: A Comprehensive Benchmark for Graph with Large Language Models
The emergence of large language models (LLMs) has revolutionized the way we interact with graphs, leading to a new paradigm called GraphLLM. Despite the rapid development of GraphLLM methods in recent years, the progress and understanding of this field remain unclear due to the lack of a benchmark with consistent experimental protocols. To bridge this gap, we introduce GLBench, the first comprehensive benchmark for evaluating GraphLLM methods in both supervised and zero-shot scenarios. GLBench provides a fair and thorough evaluation of different categories of GraphLLM methods, along with traditional baselines such as graph neural networks. Through extensive experiments on a collection of real-world datasets with consistent data processing and splitting strategies, we have uncovered several key findings. Firstly, GraphLLM methods outperform traditional baselines in supervised settings, with LLM-as-enhancers showing the most robust performance. However, using LLMs as predictors is less effective and often leads to uncontrollable output issues. We also notice that no clear scaling laws exist for current GraphLLM methods. In addition, both structures and semantics are crucial for effective zero-shot transfer, and our proposed simple baseline can even outperform several models tailored for zero-shot scenarios. The data and code of the benchmark can be found at https://github.com/NineAbyss/GLBench.
♻ ☆ TempoFormer: A Transformer for Temporally-aware Representations in Change Detection EMNLP
Dynamic representation learning plays a pivotal role in understanding the evolution of linguistic content over time. On this front both context and time dynamics as well as their interplay are of prime importance. Current approaches model context via pre-trained representations, which are typically temporally agnostic. Previous work on modelling context and temporal dynamics has used recurrent methods, which are slow and prone to overfitting. Here we introduce TempoFormer, the first task-agnostic transformer-based and temporally-aware model for dynamic representation learning. Our approach is jointly trained on inter and intra context dynamics and introduces a novel temporal variation of rotary positional embeddings. The architecture is flexible and can be used as the temporal representation foundation of other models or applied to different transformer-based architectures. We show new SOTA performance on three different real-time change detection tasks.
comment: Accepted by EMNLP Main 2024
♻ ☆ Stacking Your Transformers: A Closer Look at Model Growth for Efficient LLM Pre-Training NeurIPS 2024
LLMs are computationally expensive to pre-train due to their large scale. Model growth emerges as a promising approach by leveraging smaller models to accelerate the training of larger ones. However, the viability of these model growth methods in efficient LLM pre-training remains underexplored. This work identifies three critical $\underline{\textit{O}}$bstacles: ($\textit{O}$1) lack of comprehensive evaluation, ($\textit{O}$2) untested viability for scaling, and ($\textit{O}$3) lack of empirical guidelines. To tackle $\textit{O}$1, we summarize existing approaches into four atomic growth operators and systematically evaluate them in a standardized LLM pre-training setting. Our findings reveal that a depthwise stacking operator, called $G_{\text{stack}}$, exhibits remarkable acceleration in training, leading to decreased loss and improved overall performance on eight standard NLP benchmarks compared to strong baselines. Motivated by these promising results, we conduct extensive experiments to delve deeper into $G_{\text{stack}}$ to address $\textit{O}$2 and $\textit{O}$3. For $\textit{O}$2 (untested scalability), our study shows that $G_{\text{stack}}$ is scalable and consistently performs well, with experiments up to 7B LLMs after growth and pre-training LLMs with 750B tokens. For example, compared to a conventionally trained 7B model using 300B tokens, our $G_{\text{stack}}$ model converges to the same loss with 194B tokens, resulting in a 54.6\% speedup. We further address $\textit{O}$3 (lack of empirical guidelines) by formalizing guidelines to determine growth timing and growth factor for $G_{\text{stack}}$, making it practical in general LLM pre-training. We also provide in-depth discussions and comprehensive ablation studies of $G_{\text{stack}}$. Our code and pre-trained model are available at https://llm-stacking.github.io.
comment: NeurIPS 2024 Spotlight
♻ ☆ Good Parenting is all you need -- Multi-agentic LLM Hallucination Mitigation
This study explores the ability of Large Language Model (LLM) agents to detect and correct hallucinations in AI-generated content. A primary agent was tasked with creating a blog about a fictional Danish artist named Flipfloppidy, which was then reviewed by another agent for factual inaccuracies. Most LLMs hallucinated the existence of this artist. Across 4,900 test runs involving various combinations of primary and reviewing agents, advanced AI models such as Llama3-70b and GPT-4 variants demonstrated near-perfect accuracy in identifying hallucinations and successfully revised outputs in 85% to 100% of cases following feedback. These findings underscore the potential of advanced AI models to significantly enhance the accuracy and reliability of generated content, providing a promising approach to improving AI workflow orchestration.
♻ ☆ CartesianMoE: Boosting Knowledge Sharing among Experts via Cartesian Product Routing in Mixture-of-Experts
Large language models (LLM) have been attracting much attention from the community recently, due to their remarkable performance in all kinds of downstream tasks. According to the well-known scaling law, scaling up a dense LLM enhances its capabilities, but also significantly increases the computational complexity. Mixture-of-Experts (MoE) models address that by allowing the model size to grow without substantially raising training or inference costs. Yet MoE models face challenges regarding knowledge sharing among experts, making their performance somehow sensitive to routing accuracy. To tackle that, previous works introduced shared experts and combined their outputs with those of the top $K$ routed experts in an ``addition'' manner. In this paper, inspired by collective matrix factorization to learn shared knowledge among data, we propose CartesianMoE, which implements more effective knowledge sharing among experts in more like a ``multiplication'' manner. Extensive experimental results indicate that CartesianMoE outperforms previous MoE models for building LLMs, in terms of both perplexity and downstream task performance. And we also find that CartesianMoE achieves better expert routing robustness.
♻ ☆ F-MALLOC: Feed-forward Memory Allocation for Continual Learning in Neural Machine Translation NAACL 2024
In the evolving landscape of Neural Machine Translation (NMT), the pretrain-then-finetune paradigm has yielded impressive results. However, the persistent challenge of Catastrophic Forgetting (CF) remains a hurdle. While previous work has introduced Continual Learning (CL) methods to address CF, these approaches grapple with the delicate balance between avoiding forgetting and maintaining system extensibility. To address this, we propose a CL method, named $\textbf{F-MALLOC}$ ($\textbf{F}$eed-forward $\textbf{M}$emory $\textbf{ALLOC}ation)$. F-MALLOC is inspired by recent insights highlighting that feed-forward layers emulate neural memories and encapsulate crucial translation knowledge. It decomposes feed-forward layers into discrete memory cells and allocates these memories to different tasks. By learning to allocate and safeguard these memories, our method effectively alleviates CF while ensuring robust extendability. Besides, we propose a comprehensive assessment protocol for multi-stage CL of NMT systems. Experiments conducted following this new protocol showcase the superior performance of F-MALLOC, evidenced by higher BLEU scores and almost zero forgetting.
comment: Accepted to the main conference of NAACL 2024
♻ ☆ PLaMo-100B: A Ground-Up Language Model Designed for Japanese Proficiency
We introduce PLaMo-100B, a large-scale language model designed for Japanese proficiency. The model was trained from scratch using 2 trillion tokens, with architecture such as QK Normalization and Z-Loss to ensure training stability during the training process. Post-training techniques, including Supervised Fine-Tuning and Direct Preference Optimization, were applied to refine the model's performance. Benchmark evaluations suggest that PLaMo-100B performs well, particularly in Japanese-specific tasks, achieving results that are competitive with frontier models like GPT-4. The base model is available at https://huggingface.co/pfnet/plamo-100b.
♻ ☆ ERABAL: Enhancing Role-Playing Agents through Boundary-Aware Learning
Role-playing is an emerging application in the field of Human-Computer Interaction (HCI), primarily implemented through the alignment training of a large language model (LLM) with assigned characters. Despite significant progress, role-playing agents (RPLAs) still struggle with maintaining role-consistency across conversations, particularly when confronted with boundary queries subtly related to character attributes. In this paper, we present ERABAL, a framework aimed at enhancing RPLAs' role-playing capabilities through boundary-aware learning. ERABAL encompasses a generation pipeline for role-specific dialogues and a concomitant methodology for alignment training. Through comprehensive evaluations, we demonstrate that ERABAL is both efficient and effective. By training with significantly fewer dialogues than those used in leading approaches, ERABAL achieves notable improvements across WikiRoleEval, CharacterEval, and the role-playing subset of MT-Bench compared to the generalist baseline models. Our code and datasets will be made publicly available to support further research.
comment: arXiv admin note: substantial text overlap with arXiv:2402.10618
♻ ☆ HAF-RM: A Hybrid Alignment Framework for Reward Model Training
The reward model has become increasingly important in alignment, assessment, and data construction for large language models (LLMs). Most existing researchers focus on enhancing reward models through data improvements, following the conventional training framework for reward models that directly optimizes the predicted rewards. In this paper, we propose a hybrid alignment framework HaF-RM for reward model training by introducing an additional constraint on token-level policy probabilities in addition to the reward score. It can simultaneously supervise the internal preference model at the token level and optimize the mapping layer of the reward model at the sequence level. Theoretical justifications and experiment results on five datasets show the validity and effectiveness of our proposed hybrid framework for training a high-quality reward model. By decoupling the reward modeling procedure and incorporating hybrid supervision, our HaF-RM framework offers a principled and effective approach to enhancing the performance and alignment of reward models, a critical component in the responsible development of powerful language models. We release our code at https://haf-rm.github.io.
♻ ☆ ETF: An Entity Tracing Framework for Hallucination Detection in Code Summaries
Recent advancements in large language models (LLMs) have significantly enhanced their ability to understand both natural language and code, driving their use in tasks like natural language-to-code (NL2Code) and code summarization. However, LLMs are prone to hallucination-outputs that stray from intended meanings. Detecting hallucinations in code summarization is especially difficult due to the complex interplay between programming and natural languages. We introduce a first-of-its-kind dataset with $\sim$10K samples, curated specifically for hallucination detection in code summarization. We further propose a novel Entity Tracing Framework (ETF) that a) utilizes static program analysis to identify code entities from the program and b) uses LLMs to map and verify these entities and their intents within generated code summaries. Our experimental analysis demonstrates the effectiveness of the framework, leading to a 0.73 F1 score. This approach provides an interpretable method for detecting hallucinations by grounding entities, allowing us to evaluate summary accuracy.
comment: 11 pages, 6 Figures, 5 Tables
♻ ☆ Language Model Alignment in Multilingual Trolley Problems
We evaluate the moral alignment of large language models (LLMs) with human preferences in multilingual trolley problems. Building on the Moral Machine experiment, which captures over 40 million human judgments across 200+ countries, we develop a cross-lingual corpus of moral dilemma vignettes in over 100 languages called MultiTP. This dataset enables the assessment of LLMs' decision-making processes in diverse linguistic contexts. Our analysis explores the alignment of 19 different LLMs with human judgments, capturing preferences across six moral dimensions: species, gender, fitness, status, age, and the number of lives involved. By correlating these preferences with the demographic distribution of language speakers and examining the consistency of LLM responses to various prompt paraphrasings, our findings provide insights into cross-lingual and ethical biases of LLMs and their intersection. We discover significant variance in alignment across languages, challenging the assumption of uniform moral reasoning in AI systems and highlighting the importance of incorporating diverse perspectives in AI ethics. The results underscore the need for further research on the integration of multilingual dimensions in responsible AI research to ensure fair and equitable AI interactions worldwide. Our code and data are at https://github.com/causalNLP/moralmachine
♻ ☆ UCFE: A User-Centric Financial Expertise Benchmark for Large Language Models
This paper introduces the UCFE: User-Centric Financial Expertise benchmark, an innovative framework designed to evaluate the ability of large language models (LLMs) to handle complex real-world financial tasks. UCFE benchmark adopts a hybrid approach that combines human expert evaluations with dynamic, task-specific interactions to simulate the complexities of evolving financial scenarios. Firstly, we conducted a user study involving 804 participants, collecting their feedback on financial tasks. Secondly, based on this feedback, we created our dataset that encompasses a wide range of user intents and interactions. This dataset serves as the foundation for benchmarking 12 LLM services using the LLM-as-Judge methodology. Our results show a significant alignment between benchmark scores and human preferences, with a Pearson correlation coefficient of 0.78, confirming the effectiveness of the UCFE dataset and our evaluation approach. UCFE benchmark not only reveals the potential of LLMs in the financial sector but also provides a robust framework for assessing their performance and user satisfaction. The benchmark dataset and evaluation code are available.
♻ ☆ COMMUNITY-CROSS-INSTRUCT: Unsupervised Instruction Generation for Aligning Large Language Models to Online Communities
Social scientists use surveys to probe the opinions and beliefs of populations, but these methods are slow, costly, and prone to biases. Recent advances in large language models (LLMs) enable the creating of computational representations or "digital twins" of populations that generate human-like responses mimicking the population's language, styles, and attitudes. We introduce Community-Cross-Instruct, an unsupervised framework for aligning LLMs to online communities to elicit their beliefs. Given a corpus of a community's online discussions, Community-Cross-Instruct automatically generates instruction-output pairs by an advanced LLM to (1) finetune a foundational LLM to faithfully represent that community, and (2) evaluate the alignment of the finetuned model to the community. We demonstrate the method's utility in accurately representing political and diet communities on Reddit. Unlike prior methods requiring human-authored instructions, Community-Cross-Instruct generates instructions in a fully unsupervised manner, enhancing scalability and generalization across domains. This work enables cost-effective and automated surveying of diverse online communities.
♻ ☆ A Text is Worth Several Tokens: Text Embedding from LLMs Secretly Aligns Well with The Key Tokens
Text embeddings from large language models (LLMs) have achieved excellent results in tasks such as information retrieval, semantic textual similarity, etc. In this work, we show an interesting finding: when feeding a text into the embedding LLMs, the obtained text embedding will be able to be aligned with the key tokens in the input text. We first fully analyze this phenomenon on eight embedding LLMs and show that this phenomenon is universal and is not affected by model architecture, training strategy, and embedding method. With a deeper analysis, we then find that the main change in embedding space between the embedding LLMs and their original generative LLMs is in the first principal component. By adjusting the first principal component, we can align text embedding with the key tokens. Finally, we give several examples to demonstrate the vast application potential of this finding: (1) we propose a simple and practical sparse retrieval method based on the aligned tokens, which can achieve 80\% of the dense retrieval effect of the same model while reducing the computation significantly; (2) we show that our findings provide a fresh perspective to help understand fuzzy concepts (e.g., semantic relatedness vs. semantic similarity) and emerging technologies (e.g., instruction-following embedding) in this field.
comment: Work in Progress
♻ ☆ Improving Reinforcement Learning from Human Feedback with Efficient Reward Model Ensemble
Reinforcement Learning from Human Feedback (RLHF) is a widely adopted approach for aligning large language models with human values. However, RLHF relies on a reward model that is trained with a limited amount of human preference data, which could lead to inaccurate predictions. As a result, RLHF may produce outputs that are misaligned with human values. To mitigate this issue, we contribute a reward ensemble method that allows the reward model to make more accurate predictions. As using an ensemble of large language model-based reward models can be computationally and resource-expensive, we explore efficient ensemble methods including linear-layer ensemble and LoRA-based ensemble. Empirically, we run Best-of-$n$ and Proximal Policy Optimization with our ensembled reward models, and verify that our ensemble methods help improve the alignment performance of RLHF outputs.
♻ ☆ Position Engineering: Boosting Large Language Models through Positional Information Manipulation
The performance of large language models (LLMs) is significantly influenced by the quality of the prompts provided. In response, researchers have developed enormous prompt engineering strategies aimed at modifying the prompt text to enhance task performance. In this paper, we introduce a novel technique termed position engineering, which offers a more efficient way to guide large language models. Unlike prompt engineering, which requires substantial effort to modify the text provided to LLMs, position engineering merely involves altering the positional information in the prompt without modifying the text itself. We have evaluated position engineering in two widely-used LLM scenarios: retrieval-augmented generation (RAG) and in-context learning (ICL). Our findings show that position engineering substantially improves upon the baseline in both cases. Position engineering thus represents a promising new strategy for exploiting the capabilities of large language models.
♻ ☆ PHAnToM: Persona-based Prompting Has An Effect on Theory-of-Mind Reasoning in Large Language Models
The use of LLMs in natural language reasoning has shown mixed results, sometimes rivaling or even surpassing human performance in simpler classification tasks while struggling with social-cognitive reasoning, a domain where humans naturally excel. These differences have been attributed to many factors, such as variations in prompting and the specific LLMs used. However, no reasons appear conclusive, and no clear mechanisms have been established in prior work. In this study, we empirically evaluate how role-playing prompting influences Theory-of-Mind (ToM) reasoning capabilities. Grounding our rsearch in psychological theory, we propose the mechanism that, beyond the inherent variance in the complexity of reasoning tasks, performance differences arise because of socially-motivated prompting differences. In an era where prompt engineering with role-play is a typical approach to adapt LLMs to new contexts, our research advocates caution as models that adopt specific personas might potentially result in errors in social-cognitive reasoning.
comment: 8 pages
♻ ☆ Evaluating Human Alignment and Model Faithfulness of LLM Rationale
We study how well large language models (LLMs) explain their generations through rationales -- a set of tokens extracted from the input text that reflect the decision-making process of LLMs. Specifically, we systematically study rationales derived using two approaches: (1) popular prompting-based methods, where prompts are used to guide LLMs in generating rationales, and (2) technical attribution-based methods, which leverage attention or gradients to identify important tokens. Our analysis spans three classification datasets with annotated rationales, encompassing tasks with varying performance levels. While prompting-based self-explanations are widely used, our study reveals that these explanations are not always as "aligned" with the human rationale as attribution-based explanations. Even more so, fine-tuning LLMs to enhance classification task accuracy does not enhance the alignment of prompting-based rationales. Still, it does considerably improve the alignment of attribution-based methods (e.g., InputXGradient). More importantly, we show that prompting-based self-explanation is also less "faithful" than attribution-based explanations, failing to provide a reliable account of the model's decision-making process. To evaluate faithfulness, unlike prior studies that excluded misclassified examples, we evaluate all instances and also examine the impact of fine-tuning and accuracy on alignment and faithfulness. Our findings suggest that inconclusive faithfulness results reported in earlier studies may stem from low classification accuracy. These findings underscore the importance of more rigorous and comprehensive evaluations of LLM rationales.
♻ ☆ DNABERT-S: Pioneering Species Differentiation with Species-Aware DNA Embeddings
We introduce DNABERT-S, a tailored genome model that develops species-aware embeddings to naturally cluster and segregate DNA sequences of different species in the embedding space. Differentiating species from genomic sequences (i.e., DNA and RNA) is vital yet challenging, since many real-world species remain uncharacterized, lacking known genomes for reference. Embedding-based methods are therefore used to differentiate species in an unsupervised manner. DNABERT-S builds upon a pre-trained genome foundation model named DNABERT-2. To encourage effective embeddings to error-prone long-read DNA sequences, we introduce Manifold Instance Mixup (MI-Mix), a contrastive objective that mixes the hidden representations of DNA sequences at randomly selected layers and trains the model to recognize and differentiate these mixed proportions at the output layer. We further enhance it with the proposed Curriculum Contrastive Learning (C$^2$LR) strategy. Empirical results on 23 diverse datasets show DNABERT-S's effectiveness, especially in realistic label-scarce scenarios. For example, it identifies twice more species from a mixture of unlabeled genomic sequences, doubles the Adjusted Rand Index (ARI) in species clustering, and outperforms the top baseline's performance in 10-shot species classification with just a 2-shot training. Model, codes, and data are publicly available at \url{https://github.com/MAGICS-LAB/DNABERT_S}.
♻ ☆ LLM4Decompile: Decompiling Binary Code with Large Language Models
Decompilation aims to convert binary code to high-level source code, but traditional tools like Ghidra often produce results that are difficult to read and execute. Motivated by the advancements in Large Language Models (LLMs), we propose LLM4Decompile, the first and largest open-source LLM series (1.3B to 33B) trained to decompile binary code. We optimize the LLM training process and introduce the LLM4Decompile-End models to decompile binary directly. The resulting models significantly outperform GPT-4o and Ghidra on the HumanEval and ExeBench benchmarks by over 100% in terms of re-executability rate. Additionally, we improve the standard refinement approach to fine-tune the LLM4Decompile-Ref models, enabling them to effectively refine the decompiled code from Ghidra and achieve a further 16.2% improvement over the LLM4Decompile-End. LLM4Decompile demonstrates the potential of LLMs to revolutionize binary code decompilation, delivering remarkable improvements in readability and executability while complementing conventional tools for optimal results. Our code, dataset, and models are released at https://github.com/albertan017/LLM4Decompile
♻ ☆ From Text to Multimodality: Exploring the Evolution and Impact of Large Language Models in Medical Practice
Large Language Models (LLMs) have rapidly evolved from text-based systems to multimodal platforms, significantly impacting various sectors including healthcare. This comprehensive review explores the progression of LLMs to Multimodal Large Language Models (MLLMs) and their growing influence in medical practice. We examine the current landscape of MLLMs in healthcare, analyzing their applications across clinical decision support, medical imaging, patient engagement, and research. The review highlights the unique capabilities of MLLMs in integrating diverse data types, such as text, images, and audio, to provide more comprehensive insights into patient health. We also address the challenges facing MLLM implementation, including data limitations, technical hurdles, and ethical considerations. By identifying key research gaps, this paper aims to guide future investigations in areas such as dataset development, modality alignment methods, and the establishment of ethical guidelines. As MLLMs continue to shape the future of healthcare, understanding their potential and limitations is crucial for their responsible and effective integration into medical practice.
comment: 12 pages, 1 figure
♻ ☆ Can Large Language Models Invent Algorithms to Improve Themselves?
Large Language Models (LLMs) have shown remarkable performance improvements and are rapidly gaining adoption in industry. However, the methods for improving LLMs are still designed by humans, which restricts the invention of new model-improving algorithms to human expertise and imagination. To address this, we propose the Self-Developing framework, which enables LLMs to autonomously generate and learn model-improvement algorithms. In this framework, the seed model generates, applies, and learns model-improving algorithms, continuously improving both the seed model and the algorithms themselves. In mathematical reasoning tasks, Self-Developing not only creates models that surpass the seed model but also consistently outperforms models created using human-designed algorithms. Additionally, these LLM-discovered algorithms demonstrate strong effectiveness, including transferability to out-of-domain models.
♻ ☆ Merging LoRAs like Playing LEGO: Pushing the Modularity of LoRA to Extremes Through Rank-Wise Clustering
Low-Rank Adaptation (LoRA) has emerged as a popular technique for fine-tuning large language models (LLMs) to various domains due to its modular design and widespread availability on platforms like Huggingface. This modularity has sparked interest in combining multiple LoRAs to enhance LLM capabilities. However, existing methods for LoRA composition primarily focus on task-specific adaptations that require additional training, and current model merging techniques often fail to fully leverage LoRA's modular nature, leading to parameter interference and performance degradation. In this paper, we investigate the feasibility of disassembling and reassembling multiple LoRAs at a finer granularity, analogous to assembling LEGO blocks. We introduce the concept of Minimal Semantic Units (MSUs), where the parameters corresponding to each rank in LoRA function as independent units. These MSUs demonstrate permutation invariance and concatenation-summation equivalence properties, enabling flexible combinations to create new LoRAs. Building on these insights, we propose the LoRA-LEGO framework. This framework conducts rank-wise parameter clustering by grouping MSUs from different LoRAs into $k$ clusters. The centroid of each cluster serves as a representative MSU, enabling the assembly of a merged LoRA with an adjusted rank of $k$. Additionally, we apply a dual reweighting strategy to optimize the scale of the merged LoRA. Experiments across various benchmarks demonstrate that our method outperforms existing approaches in LoRA merging.
♻ ☆ Adaptable and Reliable Text Classification using Large Language Models ICDM
Text classification is fundamental in Natural Language Processing (NLP), and the advent of Large Language Models (LLMs) has revolutionized the field. This paper introduces an adaptable and reliable text classification paradigm, which leverages LLMs as the core component to address text classification tasks. Our system simplifies the traditional text classification workflows, reducing the need for extensive preprocessing and domain-specific expertise to deliver adaptable and reliable text classification results. We evaluated the performance of several LLMs, machine learning algorithms, and neural network-based architectures on four diverse datasets. Results demonstrate that certain LLMs surpass traditional methods in sentiment analysis, spam SMS detection, and multi-label classification. Furthermore, it is shown that the system's performance can be further enhanced through few-shot or fine-tuning strategies, making the fine-tuned model the top performer across all datasets. Source code and datasets are available in this GitHub repository: https://github.com/yeyimilk/llm-zero-shot-classifiers.
comment: ICDM Workshop ARRL 2024
♻ ☆ Zero-Shot Cross-Lingual NER Using Phonemic Representations for Low-Resource Languages EMNLP 2024
Existing zero-shot cross-lingual NER approaches require substantial prior knowledge of the target language, which is impractical for low-resource languages. In this paper, we propose a novel approach to NER using phonemic representation based on the International Phonetic Alphabet (IPA) to bridge the gap between representations of different languages. Our experiments show that our method significantly outperforms baseline models in extremely low-resource languages, with the highest average F1 score (46.38%) and lowest standard deviation (12.67), particularly demonstrating its robustness with non-Latin scripts. Our codes are available at https://github.com/Gabriel819/zeroshot_ner.git
comment: Accepted to EMNLP 2024 Main Conference
♻ ☆ Subversive Characters and Stereotyping Readers: Characterizing Queer Relationalities with Dialogue-Based Relation Extraction
Television is often seen as a site for subcultural identification and subversive fantasy, including in queer cultures. How might we measure subversion, or the degree to which the depiction of social relationship between a dyad (e.g. two characters who are colleagues) deviates from its typical representation on TV? To explore this question, we introduce the task of stereotypic relationship extraction. Built on cognitive stylistics, linguistic anthropology, and dialogue relation extraction, in this paper, we attempt to model the cognitive process of stereotyping TV characters in dialogic interactions. Given a dyad, we want to predict: what social relationship do the speakers exhibit through their words? Subversion is then characterized by the discrepancy between the distribution of the model's predictions and the ground truth labels. To demonstrate the usefulness of this task and gesture at a methodological intervention, we enclose four case studies to characterize the representation of queer relationalities in the Big Bang Theory, Frasier, and Gilmore Girls, as we explore the suspicious and reparative modes of reading with our computational methods.
comment: CHR 2024: Computational Humanities Research Conference
♻ ☆ Debiasing Text Safety Classifiers through a Fairness-Aware Ensemble
Increasing use of large language models (LLMs) demand performant guardrails to ensure the safety of inputs and outputs of LLMs. When these safeguards are trained on imbalanced data, they can learn the societal biases. We present a light-weight, post-processing method for mitigating counterfactual fairness in closed-source text safety classifiers. Our approach involves building an ensemble that not only outperforms the input classifiers and policy-aligns them, but also acts as a debiasing regularizer. We introduce two threshold-agnostic metrics to assess the counterfactual fairness of a model, and demonstrate how combining these metrics with Fair Data Reweighting (FDW) helps mitigate biases. We create an expanded Open AI dataset, and a new templated LLM-generated dataset based on user-prompts, both of which are counterfactually balanced across identity groups and cover four key areas of safety; we will work towards publicly releasing these datasets. Our results show that our approach improves counterfactual fairness with minimal impact on model performance.
♻ ☆ Improving the Language Understanding Capabilities of Large Language Models Using Reinforcement Learning
Large language models (LLMs), built on decoder-only transformers, excel in natural language generation and adapt to diverse tasks using zero-shot and few-shot prompting. However, these prompting methods often struggle on natural language understanding (NLU) tasks, where encoder-only models like BERT-base outperform LLMs on benchmarks like GLUE and SuperGLUE. This paper explores two approaches-supervised fine-tuning (SFT) and proximal policy optimization (PPO)-to enhance LLMs' NLU abilities. To reduce the cost of full-model fine-tuning, we integrate low-rank adaptation (LoRA) layers, limiting updates to these layers during both SFT and PPO. In SFT, task-specific prompts are concatenated with input queries and ground-truth labels, optimizing with next-token prediction. Despite this, LLMs still underperform compared to models like BERT-base on several NLU tasks. To close this gap, we apply PPO, a reinforcement learning technique that treats each token generation as an action and uses a reward function based on alignment with ground-truth answers. PPO then updates the model to maximize these rewards, aligning outputs with correct labels. Our experiments with LLAMA2-7B show that PPO improves performance, with a 6.3-point gain over SFT on GLUE. PPO exceeds zero-shot by 38.7 points and few-shot by 26.1 points on GLUE, while surpassing these by 28.8 and 28.5 points on SuperGLUE. Additionally, PPO outperforms BERT-large by 2.7 points on GLUE and 9.3 points on SuperGLUE. The improvements are consistent across models like Qwen2.5-7B and MPT-7B, highlighting PPO's robustness in enhancing LLMs' NLU capabilities.
♻ ☆ An Empirical Comparison of Vocabulary Expansion and Initialization Approaches for Language Models EMNLP 2024
Language Models (LMs) excel in natural language processing tasks for English but show reduced performance in most other languages. This problem is commonly tackled by continually pre-training and fine-tuning these models for said languages. A significant issue in this process is the limited vocabulary coverage in the original model's tokenizer, leading to inadequate representation of new languages and necessitating an expansion of the tokenizer. The initialization of the embeddings corresponding to new vocabulary items presents a further challenge. Current strategies require cross-lingual embeddings and lack a solid theoretical foundation as well as comparisons with strong baselines. In this paper, we first establish theoretically that initializing within the convex hull of existing embeddings is a good initialization, followed by a novel but simple approach, Constrained Word2Vec (CW2V), which does not require cross-lingual embeddings. Our study evaluates different initialization methods for expanding RoBERTa and LLaMA 2 across four languages and five tasks. The results show that CW2V performs equally well or even better than more advanced techniques. Additionally, simpler approaches like multivariate initialization perform on par with these advanced methods indicating that efficient large-scale multilingual continued pretraining can be achieved even with simpler initialization methods. We release our code publicly (https://github.com/AI4Bharat/VocabAdaptation_LLM/tree/CW2V).
comment: CONLL 2024 (EMNLP 2024)
♻ ☆ Are Large Language Models Moral Hypocrites? A Study Based on Moral Foundations
Large language models (LLMs) have taken centre stage in debates on Artificial Intelligence. Yet there remains a gap in how to assess LLMs' conformity to important human values. In this paper, we investigate whether state-of-the-art LLMs, GPT-4 and Claude 2.1 (Gemini Pro and LLAMA 2 did not generate valid results) are moral hypocrites. We employ two research instruments based on the Moral Foundations Theory: (i) the Moral Foundations Questionnaire (MFQ), which investigates which values are considered morally relevant in abstract moral judgements; and (ii) the Moral Foundations Vignettes (MFVs), which evaluate moral cognition in concrete scenarios related to each moral foundation. We characterise conflicts in values between these different abstractions of moral evaluation as hypocrisy. We found that both models displayed reasonable consistency within each instrument compared to humans, but they displayed contradictory and hypocritical behaviour when we compared the abstract values present in the MFQ to the evaluation of concrete moral violations of the MFV.
comment: Final version available at: https://ojs.aaai.org/index.php/AIES/article/view/31704 13 pages, 4 figures, 2 tables
♻ ☆ NVLM: Open Frontier-Class Multimodal LLMs
We introduce NVLM 1.0, a family of frontier-class multimodal large language models (LLMs) that achieve state-of-the-art results on vision-language tasks, rivaling the leading proprietary models (e.g., GPT-4o) and open-access models (e.g., Llama 3-V 405B and InternVL 2). Remarkably, NVLM 1.0 shows improved text-only performance over its LLM backbone after multimodal training. In terms of model design, we perform a comprehensive comparison between decoder-only multimodal LLMs (e.g., LLaVA) and cross-attention-based models (e.g., Flamingo). Based on the strengths and weaknesses of both approaches, we propose a novel architecture that enhances both training efficiency and multimodal reasoning capabilities. Furthermore, we introduce a 1-D tile-tagging design for tile-based dynamic high-resolution images, which significantly boosts performance on multimodal reasoning and OCR-related tasks. Regarding training data, we meticulously curate and provide detailed information on our multimodal pretraining and supervised fine-tuning datasets. Our findings indicate that dataset quality and task diversity are more important than scale, even during the pretraining phase, across all architectures. Notably, we develop production-grade multimodality for the NVLM-1.0 models, enabling them to excel in vision-language tasks while maintaining and even improving text-only performance compared to their LLM backbones. To achieve this, we craft and integrate a high-quality text-only dataset into multimodal training, alongside a substantial amount of multimodal math and reasoning data, leading to enhanced math and coding capabilities across modalities. To advance research in the field, we release the model weights at https://huggingface.co/nvidia/NVLM-D-72B and will open-source the training code for the community soon.
comment: Fixed the typos. For more information, please visit our project page at: https://research.nvidia.com/labs/adlr/NVLM-1
♻ ☆ How to Evaluate Reward Models for RLHF
We introduce a new benchmark for reward models that quantifies their ability to produce strong language models through RLHF (Reinforcement Learning from Human Feedback). The gold-standard approach is to run a full RLHF training pipeline and directly probe downstream LLM performance. However, this process is prohibitively expensive. To address this, we build a predictive model of downstream LLM performance by evaluating the reward model on proxy tasks. These proxy tasks consist of a large-scale human preference and a verifiable correctness preference dataset, in which we measure 12 metrics across 12 domains. To investigate which reward model metrics are most correlated to gold-standard RLHF outcomes, we launch an end-to-end RLHF experiment on a large-scale crowdsourced human preference platform to view real reward model downstream performance as ground truth. Ultimately, we compile our data and findings into Preference Proxy Evaluations (PPE), the first reward model benchmark explicitly linked to post-RLHF real-world human preference performance, which we open-source for public use and further development. Our code and evaluations can be found at https://github.com/lmarena/PPE .
♻ ☆ A Comparative Study on Reasoning Patterns of OpenAI's o1 Model
Enabling Large Language Models (LLMs) to handle a wider range of complex tasks (e.g., coding, math) has drawn great attention from many researchers. As LLMs continue to evolve, merely increasing the number of model parameters yields diminishing performance improvements and heavy computational costs. Recently, OpenAI's o1 model has shown that inference strategies (i.e., Test-time Compute methods) can also significantly enhance the reasoning capabilities of LLMs. However, the mechanisms behind these methods are still unexplored. In our work, to investigate the reasoning patterns of o1, we compare o1 with existing Test-time Compute methods (BoN, Step-wise BoN, Agent Workflow, and Self-Refine) by using OpenAI's GPT-4o as a backbone on general reasoning benchmarks in three domains (i.e., math, coding, commonsense reasoning). Specifically, first, our experiments show that the o1 model has achieved the best performance on most datasets. Second, as for the methods of searching diverse responses (e.g., BoN), we find the reward models' capability and the search space both limit the upper boundary of these methods. Third, as for the methods that break the problem into many sub-problems, the Agent Workflow has achieved better performance than Step-wise BoN due to the domain-specific system prompt for planning better reasoning processes. Fourth, it is worth mentioning that we have summarized six reasoning patterns of o1, and provided a detailed analysis on several reasoning benchmarks.
♻ ☆ NutriBench: A Dataset for Evaluating Large Language Models in Carbohydrate Estimation from Meal Descriptions
Accurate nutrition estimation helps people make informed dietary choices and is essential in the prevention of serious health complications. We present NutriBench, the first publicly available natural language meal description nutrition benchmark. NutriBench consists of 11,857 meal descriptions generated from real-world global dietary intake data. The data is human-verified and annotated with macro-nutrient labels, including carbohydrates, proteins, fats, and calories. We conduct an extensive evaluation of NutriBench on the task of carbohydrate estimation, testing twelve leading Large Language Models (LLMs), including GPT-4o, Llama3.1, Qwen2, Gemma2, and OpenBioLLM models, using standard, Chain-of-Thought and Retrieval-Augmented Generation strategies. Additionally, we present a study involving professional nutritionists, finding that LLMs can provide more accurate and faster estimates. Finally, we perform a real-world risk assessment by simulating the effect of carbohydrate predictions on the blood glucose levels of individuals with diabetes. Our work highlights the opportunities and challenges of using LLMs for nutrition estimation, demonstrating their potential to aid professionals and laypersons and improve health outcomes. Our benchmark is publicly available at: https://mehak126.github.io/nutribench.html
♻ ☆ Characterizing the Accuracy -- Efficiency Trade-off of Low-rank Decomposition in Language Models
Recent large language models (LLMs) employ billions of parameters to enable broad problem-solving capabilities. Such language models also tend to be memory-bound because of the dominance of matrix-vector and matrix-matrix multiplications with low arithmetic intensity. Therefore, optimizing the memory footprint and traffic is an important optimization direction for LLMs today. Model compression methods such as quantization and parameter pruning have been actively explored to achieve memory footprint and traffic optimization. However, the accuracy-efficiency trade-off of rank pruning (i.e., low-rank decomposition) for LLMs is not well-understood yet. Therefore, in this work, we characterize the accuracy-efficiency trade-off of a low-rank decomposition method, specifically Tucker decomposition, on recent language models, including an open-source LLM, Llama 2. We formalize the low-rank decomposition design space and show that the decomposition design space is enormous (e.g., O($2^{39}$) for Llama2-7B). To navigate such a vast design space, we formulate it and perform thorough case studies of accuracy-efficiency trade-offs using six widely used LLM benchmarks on BERT and Llama 2 models. Our results show that we can achieve a 9\% model size reduction with minimal accuracy drops, which range from 4\%p (\%p refers to "percentage point," which refers to the absolute difference between two percentage numbers; 74\% -> 78\% = 4\%p increase) to 10\%p, depending on the difficulty of the benchmark, without any retraining to recover accuracy after decomposition. The results show that low-rank decomposition can be a promising direction for LLM-based applications that require real-time service at scale (e.g., AI agent and real-time coding assistant), where the latency is as important as the model accuracy.
♻ ☆ LLM Processes: Numerical Predictive Distributions Conditioned on Natural Language
Machine learning practitioners often face significant challenges in formally integrating their prior knowledge and beliefs into predictive models, limiting the potential for nuanced and context-aware analyses. Moreover, the expertise needed to integrate this prior knowledge into probabilistic modeling typically limits the application of these models to specialists. Our goal is to build a regression model that can process numerical data and make probabilistic predictions at arbitrary locations, guided by natural language text which describes a user's prior knowledge. Large Language Models (LLMs) provide a useful starting point for designing such a tool since they 1) provide an interface where users can incorporate expert insights in natural language and 2) provide an opportunity for leveraging latent problem-relevant knowledge encoded in LLMs that users may not have themselves. We start by exploring strategies for eliciting explicit, coherent numerical predictive distributions from LLMs. We examine these joint predictive distributions, which we call LLM Processes, over arbitrarily-many quantities in settings such as forecasting, multi-dimensional regression, black-box optimization, and image modeling. We investigate the practical details of prompting to elicit coherent predictive distributions, and demonstrate their effectiveness at regression. Finally, we demonstrate the ability to usefully incorporate text into numerical predictions, improving predictive performance and giving quantitative structure that reflects qualitative descriptions. This lets us begin to explore the rich, grounded hypothesis space that LLMs implicitly encode.
♻ ☆ Token-wise Influential Training Data Retrieval for Large Language Models ACL 2024
Given a Large Language Model (LLM) generation, how can we identify which training data led to this generation? In this paper, we proposed RapidIn, a scalable framework adapting to LLMs for estimating the influence of each training data. The proposed framework consists of two stages: caching and retrieval. First, we compress the gradient vectors by over 200,000x, allowing them to be cached on disk or in GPU/CPU memory. Then, given a generation, RapidIn efficiently traverses the cached gradients to estimate the influence within minutes, achieving over a 6,326x speedup. Moreover, RapidIn supports multi-GPU parallelization to substantially accelerate caching and retrieval. Our empirical result confirms the efficiency and effectiveness of RapidIn.
comment: Accepted to ACL 2024. Keywords: Influence Function, Influence Estimation, Training Data Attribution
♻ ☆ Learning to Poison Large Language Models During Instruction Tuning
The advent of Large Language Models (LLMs) has marked significant achievements in language processing and reasoning capabilities. Despite their advancements, LLMs face vulnerabilities to data poisoning attacks, where adversaries insert backdoor triggers into training data to manipulate outputs for malicious purposes. This work further identifies additional security risks in LLMs by designing a new data poisoning attack tailored to exploit the instruction tuning process. We propose a novel gradient-guided backdoor trigger learning (GBTL) algorithm to identify adversarial triggers efficiently, ensuring an evasion of detection by conventional defenses while maintaining content integrity. Through experimental validation across various tasks, including sentiment analysis, domain generation, and question answering, our poisoning strategy demonstrates a high success rate in compromising various LLMs' outputs. We further propose two defense strategies against data poisoning attacks, including in-context learning (ICL) and continuous learning (CL), which effectively rectify the behavior of LLMs and significantly reduce the decline in performance. Our work highlights the significant security risks present during the instruction tuning of LLMs and emphasizes the necessity of safeguarding LLMs against data poisoning attacks.
♻ ☆ On the Diversity of Synthetic Data and its Impact on Training Large Language Models
The rise of Large Language Models (LLMs) has accentuated the need for diverse, high-quality pre-training data. Synthetic data emerges as a viable solution to the challenges of data scarcity and inaccessibility. While previous literature has focused predominantly on the quality and quantity of real data, our work enables the measurement of diversity in synthetic data and explores its impact on LLM performance. We study the downstream effects of synthetic data diversity during both the pre-training and fine-tuning stages by introducing a new diversity metric, \textit{LLM cluster-agent}, designed to evaluate the diversity of synthetic datasets. Through a series of controlled experiments with models of 350M and 1.4B parameters, we demonstrate that the proposed cluster-based LLM scoring of diversity correlates positively with both pre-training and supervised fine-tuning performance. Our findings also reveal that synthetic data diversity in pre-training affects supervised fine-tuning more significantly than pre-training itself, even for smaller models. We hope this study advances our understanding of the optimal use of synthetic data in LLM training and opens new avenues for efficient data generation processes.
♻ ☆ CDQuant: Greedy Coordinate Descent for Accurate LLM Quantization
Large language models (LLMs) have recently demonstrated remarkable performance across diverse language tasks. But their deployment is often constrained by their substantial computational and storage requirements. Quantization has emerged as a key technique for addressing this challenge, enabling the compression of large models with minimal impact on performance. The recent GPTQ algorithm, a post-training quantization (PTQ) method, has proven highly effective for compressing LLMs, sparking a wave of research that leverages GPTQ as a core component. Recognizing the pivotal role of GPTQ in the PTQ landscape, we introduce CDQuant, a simple and scalable alternative to GPTQ with improved performance. CDQuant uses greedy coordinate descent to minimize the layer-wise reconstruction loss to achieve high-quality quantized weights. Our algorithm is easy to implement and scales efficiently to models with hundreds of billions of parameters. We perform extensive evaluation on Gemma, and PaLM2 model families, and demonstrate that CDQuant consistently outperforms GPTQ in 2-4 bit weight quantization. Moreover, CDQuant improves the performance of state-of-the-art PTQ techniques such as QuIP and FrameQuant when used as a replacement for their GPTQ component, resulting in further gains in quality.
♻ ☆ S2-Attention: Hardware-Aware Context Sharding Among Attention Heads
Sparse attention, which selectively attends to a subset of tokens in the context was supposed to be efficient. However, its theoretical reduction in FLOPs has rarely translated into wall-clock speed-up over its dense attention counterparts due to the lack of hardware-aware optimizations like FlashAttention. Meanwhile, it remains unclear whether sparse attention can maintain the model's quality at a scale of today's large language models (LLMs) and how. This paper presents Sparsely-Sharded(S2) Attention, a Triton library that provides kernel optimization for sparse attention customizable at both per-head and per-context-range levels. S2-Attention enables the exploration of novel and high-performance sparse attention techniques, which we demonstrate through extensive ablations across a wide range of sparse attention designs at various model scales. From these insights, we present several basic guidelines to design sparse attention that can achieve not only practical efficiency improvements, but also strong downstream performance. To achieve high parallelization and optimized memory IO, sparse attention should shard the context heterogeneously across attention heads, where each head attends to a different subset of tokens while collectively covering the full context. Meanwhile, we find hybrid architectures combining sparse and dense attention particularly beneficial in practice. S2-Attention achieves wall-clock speedup of 8.79X, 15.87X, 25.3X compared to the strong FlashAttention-2 baseline with strong downstream performance on-par with full attention and perfect retrieval performance at a 128k context length. At inference, for 7B models, our model, with the help of our S2-Attention kernel, achieves 4.5x speed-up compared to dense counterparts. S2-Attention is released with easy-to-customize APIs for direct usage in Megatron and vLLM.
comment: 10 pages
♻ ☆ FLAG: Financial Long Document Classification via AMR-based GNN
The advent of large language models (LLMs) has initiated much research into their various financial applications. However, in applying LLMs on long documents, semantic relations are not explicitly incorporated, and a full or arbitrarily sparse attention operation is employed. In recent years, progress has been made in Abstract Meaning Representation (AMR), which is a graph-based representation of text to preserve its semantic relations. Since AMR can represent semantic relationships at a deeper level, it can be beneficially utilized by graph neural networks (GNNs) for constructing effective document-level graph representations built upon LLM embeddings to predict target metrics in the financial domain. We propose FLAG: Financial Long document classification via AMR-based GNN, an AMR graph based framework to generate document-level embeddings for long financial document classification. We construct document-level graphs from sentence-level AMR graphs, endow them with specialized LLM word embeddings in the financial domain, apply a deep learning mechanism that utilizes a GNN, and examine the efficacy of our AMR-based approach in predicting labeled target data from long financial documents. Extensive experiments are conducted on a dataset of quarterly earnings calls transcripts of companies in various sectors of the economy, as well as on a corpus of more recent earnings calls of companies in the S&P 1500 Composite Index. We find that our AMR-based approach outperforms fine-tuning LLMs directly on text in predicting stock price movement trends at different time horizons in both datasets. Our work also outperforms previous work utilizing document graphs and GNNs for text classification.
comment: 8 pages, 3 figures, to be published in CIFEr Conference 2024 as "Semantic Graph Learning for Trend Prediction from Long Financial Documents"
♻ ☆ The Causal Influence of Grammatical Gender on Distributional Semantics
How much meaning influences gender assignment across languages is an active area of research in linguistics and cognitive science. We can view current approaches as aiming to determine where gender assignment falls on a spectrum, from being fully arbitrarily determined to being largely semantically determined. For the latter case, there is a formulation of the neo-Whorfian hypothesis, which claims that even inanimate noun gender influences how people conceive of and talk about objects (using the choice of adjective used to modify inanimate nouns as a proxy for meaning). We offer a novel, causal graphical model that jointly represents the interactions between a noun's grammatical gender, its meaning, and adjective choice. In accordance with past results, we find a significant relationship between the gender of nouns and the adjectives that modify them. However, when we control for the meaning of the noun, the relationship between grammatical gender and adjective choice is near zero and insignificant.
Computer Vision and Pattern Recognition 141
☆ Altogether: Image Captioning via Re-aligning Alt-text EMNLP 2024
This paper focuses on creating synthetic data to improve the quality of image captions. Existing works typically have two shortcomings. First, they caption images from scratch, ignoring existing alt-text metadata, and second, lack transparency if the captioners' training data (e.g. GPT) is unknown. In this paper, we study a principled approach Altogether based on the key idea to edit and re-align existing alt-texts associated with the images. To generate training data, we perform human annotation where annotators start with the existing alt-text and re-align it to the image content in multiple rounds, consequently constructing captions with rich visual concepts. This differs from prior work that carries out human annotation as a one-time description task solely based on images and annotator knowledge. We train a captioner on this data that generalizes the process of re-aligning alt-texts at scale. Our results show our Altogether approach leads to richer image captions that also improve text-to-image generation and zero-shot image classification tasks.
comment: accepted by EMNLP 2024; MetaCLIPv2
☆ SpectroMotion: Dynamic 3D Reconstruction of Specular Scenes
We present SpectroMotion, a novel approach that combines 3D Gaussian Splatting (3DGS) with physically-based rendering (PBR) and deformation fields to reconstruct dynamic specular scenes. Previous methods extending 3DGS to model dynamic scenes have struggled to accurately represent specular surfaces. Our method addresses this limitation by introducing a residual correction technique for accurate surface normal computation during deformation, complemented by a deformable environment map that adapts to time-varying lighting conditions. We implement a coarse-to-fine training strategy that significantly enhances both scene geometry and specular color prediction. We demonstrate that our model outperforms prior methods for view synthesis of scenes containing dynamic specular objects and that it is the only existing 3DGS method capable of synthesizing photorealistic real-world dynamic specular scenes, outperforming state-of-the-art methods in rendering complex, dynamic, and specular scenes.
comment: Project page: https://cdfan0627.github.io/spectromotion/
☆ JMMMU: A Japanese Massive Multi-discipline Multimodal Understanding Benchmark for Culture-aware Evaluation
Accelerating research on Large Multimodal Models (LMMs) in non-English languages is crucial for enhancing user experiences across broader populations. In this paper, we introduce JMMMU (Japanese MMMU), the first large-scale Japanese benchmark designed to evaluate LMMs on expert-level tasks based on the Japanese cultural context. To facilitate comprehensive culture-aware evaluation, JMMMU features two complementary subsets: (i) culture-agnostic (CA) subset, where the culture-independent subjects (e.g., Math) are selected and translated into Japanese, enabling one-to-one comparison with its English counterpart MMMU; and (ii) culture-specific (CS) subset, comprising newly crafted subjects that reflect Japanese cultural context. Using the CA subset, we observe performance drop in many LMMs when evaluated in Japanese, which is purely attributable to language variation. Using the CS subset, we reveal their inadequate Japanese cultural understanding. Further, by combining both subsets, we identify that some LMMs perform well on the CA subset but not on the CS subset, exposing a shallow understanding of the Japanese language that lacks depth in cultural understanding. We hope this work will not only help advance LMM performance in Japanese but also serve as a guideline to create high-standard, culturally diverse benchmarks for multilingual LMM development. The project page is https://mmmu-japanese-benchmark.github.io/JMMMU/.
comment: Project page: https://mmmu-japanese-benchmark.github.io/JMMMU/
☆ PyramidDrop: Accelerating Your Large Vision-Language Models via Pyramid Visual Redundancy Reduction
In large vision-language models (LVLMs), images serve as inputs that carry a wealth of information. As the idiom "A picture is worth a thousand words" implies, representing a single image in current LVLMs can require hundreds or even thousands of tokens. This results in significant computational costs, which grow quadratically as input image resolution increases, thereby severely impacting the efficiency of both training and inference. Previous approaches have attempted to reduce the number of image tokens either before or within the early layers of LVLMs. However, these strategies inevitably result in the loss of crucial image information, ultimately diminishing model performance. To address this challenge, we conduct an empirical study revealing that all visual tokens are necessary for LVLMs in the shallow layers, and token redundancy progressively increases in the deeper layers of the model. To this end, we propose PyramidDrop, a visual redundancy reduction strategy for LVLMs to boost their efficiency in both training and inference with neglectable performance loss. Specifically, we partition the LVLM into several stages and drop part of the image tokens at the end of each stage with a pre-defined ratio, creating pyramid-like visual tokens across model layers. The dropping is based on a lightweight similarity calculation with a negligible time overhead. Extensive experiments demonstrate that PyramidDrop can achieve a 40% training time and 55% inference FLOPs acceleration of LLaVA-NeXT with comparable performance. Besides, the PyramidDrop could also serve as a plug-and-play strategy for inference acceleration without training, with better performance and lower inference cost than counterparts. We hope that the insights and approach introduced by PyramidDrop will inspire future research to further investigate the role of image tokens in LVLMs.
comment: 10 pages
☆ Breaking the Memory Barrier: Near Infinite Batch Size Scaling for Contrastive Loss
Contrastive loss is a powerful approach for representation learning, where larger batch sizes enhance performance by providing more negative samples to better distinguish between similar and dissimilar data. However, scaling batch sizes is constrained by the quadratic growth in GPU memory consumption, primarily due to the full instantiation of the similarity matrix. To address this, we propose a tile-based computation strategy that partitions the contrastive loss calculation into arbitrary small blocks, avoiding full materialization of the similarity matrix. Furthermore, we introduce a multi-level tiling strategy to leverage the hierarchical structure of distributed systems, employing ring-based communication at the GPU level to optimize synchronization and fused kernels at the CUDA core level to reduce I/O overhead. Experimental results show that the proposed method scales batch sizes to unprecedented levels. For instance, it enables contrastive training of a CLIP-ViT-L/14 model with a batch size of 4M or 12M using 8 or 32 A800 80GB without sacrificing any accuracy. Compared to SOTA memory-efficient solutions, it achieves a two-order-of-magnitude reduction in memory while maintaining comparable speed. The code will be made publicly available.
☆ LVSM: A Large View Synthesis Model with Minimal 3D Inductive Bias
We propose the Large View Synthesis Model (LVSM), a novel transformer-based approach for scalable and generalizable novel view synthesis from sparse-view inputs. We introduce two architectures: (1) an encoder-decoder LVSM, which encodes input image tokens into a fixed number of 1D latent tokens, functioning as a fully learned scene representation, and decodes novel-view images from them; and (2) a decoder-only LVSM, which directly maps input images to novel-view outputs, completely eliminating intermediate scene representations. Both models bypass the 3D inductive biases used in previous methods -- from 3D representations (e.g., NeRF, 3DGS) to network designs (e.g., epipolar projections, plane sweeps) -- addressing novel view synthesis with a fully data-driven approach. While the encoder-decoder model offers faster inference due to its independent latent representation, the decoder-only LVSM achieves superior quality, scalability, and zero-shot generalization, outperforming previous state-of-the-art methods by 1.5 to 3.5 dB PSNR. Comprehensive evaluations across multiple datasets demonstrate that both LVSM variants achieve state-of-the-art novel view synthesis quality. Notably, our models surpass all previous methods even with reduced computational resources (1-2 GPUs). Please see our website for more details: https://haian-jin.github.io/projects/LVSM/ .
comment: project page: https://haian-jin.github.io/projects/LVSM/
☆ Frontiers in Intelligent Colonoscopy
Colonoscopy is currently one of the most sensitive screening methods for colorectal cancer. This study investigates the frontiers of intelligent colonoscopy techniques and their prospective implications for multimodal medical applications. With this goal, we begin by assessing the current data-centric and model-centric landscapes through four tasks for colonoscopic scene perception, including classification, detection, segmentation, and vision-language understanding. This assessment enables us to identify domain-specific challenges and reveals that multimodal research in colonoscopy remains open for further exploration. To embrace the coming multimodal era, we establish three foundational initiatives: a large-scale multimodal instruction tuning dataset ColonINST, a colonoscopy-designed multimodal language model ColonGPT, and a multimodal benchmark. To facilitate ongoing monitoring of this rapidly evolving field, we provide a public website for the latest updates: https://github.com/ai4colonoscopy/IntelliScope.
comment: [work in progress] A comprehensive survey of intelligent colonoscopy in the multimodal era
☆ Automated Spinal MRI Labelling from Reports Using a Large Language Model MICCAI 2024
We propose a general pipeline to automate the extraction of labels from radiology reports using large language models, which we validate on spinal MRI reports. The efficacy of our labelling method is measured on five distinct conditions: spinal cancer, stenosis, spondylolisthesis, cauda equina compression and herniation. Using open-source models, our method equals or surpasses GPT-4 on a held-out set of reports. Furthermore, we show that the extracted labels can be used to train imaging models to classify the identified conditions in the accompanying MR scans. All classifiers trained using automated labels achieve comparable performance to models trained using scans manually annotated by clinicians. Code can be found at https://github.com/robinyjpark/AutoLabelClassifier.
comment: Accepted to Medical Image Computing and Computer Assisted Intervention (MICCAI 2024, Spotlight). 11 pages plus appendix
☆ EPContrast: Effective Point-level Contrastive Learning for Large-scale Point Cloud Understanding
The acquisition of inductive bias through point-level contrastive learning holds paramount significance in point cloud pre-training. However, the square growth in computational requirements with the scale of the point cloud poses a substantial impediment to the practical deployment and execution. To address this challenge, this paper proposes an Effective Point-level Contrastive Learning method for large-scale point cloud understanding dubbed \textbf{EPContrast}, which consists of AGContrast and ChannelContrast. In practice, AGContrast constructs positive and negative pairs based on asymmetric granularity embedding, while ChannelContrast imposes contrastive supervision between channel feature maps. EPContrast offers point-level contrastive loss while concurrently mitigating the computational resource burden. The efficacy of EPContrast is substantiated through comprehensive validation on S3DIS and ScanNetV2, encompassing tasks such as semantic segmentation, instance segmentation, and object detection. In addition, rich ablation experiments demonstrate remarkable bias induction capabilities under label-efficient and one-epoch training settings.
☆ Emphasizing Discriminative Features for Dataset Distillation in Complex Scenarios
Dataset distillation has demonstrated strong performance on simple datasets like CIFAR, MNIST, and TinyImageNet but struggles to achieve similar results in more complex scenarios. In this paper, we propose EDF (emphasizes the discriminative features), a dataset distillation method that enhances key discriminative regions in synthetic images using Grad-CAM activation maps. Our approach is inspired by a key observation: in simple datasets, high-activation areas typically occupy most of the image, whereas in complex scenarios, the size of these areas is much smaller. Unlike previous methods that treat all pixels equally when synthesizing images, EDF uses Grad-CAM activation maps to enhance high-activation areas. From a supervision perspective, we downplay supervision signals that have lower losses, as they contain common patterns. Additionally, to help the DD community better explore complex scenarios, we build the Complex Dataset Distillation (Comp-DD) benchmark by meticulously selecting sixteen subsets, eight easy and eight hard, from ImageNet-1K. In particular, EDF consistently outperforms SOTA results in complex scenarios, such as ImageNet-1K subsets. Hopefully, more researchers will be inspired and encouraged to improve the practicality and efficacy of DD. Our code and benchmark will be made public at https://github.com/NUS-HPC-AI-Lab/EDF.
comment: 24 pages, 13 figures
☆ KANICE: Kolmogorov-Arnold Networks with Interactive Convolutional Elements
We introduce KANICE (Kolmogorov-Arnold Networks with Interactive Convolutional Elements), a novel neural architecture that combines Convolutional Neural Networks (CNNs) with Kolmogorov-Arnold Network (KAN) principles. KANICE integrates Interactive Convolutional Blocks (ICBs) and KAN linear layers into a CNN framework. This leverages KANs' universal approximation capabilities and ICBs' adaptive feature learning. KANICE captures complex, non-linear data relationships while enabling dynamic, context-dependent feature extraction based on the Kolmogorov-Arnold representation theorem. We evaluated KANICE on four datasets: MNIST, Fashion-MNIST, EMNIST, and SVHN, comparing it against standard CNNs, CNN-KAN hybrids, and ICB variants. KANICE consistently outperformed baseline models, achieving 99.35% accuracy on MNIST and 90.05% on the SVHN dataset. Furthermore, we introduce KANICE-mini, a compact variant designed for efficiency. A comprehensive ablation study demonstrates that KANICE-mini achieves comparable performance to KANICE with significantly fewer parameters. KANICE-mini reached 90.00% accuracy on SVHN with 2,337,828 parameters, compared to KANICE's 25,432,000. This study highlights the potential of KAN-based architectures in balancing performance and computational efficiency in image classification tasks. Our work contributes to research in adaptive neural networks, integrates mathematical theorems into deep learning architectures, and explores the trade-offs between model complexity and performance, advancing computer vision and pattern recognition. The source code for this paper is publicly accessible through our GitHub repository (https://github.com/m-ferdaus/kanice).
☆ Are Visual-Language Models Effective in Action Recognition? A Comparative Study
Current vision-language foundation models, such as CLIP, have recently shown significant improvement in performance across various downstream tasks. However, whether such foundation models significantly improve more complex fine-grained action recognition tasks is still an open question. To answer this question and better find out the future research direction on human behavior analysis in-the-wild, this paper provides a large-scale study and insight on current state-of-the-art vision foundation models by comparing their transfer ability onto zero-shot and frame-wise action recognition tasks. Extensive experiments are conducted on recent fine-grained, human-centric action recognition datasets (e.g., Toyota Smarthome, Penn Action, UAV-Human, TSU, Charades) including action classification and segmentation.
☆ LiNeS: Post-training Layer Scaling Prevents Forgetting and Enhances Model Merging
Large pre-trained models exhibit impressive zero-shot performance across diverse tasks, but fine-tuning often leads to catastrophic forgetting, where improvements on a target domain degrade generalization on other tasks. To address this challenge, we introduce LiNeS, Layer-increasing Network Scaling, a post-training editing technique designed to preserve pre-trained generalization while enhancing fine-tuned task performance. LiNeS scales parameter updates linearly based on their layer depth within the network, maintaining shallow layers close to their pre-trained values to preserve general features while allowing deeper layers to retain task-specific representations. We further extend this approach to multi-task model merging scenarios, where layer-wise scaling of merged parameters reduces negative task interference. LiNeS demonstrates significant improvements in both single-task and multi-task settings across various benchmarks in vision and natural language processing. It mitigates forgetting, enhances out-of-distribution generalization, integrates seamlessly with existing multi-task model merging baselines improving their performance across benchmarks and model sizes, and can boost generalization when merging LLM policies aligned with different rewards via RLHF. Importantly, our method is simple to implement and complementary to many existing techniques.
comment: The first two authors contributed equally to this work; Project website: \url{https://lines-merging.github.io/}
☆ YOLO-TS: Real-Time Traffic Sign Detection with Enhanced Accuracy Using Optimized Receptive Fields and Anchor-Free Fusion
Ensuring safety in both autonomous driving and advanced driver-assistance systems (ADAS) depends critically on the efficient deployment of traffic sign recognition technology. While current methods show effectiveness, they often compromise between speed and accuracy. To address this issue, we present a novel real-time and efficient road sign detection network, YOLO-TS. This network significantly improves performance by optimizing the receptive fields of multi-scale feature maps to align more closely with the size distribution of traffic signs in various datasets. Moreover, our innovative feature-fusion strategy, leveraging the flexibility of Anchor-Free methods, allows for multi-scale object detection on a high-resolution feature map abundant in contextual information, achieving remarkable enhancements in both accuracy and speed. To mitigate the adverse effects of the grid pattern caused by dilated convolutions on the detection of smaller objects, we have devised a unique module that not only mitigates this grid effect but also widens the receptive field to encompass an extensive range of spatial contextual information, thus boosting the efficiency of information usage. Evaluation on challenging public datasets, TT100K and CCTSDB2021, demonstrates that YOLO-TS surpasses existing state-of-the-art methods in terms of both accuracy and speed. The code for our method will be available.
comment: 13 pages, 9 figures and 7 tables
☆ AlphaChimp: Tracking and Behavior Recognition of Chimpanzees
Understanding non-human primate behavior is crucial for improving animal welfare, modeling social behavior, and gaining insights into both distinctly human and shared behaviors. Despite recent advances in computer vision, automated analysis of primate behavior remains challenging due to the complexity of their social interactions and the lack of specialized algorithms. Existing methods often struggle with the nuanced behaviors and frequent occlusions characteristic of primate social dynamics. This study aims to develop an effective method for automated detection, tracking, and recognition of chimpanzee behaviors in video footage. Here we show that our proposed method, AlphaChimp, an end-to-end approach that simultaneously detects chimpanzee positions and estimates behavior categories from videos, significantly outperforms existing methods in behavior recognition. AlphaChimp achieves approximately 10% higher tracking accuracy and a 20% improvement in behavior recognition compared to state-of-the-art methods, particularly excelling in the recognition of social behaviors. This superior performance stems from AlphaChimp's innovative architecture, which integrates temporal feature fusion with a Transformer-based self-attention mechanism, enabling more effective capture and interpretation of complex social interactions among chimpanzees. Our approach bridges the gap between computer vision and primatology, enhancing technical capabilities and deepening our understanding of primate communication and sociality. We release our code and models and hope this will facilitate future research in animal social dynamics. This work contributes to ethology, cognitive science, and artificial intelligence, offering new perspectives on social intelligence.
comment: An eXpressive extension of ChimpACT [arXiv:2310.16447], proposes AlphaChimp for tracking and behavior recognition of chimpanzees. arXiv admin note: substantial text overlap with arXiv:2310.16447
☆ CLAP: Concave Linear APproximation for Quadratic Graph Matching SC
Solving point-wise feature correspondence in visual data is a fundamental problem in computer vision. A powerful model that addresses this challenge is to formulate it as graph matching, which entails solving a Quadratic Assignment Problem (QAP) with node-wise and edge-wise constraints. However, solving such a QAP can be both expensive and difficult due to numerous local extreme points. In this work, we introduce a novel linear model and solver designed to accelerate the computation of graph matching. Specifically, we employ a positive semi-definite matrix approximation to establish the structural attribute constraint.We then transform the original QAP into a linear model that is concave for maximization. This model can subsequently be solved using the Sinkhorn optimal transport algorithm, known for its enhanced efficiency and numerical stability compared to existing approaches. Experimental results on the widely used benchmark PascalVOC showcase that our algorithm achieves state-of-the-art performance with significantly improved efficiency. Source code: https://github.com/xmlyqing00/clap
comment: Accepted as an oral paper in International Symposium on Visual Computing (ISCV2024)
☆ Masked Differential Privacy
Privacy-preserving computer vision is an important emerging problem in machine learning and artificial intelligence. The prevalent methods tackling this problem use differential privacy or anonymization and obfuscation techniques to protect the privacy of individuals. In both cases, the utility of the trained model is sacrificed heavily in this process. In this work, we propose an effective approach called masked differential privacy (MaskDP), which allows for controlling sensitive regions where differential privacy is applied, in contrast to applying DP on the entire input. Our method operates selectively on the data and allows for defining non-sensitive spatio-temporal regions without DP application or combining differential privacy with other privacy techniques within data samples. Experiments on four challenging action recognition datasets demonstrate that our proposed techniques result in better utility-privacy trade-offs compared to standard differentially private training in the especially demanding $\epsilon<1$ regime.
☆ A Survey on Deep Learning-based Gaze Direction Regression: Searching for the State-of-the-art SP
In this paper, we present a survey of deep learning-based methods for the regression of gaze direction vector from head and eye images. We describe in detail numerous published methods with a focus on the input data, architecture of the model, and loss function used to supervise the model. Additionally, we present a list of datasets that can be used to train and evaluate gaze direction regression methods. Furthermore, we noticed that the results reported in the literature are often not comparable one to another due to differences in the validation or even test subsets used. To address this problem, we re-evaluated several methods on the commonly used in-the-wild Gaze360 dataset using the same validation setup. The experimental results show that the latest methods, although claiming state-of-the-art results, significantly underperform compared with some older methods. Finally, we show that the temporal models outperform the static models under static test conditions.
comment: Accepted on SPRA 2024 (Istanbul, Turkey)
☆ Neuronal Competition Groups with Supervised STDP for Spike-Based Classification
Spike Timing-Dependent Plasticity (STDP) is a promising substitute to backpropagation for local training of Spiking Neural Networks (SNNs) on neuromorphic hardware. STDP allows SNNs to address classification tasks by combining unsupervised STDP for feature extraction and supervised STDP for classification. Unsupervised STDP is usually employed with Winner-Takes-All (WTA) competition to learn distinct patterns. However, WTA for supervised STDP classification faces unbalanced competition challenges. In this paper, we propose a method to effectively implement WTA competition in a spiking classification layer employing first-spike coding and supervised STDP training. We introduce the Neuronal Competition Group (NCG), an architecture that improves classification capabilities by promoting the learning of various patterns per class. An NCG is a group of neurons mapped to a specific class, implementing intra-class WTA and a novel competition regulation mechanism based on two-compartment thresholds. We incorporate our proposed architecture into spiking classification layers trained with state-of-the-art supervised STDP rules. On top of two different unsupervised feature extractors, we obtain significant accuracy improvements on image recognition datasets such as CIFAR-10 and CIFAR-100. We show that our competition regulation mechanism is crucial for ensuring balanced competition and improved class separation.
☆ Multi Kernel Estimation based Object Segmentation
This paper presents a novel approach for multi-kernel estimation by enhancing the KernelGAN algorithm, which traditionally estimates a single kernel for the entire image. We introduce Multi-KernelGAN, which extends KernelGAN's capabilities by estimating two distinct kernels based on object segmentation masks. Our approach is validated through three distinct methods: texture-based patch Fast Fourier Transform (FFT) calculation, detail-based segmentation, and deep learning-based object segmentation using YOLOv8 and the Segment Anything Model (SAM). Among these methods, the combination of YOLO and SAM yields the best results for kernel estimation. Experimental results demonstrate that our multi-kernel estimation technique outperforms conventional single-kernel methods in super-resolution tasks.
☆ LFME: A Simple Framework for Learning from Multiple Experts in Domain Generalization NeurIPS 2024
Domain generalization (DG) methods aim to maintain good performance in an unseen target domain by using training data from multiple source domains. While success on certain occasions are observed, enhancing the baseline across most scenarios remains challenging. This work introduces a simple yet effective framework, dubbed learning from multiple experts (LFME), that aims to make the target model an expert in all source domains to improve DG. Specifically, besides learning the target model used in inference, LFME will also train multiple experts specialized in different domains, whose output probabilities provide professional guidance by simply regularizing the logit of the target model. Delving deep into the framework, we reveal that the introduced logit regularization term implicitly provides effects of enabling the target model to harness more information, and mining hard samples from the experts during training. Extensive experiments on benchmarks from different DG tasks demonstrate that LFME is consistently beneficial to the baseline and can achieve comparable performance to existing arts. Code is available at~\url{https://github.com/liangchen527/LFME}.
comment: Accepted by NeurIPS 2024
☆ SPVSoAP3D: A Second-order Average Pooling Approach to enhance 3D Place Recognition in Horticultural Environments IROS 2024
3D LiDAR-based place recognition has been extensively researched in urban environments, yet it remains underexplored in agricultural settings. Unlike urban contexts, horticultural environments, characterized by their permeability to laser beams, result in sparse and overlapping LiDAR scans with suboptimal geometries. This phenomenon leads to intra- and inter-row descriptor ambiguity. In this work, we address this challenge by introducing SPVSoAP3D, a novel modeling approach that combines a voxel-based feature extraction network with an aggregation technique based on a second-order average pooling operator, complemented by a descriptor enhancement stage. Furthermore, we augment the existing HORTO-3DLM dataset by introducing two new sequences derived from horticultural environments. We evaluate the performance of SPVSoAP3D against state-of-the-art (SOTA) models, including OverlapTransformer, PointNetVLAD, and LOGG3D-Net, utilizing a cross-validation protocol on both the newly introduced sequences and the existing HORTO-3DLM dataset. The findings indicate that the average operator is more suitable for horticultural environments compared to the max operator and other first-order pooling techniques. Additionally, the results highlight the improvements brought by the descriptor enhancement stage.
comment: This work has been accepted to IROS 2024
☆ Joint Point Cloud Upsampling and Cleaning with Octree-based CNNs
Recovering dense and uniformly distributed point clouds from sparse or noisy data remains a significant challenge. Recently, great progress has been made on these tasks, but usually at the cost of increasingly intricate modules or complicated network architectures, leading to long inference time and huge resource consumption. Instead, we embrace simplicity and present a simple yet efficient method for jointly upsampling and cleaning point clouds. Our method leverages an off-the-shelf octree-based 3D U-Net (OUNet) with minor modifications, enabling the upsampling and cleaning tasks within a single network. Our network directly processes each input point cloud as a whole instead of processing each point cloud patch as in previous works, which significantly eases the implementation and brings at least 47 times faster inference. Extensive experiments demonstrate that our method achieves state-of-the-art performances under huge efficiency advantages on a series of benchmarks. We expect our method to serve simple baselines and inspire researchers to rethink the method design on point cloud upsampling and cleaning.
comment: Accepted by Computational Visual Media
☆ AGSENet: A Robust Road Ponding Detection Method for Proactive Traffic Safety
Road ponding, a prevalent traffic hazard, poses a serious threat to road safety by causing vehicles to lose control and leading to accidents ranging from minor fender benders to severe collisions. Existing technologies struggle to accurately identify road ponding due to complex road textures and variable ponding coloration influenced by reflection characteristics. To address this challenge, we propose a novel approach called Self-Attention-based Global Saliency-Enhanced Network (AGSENet) for proactive road ponding detection and traffic safety improvement. AGSENet incorporates saliency detection techniques through the Channel Saliency Information Focus (CSIF) and Spatial Saliency Information Enhancement (SSIE) modules. The CSIF module, integrated into the encoder, employs self-attention to highlight similar features by fusing spatial and channel information. The SSIE module, embedded in the decoder, refines edge features and reduces noise by leveraging correlations across different feature levels. To ensure accurate and reliable evaluation, we corrected significant mislabeling and missing annotations in the Puddle-1000 dataset. Additionally, we constructed the Foggy-Puddle and Night-Puddle datasets for road ponding detection in low-light and foggy conditions, respectively. Experimental results demonstrate that AGSENet outperforms existing methods, achieving IoU improvements of 2.03\%, 0.62\%, and 1.06\% on the Puddle-1000, Foggy-Puddle, and Night-Puddle datasets, respectively, setting a new state-of-the-art in this field. Finally, we verified the algorithm's reliability on edge computing devices. This work provides a valuable reference for proactive warning research in road traffic safety.
comment: 21 pages, 15 figures
☆ E-3DGS: Gaussian Splatting with Exposure and Motion Events
Estimating Neural Radiance Fields (NeRFs) from images captured under optimal conditions has been extensively explored in the vision community. However, robotic applications often face challenges such as motion blur, insufficient illumination, and high computational overhead, which adversely affect downstream tasks like navigation, inspection, and scene visualization. To address these challenges, we propose E-3DGS, a novel event-based approach that partitions events into motion (from camera or object movement) and exposure (from camera exposure), using the former to handle fast-motion scenes and using the latter to reconstruct grayscale images for high-quality training and optimization of event-based 3D Gaussian Splatting (3DGS). We introduce a novel integration of 3DGS with exposure events for high-quality reconstruction of explicit scene representations. Our versatile framework can operate on motion events alone for 3D reconstruction, enhance quality using exposure events, or adopt a hybrid mode that balances quality and effectiveness by optimizing with initial exposure events followed by high-speed motion events. We also introduce EME-3D, a real-world 3D dataset with exposure events, motion events, camera calibration parameters, and sparse point clouds. Our method is faster and delivers better reconstruction quality than event-based NeRF while being more cost-effective than NeRF methods that combine event and RGB data by using a single event sensor. By combining motion and exposure events, E-3DGS sets a new benchmark for event-based 3D reconstruction with robust performance in challenging conditions and lower hardware demands. The source code and dataset will be available at https://github.com/MasterHow/E-3DGS.
comment: The source code and dataset will be available at https://github.com/MasterHow/E-3DGS
☆ Multi-Layer Gaussian Splatting for Immersive Anatomy Visualization
In medical image visualization, path tracing of volumetric medical data like CT scans produces lifelike three-dimensional visualizations. Immersive VR displays can further enhance the understanding of complex anatomies. Going beyond the diagnostic quality of traditional 2D slices, they enable interactive 3D evaluation of anatomies, supporting medical education and planning. Rendering high-quality visualizations in real-time, however, is computationally intensive and impractical for compute-constrained devices like mobile headsets. We propose a novel approach utilizing GS to create an efficient but static intermediate representation of CT scans. We introduce a layered GS representation, incrementally including different anatomical structures while minimizing overlap and extending the GS training to remove inactive Gaussians. We further compress the created model with clustering across layers. Our approach achieves interactive frame rates while preserving anatomical structures, with quality adjustable to the target hardware. Compared to standard GS, our representation retains some of the explorative qualities initially enabled by immersive path tracing. Selective activation and clipping of layers are possible at rendering time, adding a degree of interactivity to otherwise static GS models. This could enable scenarios where high computational demands would otherwise prohibit using path-traced medical volumes.
☆ Leaky ReLUs That Differ in Forward and Backward Pass Facilitate Activation Maximization in Deep Neural Networks
Activation maximization (AM) strives to generate optimal input stimuli, revealing features that trigger high responses in trained deep neural networks. AM is an important method of explainable AI. We demonstrate that AM fails to produce optimal input stimuli for simple functions containing ReLUs or Leaky ReLUs, casting doubt on the practical usefulness of AM and the visual interpretation of the generated images. This paper proposes a solution based on using Leaky ReLUs with a high negative slope in the backward pass while keeping the original, usually zero, slope in the forward pass. The approach significantly increases the maxima found by AM. The resulting ProxyGrad algorithm implements a novel optimization technique for neural networks that employs a secondary network as a proxy for gradient computation. This proxy network is designed to have a simpler loss landscape with fewer local maxima than the original network. Our chosen proxy network is an identical copy of the original network, including its weights, with distinct negative slopes in the Leaky ReLUs. Moreover, we show that ProxyGrad can be used to train the weights of Convolutional Neural Networks for classification such that, on some of the tested benchmarks, they outperform traditional networks.
☆ PGCS: Physical Law embedded Generative Cloud Synthesis in Remote Sensing Images
Data quantity and quality are both critical for information extraction and analyzation in remote sensing. However, the current remote sensing datasets often fail to meet these two requirements, for which cloud is a primary factor degrading the data quantity and quality. This limitation affects the precision of results in remote sensing application, particularly those derived from data-driven techniques. In this paper, a physical law embedded generative cloud synthesis method (PGCS) is proposed to generate diverse realistic cloud images to enhance real data and promote the development of algorithms for subsequent tasks, such as cloud correction, cloud detection, and data augmentation for classification, recognition, and segmentation. The PGCS method involves two key phases: spatial synthesis and spectral synthesis. In the spatial synthesis phase, a style-based generative adversarial network is utilized to simulate the spatial characteristics, generating an infinite number of single-channel clouds. In the spectral synthesis phase, the atmospheric scattering law is embedded through a local statistics and global fitting method, converting the single-channel clouds into multi-spectral clouds. The experimental results demonstrate that PGCS achieves a high accuracy in both phases and performs better than three other existing cloud synthesis methods. Two cloud correction methods are developed from PGCS and exhibits a superior performance compared to state-of-the-art methods in the cloud correction task. Furthermore, the application of PGCS with data from various sensors was investigated and successfully extended. Code will be provided at https://github.com/Liying-Xu/PGCS.
comment: 20 pages, 16 figures
☆ Towards Real Zero-Shot Camouflaged Object Segmentation without Camouflaged Annotations
Camouflaged Object Segmentation (COS) faces significant challenges due to the scarcity of annotated data, where meticulous pixel-level annotation is both labor-intensive and costly, primarily due to the intricate object-background boundaries. Addressing the core question, "Can COS be effectively achieved in a zero-shot manner without manual annotations for any camouflaged object?" we affirmatively respond and introduce a robust zero-shot COS framework. This framework leverages the inherent local pattern bias of COS and employs a broad semantic feature space derived from salient object segmentation (SOS) for efficient zero-shot transfer. We incorporate an Masked Image Modeling (MIM) based image encoder optimized for Parameter-Efficient Fine-Tuning (PEFT), a Multimodal Large Language Model (M-LLM), and a Multi-scale Fine-grained Alignment (MFA) mechanism. The MIM pre-trained image encoder focuses on capturing essential low-level features, while the M-LLM generates caption embeddings processed alongside these visual cues. These embeddings are precisely aligned using MFA, enabling our framework to accurately interpret and navigate complex semantic contexts. To optimize operational efficiency, we introduce a learnable codebook that represents the M-LLM during inference, significantly reducing computational overhead. Our framework demonstrates its versatility and efficacy through rigorous experimentation, achieving state-of-the-art performance in zero-shot COS with $F_{\beta}^w$ scores of 72.9\% on CAMO and 71.7\% on COD10K. By removing the M-LLM during inference, we achieve an inference speed comparable to that of traditional end-to-end models, reaching 18.1 FPS. Code: https://github.com/R-LEI360725/ZSCOS-CaMF
☆ ISImed: A Framework for Self-Supervised Learning using Intrinsic Spatial Information in Medical Images
This paper demonstrates that spatial information can be used to learn interpretable representations in medical images using Self-Supervised Learning (SSL). Our proposed method, ISImed, is based on the observation that medical images exhibit a much lower variability among different images compared to classic data vision benchmarks. By leveraging this resemblance of human body structures across multiple images, we establish a self-supervised objective that creates a latent representation capable of capturing its location in the physical realm. More specifically, our method involves sampling image crops and creating a distance matrix that compares the learned representation vectors of all possible combinations of these crops to the true distance between them. The intuition is, that the learned latent space is a positional encoding for a given image crop. We hypothesize, that by learning these positional encodings, comprehensive image representations have to be generated. To test this hypothesis and evaluate our method, we compare our learned representation with two state-of-the-art SSL benchmarking methods on two publicly available medical imaging datasets. We show that our method can efficiently learn representations that capture the underlying structure of the data and can be used to transfer to a downstream classification task.
comment: 11 pages, 4 figures
☆ IdenBAT: Disentangled Representation Learning for Identity-Preserved Brain Age Transformation
Brain age transformation aims to convert reference brain images into synthesized images that accurately reflect the age-specific features of a target age group. The primary objective of this task is to modify only the age-related attributes of the reference image while preserving all other age-irrelevant attributes. However, achieving this goal poses substantial challenges due to the inherent entanglement of various image attributes within features extracted from a backbone encoder, resulting in simultaneous alterations during the image generation. To address this challenge, we propose a novel architecture that employs disentangled representation learning for identity-preserved brain age transformation called IdenBAT. This approach facilitates the decomposition of image features, ensuring the preservation of individual traits while selectively transforming age-related characteristics to match those of the target age group. Through comprehensive experiments conducted on both 2D and full-size 3D brain datasets, our method adeptly converts input images to target age while retaining individual characteristics accurately. Furthermore, our approach demonstrates superiority over existing state-of-the-art regarding performance fidelity.
comment: 16 pages, 8 figures, 2 tables
☆ DiP-GO: A Diffusion Pruner via Few-step Gradient Optimization
Diffusion models have achieved remarkable progress in the field of image generation due to their outstanding capabilities. However, these models require substantial computing resources because of the multi-step denoising process during inference. While traditional pruning methods have been employed to optimize these models, the retraining process necessitates large-scale training datasets and extensive computational costs to maintain generalization ability, making it neither convenient nor efficient. Recent studies attempt to utilize the similarity of features across adjacent denoising stages to reduce computational costs through simple and static strategies. However, these strategies cannot fully harness the potential of the similar feature patterns across adjacent timesteps. In this work, we propose a novel pruning method that derives an efficient diffusion model via a more intelligent and differentiable pruner. At the core of our approach is casting the model pruning process into a SubNet search process. Specifically, we first introduce a SuperNet based on standard diffusion via adding some backup connections built upon the similar features. We then construct a plugin pruner network and design optimization losses to identify redundant computation. Finally, our method can identify an optimal SubNet through few-step gradient optimization and a simple post-processing procedure. We conduct extensive experiments on various diffusion models including Stable Diffusion series and DiTs. Our DiP-GO approach achieves 4.4 x speedup for SD-1.5 without any loss of accuracy, significantly outperforming the previous state-of-the-art methods.
☆ LIMIS: Towards Language-based Interactive Medical Image Segmentation
Within this work, we introduce LIMIS: The first purely language-based interactive medical image segmentation model. We achieve this by adapting Grounded SAM to the medical domain and designing a language-based model interaction strategy that allows radiologists to incorporate their knowledge into the segmentation process. LIMIS produces high-quality initial segmentation masks by leveraging medical foundation models and allows users to adapt segmentation masks using only language, opening up interactive segmentation to scenarios where physicians require using their hands for other tasks. We evaluate LIMIS on three publicly available medical datasets in terms of performance and usability with experts from the medical domain confirming its high-quality segmentation masks and its interactive usability.
☆ Hierarchical Clustering for Conditional Diffusion in Image Generation ICLR 2025
Finding clusters of data points with similar characteristics and generating new cluster-specific samples can significantly enhance our understanding of complex data distributions. While clustering has been widely explored using Variational Autoencoders, these models often lack generation quality in real-world datasets. This paper addresses this gap by introducing TreeDiffusion, a deep generative model that conditions Diffusion Models on hierarchical clusters to obtain high-quality, cluster-specific generations. The proposed pipeline consists of two steps: a VAE-based clustering model that learns the hierarchical structure of the data, and a conditional diffusion model that generates realistic images for each cluster. We propose this two-stage process to ensure that the generated samples remain representative of their respective clusters and enhance image fidelity to the level of diffusion models. A key strength of our method is its ability to create images for each cluster, providing better visualization of the learned representations by the clustering model, as demonstrated through qualitative results. This method effectively addresses the generative limitations of VAE-based approaches while preserving their clustering performance. Empirically, we demonstrate that conditioning diffusion models on hierarchical clusters significantly enhances generative performance, thereby advancing the state of generative clustering models.
comment: 25 pages, submitted to ICLR 2025
☆ Mitigating Vanishing Activations in Deep CapsNets Using Channel Pruning
Capsule Networks outperform Convolutional Neural Networks in learning the part-whole relationships with viewpoint invariance, and the credit goes to their multidimensional capsules. It was assumed that increasing the number of capsule layers in the capsule networks would enhance the model performance. However, recent studies found that Capsule Networks lack scalability due to vanishing activations in the capsules of deeper layers. This paper thoroughly investigates the vanishing activation problem in deep Capsule Networks. To analyze this issue and understand how increasing capsule dimensions can facilitate deeper networks, various Capsule Network models are constructed and evaluated with different numbers of capsules, capsule dimensions, and intermediate layers for this paper. Unlike traditional model pruning, which reduces the number of model parameters and expedites model training, this study uses pruning to mitigate the vanishing activations in the deeper capsule layers. In addition, the backbone network and capsule layers are pruned with different pruning ratios to reduce the number of inactive capsules and achieve better model accuracy than the unpruned models.
☆ MBD: Multi b-value Denoising of Diffusion Magnetic Resonance Images
We propose a novel approach to denoising diffusion magnetic resonance images (dMRI) using convolutional neural networks, that exploits the benefits of data acquired at multiple b-values to offset the need for many redundant observations. Denoising is especially relevant in dMRI since noise can have a deleterious impact on both quantification accuracy and image preprocessing. The most successful methods proposed to date, like Marchenko-Pastur Principal Component Analysis (MPPCA) denoising, are tailored to diffusion-weighting repeated for many encoding directions. They exploit high redundancy of the dataset that oversamples the diffusion-encoding direction space, since many directions have collinear components. However, there are many dMRI techniques that do not entail a large number of encoding directions or repetitions, and are therefore less suited to this approach. For example, clinical dMRI exams may include as few as three encoding directions, with low or negligible data redundancy across directions. Moreover, promising new dMRI approaches, like spherical b-tensor encoding (STE), benefit from high b-values while sensitizing the signal to diffusion along all directions in just a single shot. We introduce a convolutional neural network approach that we call multi-b-value-based denoising (MBD). MBD exploits the similarity in diffusion-weighted images (DWI) across different b-values but along the same diffusion encoding direction. It allows denoising of diffusion images with high noise variance while avoiding blurring, and using just a small number input images.
comment: this is a biomedical engineering work using machine learning to enhance medical images
☆ Enhancing Generalization in Convolutional Neural Networks through Regularization with Edge and Line Features
This paper proposes a novel regularization approach to bias Convolutional Neural Networks (CNNs) toward utilizing edge and line features in their hidden layers. Rather than learning arbitrary kernels, we constrain the convolution layers to edge and line detection kernels. This intentional bias regularizes the models, improving generalization performance, especially on small datasets. As a result, test accuracies improve by margins of 5-11 percentage points across four challenging fine-grained classification datasets with limited training data and an identical number of trainable parameters. Instead of traditional convolutional layers, we use Pre-defined Filter Modules, which convolve input data using a fixed set of 3x3 pre-defined edge and line filters. A subsequent ReLU erases information that did not trigger any positive response. Next, a 1x1 convolutional layer generates linear combinations. Notably, the pre-defined filters are a fixed component of the architecture, remaining unchanged during the training phase. Our findings reveal that the number of dimensions spanned by the set of pre-defined filters has a low impact on recognition performance. However, the size of the set of filters matters, with nine or more filters providing optimal results.
☆ VistaDream: Sampling multiview consistent images for single-view scene reconstruction
In this paper, we propose VistaDream a novel framework to reconstruct a 3D scene from a single-view image. Recent diffusion models enable generating high-quality novel-view images from a single-view input image. Most existing methods only concentrate on building the consistency between the input image and the generated images while losing the consistency between the generated images. VistaDream addresses this problem by a two-stage pipeline. In the first stage, VistaDream begins with building a global coarse 3D scaffold by zooming out a little step with inpainted boundaries and an estimated depth map. Then, on this global scaffold, we use iterative diffusion-based RGB-D inpainting to generate novel-view images to inpaint the holes of the scaffold. In the second stage, we further enhance the consistency between the generated novel-view images by a novel training-free Multiview Consistency Sampling (MCS) that introduces multi-view consistency constraints in the reverse sampling process of diffusion models. Experimental results demonstrate that without training or fine-tuning existing diffusion models, VistaDream achieves consistent and high-quality novel view synthesis using just single-view images and outperforms baseline methods by a large margin. The code, videos, and interactive demos are available at https://vistadream-project-page.github.io/.
comment: Project Page: https://vistadream-project-page.github.io/
☆ Network Inversion for Training-Like Data Reconstruction
Machine Learning models are often trained on proprietary and private data that cannot be shared, though the trained models themselves are distributed openly assuming that sharing model weights is privacy preserving, as training data is not expected to be inferred from the model weights. In this paper, we present Training-Like Data Reconstruction (TLDR), a network inversion-based approach to reconstruct training-like data from trained models. To begin with, we introduce a comprehensive network inversion technique that learns the input space corresponding to different classes in the classifier using a single conditioned generator. While inversion may typically return random and arbitrary input images for a given output label, we modify the inversion process to incentivize the generator to reconstruct training-like data by exploiting key properties of the classifier with respect to the training data along with some prior knowledge about the images. To validate our approach, we conduct empirical evaluations on multiple standard vision classification datasets, thereby highlighting the potential privacy risks involved in sharing machine learning models.
☆ Rethinking generalization of classifiers in separable classes scenarios and over-parameterized regimes
We investigate the learning dynamics of classifiers in scenarios where classes are separable or classifiers are over-parameterized. In both cases, Empirical Risk Minimization (ERM) results in zero training error. However, there are many global minima with a training error of zero, some of which generalize well and some of which do not. We show that in separable classes scenarios the proportion of "bad" global minima diminishes exponentially with the number of training data n. Our analysis provides bounds and learning curves dependent solely on the density distribution of the true error for the given classifier function set, irrespective of the set's size or complexity (e.g., number of parameters). This observation may shed light on the unexpectedly good generalization of over-parameterized Neural Networks. For the over-parameterized scenario, we propose a model for the density distribution of the true error, yielding learning curves that align with experiments on MNIST and CIFAR-10.
☆ Nash Meets Wertheimer: Using Good Continuation in Jigsaw Puzzles ACCV2024
Jigsaw puzzle solving is a challenging task for computer vision since it requires high-level spatial and semantic reasoning. To solve the problem, existing approaches invariably use color and/or shape information but in many real-world scenarios, such as in archaeological fresco reconstruction, this kind of clues is often unreliable due to severe physical and pictorial deterioration of the individual fragments. This makes state-of-the-art approaches entirely unusable in practice. On the other hand, in such cases, simple geometrical patterns such as lines or curves offer a powerful yet unexplored clue. In an attempt to fill in this gap, in this paper we introduce a new challenging version of the puzzle solving problem in which one deliberately ignores conventional color and shape features and relies solely on the presence of linear geometrical patterns. The reconstruction process is then only driven by one of the most fundamental principles of Gestalt perceptual organization, namely Wertheimer's {\em law of good continuation}. In order to tackle this problem, we formulate the puzzle solving problem as the problem of finding a Nash equilibrium of a (noncooperative) multiplayer game and use classical multi-population replicator dynamics to solve it. The proposed approach is general and allows us to deal with pieces of arbitrary shape, size and orientation. We evaluate our approach on both synthetic and real-world data and compare it with state-of-the-art algorithms. The results show the intrinsic complexity of our purely line-based puzzle problem as well as the relative effectiveness of our game-theoretic formulation.
comment: to be published in ACCV2024
☆ Bridging the Modality Gap: Dimension Information Alignment and Sparse Spatial Constraint for Image-Text Matching
Many contrastive learning based models have achieved advanced performance in image-text matching tasks. The key of these models lies in analyzing the correlation between image-text pairs, which involves cross-modal interaction of embeddings in corresponding dimensions. However, the embeddings of different modalities are from different models or modules, and there is a significant modality gap. Directly interacting such embeddings lacks rationality and may capture inaccurate correlation. Therefore, we propose a novel method called DIAS to bridge the modality gap from two aspects: (1) We align the information representation of embeddings from different modalities in corresponding dimension to ensure the correlation calculation is based on interactions of similar information. (2) The spatial constraints of inter- and intra-modalities unmatched pairs are introduced to ensure the effectiveness of semantic alignment of the model. Besides, a sparse correlation algorithm is proposed to select strong correlated spatial relationships, enabling the model to learn more significant features and avoid being misled by weak correlation. Extensive experiments demonstrate the superiority of DIAS, achieving 4.3\%-10.2\% rSum improvements on Flickr30k and MSCOCO benchmarks.
☆ MPDS: A Movie Posters Dataset for Image Generation with Diffusion Model
Movie posters are vital for captivating audiences, conveying themes, and driving market competition in the film industry. While traditional designs are laborious, intelligent generation technology offers efficiency gains and design enhancements. Despite exciting progress in image generation, current models often fall short in producing satisfactory poster results. The primary issue lies in the absence of specialized poster datasets for targeted model training. In this work, we propose a Movie Posters DataSet (MPDS), tailored for text-to-image generation models to revolutionize poster production. As dedicated to posters, MPDS stands out as the first image-text pair dataset to our knowledge, composing of 373k+ image-text pairs and 8k+ actor images (covering 4k+ actors). Detailed poster descriptions, such as movie titles, genres, casts, and synopses, are meticulously organized and standardized based on public movie synopsis, also named movie-synopsis prompt. To bolster poster descriptions as well as reduce differences from movie synopsis, further, we leverage a large-scale vision-language model to automatically produce vision-perceptive prompts for each poster, then perform manual rectification and integration with movie-synopsis prompt. In addition, we introduce a prompt of poster captions to exhibit text elements in posters like actor names and movie titles. For movie poster generation, we develop a multi-condition diffusion framework that takes poster prompt, poster caption, and actor image (for personalization) as inputs, yielding excellent results through the learning of a diffusion model. Experiments demonstrate the valuable role of our proposed MPDS dataset in advancing personalized movie poster generation. MPDS is available at https://anonymous.4open.science/r/MPDS-373k-BD3B.
☆ PerspectiveNet: Multi-View Perception for Dynamic Scene Understanding
Generating detailed descriptions from multiple cameras and viewpoints is challenging due to the complex and inconsistent nature of visual data. In this paper, we introduce PerspectiveNet, a lightweight yet efficient model for generating long descriptions across multiple camera views. Our approach utilizes a vision encoder, a compact connector module to convert visual features into a fixed-size tensor, and large language models (LLMs) to harness the strong natural language generation capabilities of LLMs. The connector module is designed with three main goals: mapping visual features onto LLM embeddings, emphasizing key information needed for description generation, and producing a fixed-size feature matrix. Additionally, we augment our solution with a secondary task, the correct frame sequence detection, enabling the model to search for the correct sequence of frames to generate descriptions. Finally, we integrate the connector module, the secondary task, the LLM, and a visual feature extraction model into a single architecture, which is trained for the Traffic Safety Description and Analysis task. This task requires generating detailed, fine-grained descriptions of events from multiple cameras and viewpoints. The resulting model is lightweight, ensuring efficient training and inference, while remaining highly effective.
comment: 6 pages, 2 figures
☆ AttriPrompter: Auto-Prompting with Attribute Semantics for Zero-shot Nuclei Detection via Visual-Language Pre-trained Models
Large-scale visual-language pre-trained models (VLPMs) have demonstrated exceptional performance in downstream object detection through text prompts for natural scenes. However, their application to zero-shot nuclei detection on histopathology images remains relatively unexplored, mainly due to the significant gap between the characteristics of medical images and the web-originated text-image pairs used for pre-training. This paper aims to investigate the potential of the object-level VLPM, Grounded Language-Image Pre-training (GLIP), for zero-shot nuclei detection. Specifically, we propose an innovative auto-prompting pipeline, named AttriPrompter, comprising attribute generation, attribute augmentation, and relevance sorting, to avoid subjective manual prompt design. AttriPrompter utilizes VLPMs' text-to-image alignment to create semantically rich text prompts, which are then fed into GLIP for initial zero-shot nuclei detection. Additionally, we propose a self-trained knowledge distillation framework, where GLIP serves as the teacher with its initial predictions used as pseudo labels, to address the challenges posed by high nuclei density, including missed detections, false positives, and overlapping instances. Our method exhibits remarkable performance in label-free nuclei detection, outperforming all existing unsupervised methods and demonstrating excellent generality. Notably, this work highlights the astonishing potential of VLPMs pre-trained on natural image-text pairs for downstream tasks in the medical field as well. Code will be released at https://github.com/wuyongjianCODE/AttriPrompter.
comment: This article has been accepted for publication in a future issue of IEEE Transactions on Medical Imaging (TMI), but has not been fully edited. Content may change prior to final publication. Citation information: DOI: https://doi.org/10.1109/TMI.2024.3473745 . Code: https://github.com/wuyongjianCODE/AttriPrompter
☆ Evaluating the Effectiveness of Attack-Agnostic Features for Morphing Attack Detection
Morphing attacks have diversified significantly over the past years, with new methods based on generative adversarial networks (GANs) and diffusion models posing substantial threats to face recognition systems. Recent research has demonstrated the effectiveness of features extracted from large vision models pretrained on bonafide data only (attack-agnostic features) for detecting deep generative images. Building on this, we investigate the potential of these image representations for morphing attack detection (MAD). We develop supervised detectors by training a simple binary linear SVM on the extracted features and one-class detectors by modeling the distribution of bonafide features with a Gaussian Mixture Model (GMM). Our method is evaluated across a comprehensive set of attacks and various scenarios, including generalization to unseen attacks, different source datasets, and print-scan data. Our results indicate that attack-agnostic features can effectively detect morphing attacks, outperforming traditional supervised and one-class detectors from the literature in most scenarios. Additionally, we provide insights into the strengths and limitations of each considered representation and discuss potential future research directions to further enhance the robustness and generalizability of our approach.
comment: Published in the 2024 IEEE International Joint Conference on Biometrics (IJCB)
☆ One-Step Diffusion Distillation through Score Implicit Matching NeurIPS 2024
Despite their strong performances on many generative tasks, diffusion models require a large number of sampling steps in order to generate realistic samples. This has motivated the community to develop effective methods to distill pre-trained diffusion models into more efficient models, but these methods still typically require few-step inference or perform substantially worse than the underlying model. In this paper, we present Score Implicit Matching (SIM) a new approach to distilling pre-trained diffusion models into single-step generator models, while maintaining almost the same sample generation ability as the original model as well as being data-free with no need of training samples for distillation. The method rests upon the fact that, although the traditional score-based loss is intractable to minimize for generator models, under certain conditions we can efficiently compute the gradients for a wide class of score-based divergences between a diffusion model and a generator. SIM shows strong empirical performances for one-step generators: on the CIFAR10 dataset, it achieves an FID of 2.06 for unconditional generation and 1.96 for class-conditional generation. Moreover, by applying SIM to a leading transformer-based diffusion model, we distill a single-step generator for text-to-image (T2I) generation that attains an aesthetic score of 6.42 with no performance decline over the original multi-step counterpart, clearly outperforming the other one-step generators including SDXL-TURBO of 5.33, SDXL-LIGHTNING of 5.34 and HYPER-SDXL of 5.85. We will release this industry-ready one-step transformer-based T2I generator along with this paper.
comment: Accepted by NeurIPS 2024
☆ The Scene Language: Representing Scenes with Programs, Words, and Embeddings
We introduce the Scene Language, a visual scene representation that concisely and precisely describes the structure, semantics, and identity of visual scenes. It represents a scene with three key components: a program that specifies the hierarchical and relational structure of entities in the scene, words in natural language that summarize the semantic class of each entity, and embeddings that capture the visual identity of each entity. This representation can be inferred from pre-trained language models via a training-free inference technique, given text or image inputs. The resulting scene can be rendered into images using traditional, neural, or hybrid graphics renderers. Together, this forms a robust, automated system for high-quality 3D and 4D scene generation. Compared with existing representations like scene graphs, our proposed Scene Language generates complex scenes with higher fidelity, while explicitly modeling the scene structures to enable precise control and editing.
comment: Project page: https://ai.stanford.edu/~yzzhang/projects/scene-language/
☆ DSORT-MCU: Detecting Small Objects in Real-Time on Microcontroller Units
Advances in lightweight neural networks have revolutionized computer vision in a broad range of IoT applications, encompassing remote monitoring and process automation. However, the detection of small objects, which is crucial for many of these applications, remains an underexplored area in current computer vision research, particularly for low-power embedded devices that host resource-constrained processors. To address said gap, this paper proposes an adaptive tiling method for lightweight and energy-efficient object detection networks, including YOLO-based models and the popular FOMO network. The proposed tiling enables object detection on low-power MCUs with no compromise on accuracy compared to large-scale detection models. The benefit of the proposed method is demonstrated by applying it to FOMO and TinyissimoYOLO networks on a novel RISC-V-based MCU with built-in ML accelerators. Extensive experimental results show that the proposed tiling method boosts the F1-score by up to 225% for both FOMO and TinyissimoYOLO networks while reducing the average object count error by up to 76% with FOMO and up to 89% for TinyissimoYOLO. Furthermore, the findings of this work indicate that using a soft F1 loss over the popular binary cross-entropy loss can serve as an implicit non-maximum suppression for the FOMO network. To evaluate the real-world performance, the networks are deployed on the RISC-V based GAP9 microcontroller from GreenWaves Technologies, showcasing the proposed method's ability to strike a balance between detection performance ($58% - 95%$ F1 score), low latency (0.6 ms/Inference - 16.2 ms/Inference}), and energy efficiency (31 uJ/Inference} - 1.27 mJ/Inference) while performing multiple predictions using high-resolution images on a MCU.
comment: arXiv admin note: text overlap with arXiv:2311.07163
☆ SpikMamba: When SNN meets Mamba in Event-based Human Action Recognition
Human action recognition (HAR) plays a key role in various applications such as video analysis, surveillance, autonomous driving, robotics, and healthcare. Most HAR algorithms are developed from RGB images, which capture detailed visual information. However, these algorithms raise concerns in privacy-sensitive environments due to the recording of identifiable features. Event cameras offer a promising solution by capturing scene brightness changes sparsely at the pixel level, without capturing full images. Moreover, event cameras have high dynamic ranges that can effectively handle scenarios with complex lighting conditions, such as low light or high contrast environments. However, using event cameras introduces challenges in modeling the spatially sparse and high temporal resolution event data for HAR. To address these issues, we propose the SpikMamba framework, which combines the energy efficiency of spiking neural networks and the long sequence modeling capability of Mamba to efficiently capture global features from spatially sparse and high a temporal resolution event data. Additionally, to improve the locality of modeling, a spiking window-based linear attention mechanism is used. Extensive experiments show that SpikMamba achieves remarkable recognition performance, surpassing the previous state-of-the-art by 1.45%, 7.22%, 0.15%, and 3.92% on the PAF, HARDVS, DVS128, and E-FAction datasets, respectively. The code is available at https://github.com/Typistchen/SpikMamba.
comment: 8 pages, 4 figures
☆ Time-Resolved MNIST Dataset for Single-Photon Recognition ECCV 2024
Time-resolved single photon imaging is a promising imaging modality characterized by the unique capability of timestamping the arrivals of single photons. Single-Photon Avalanche Diodes (SPADs) are the leading technology for implementing modern time-resolved pixels, suitable for passive imaging with asynchronous readout. However, they are currently limited to small sized arrays, thus there is a lack of datasets for passive time-resolved SPAD imaging, which in turn hinders research on this peculiar imaging data. In this paper we describe a realistic simulation process for SPAD imaging, which takes into account both the stochastic nature of photon arrivals and all the noise sources involved in the acquisition process of time-resolved SPAD arrays. We have implemented this simulator in a software prototype able to generate arbitrary-sized time-resolved SPAD arrays operating in passive mode. Starting from a reference image, our simulator generates a realistic stream of timestamped photon detections. We use our simulator to generate a time-resolved version of MNIST, which we make publicly available. Our dataset has the purpose of encouraging novel research directions in time-resolved SPAD imaging, as well as investigating the performance of CNN classifiers in extremely low-light conditions.
comment: 12 pages, 4 figures. Accepted for Workshop on Synthetic Data for Computer Vision at ECCV 2024
☆ Polyp-E: Benchmarking the Robustness of Deep Segmentation Models via Polyp Editing
Automatic polyp segmentation is helpful to assist clinical diagnosis and treatment. In daily clinical practice, clinicians exhibit robustness in identifying polyps with both location and size variations. It is uncertain if deep segmentation models can achieve comparable robustness in automated colonoscopic analysis. To benchmark the model robustness, we focus on evaluating the robustness of segmentation models on the polyps with various attributes (e.g. location and size) and healthy samples. Based on the Latent Diffusion Model, we perform attribute editing on real polyps and build a new dataset named Polyp-E. Our synthetic dataset boasts exceptional realism, to the extent that clinical experts find it challenging to discern them from real data. We evaluate several existing polyp segmentation models on the proposed benchmark. The results reveal most of the models are highly sensitive to attribute variations. As a novel data augmentation technique, the proposed editing pipeline can improve both in-distribution and out-of-distribution generalization ability. The code and datasets will be released.
☆ Progressive Compositionality In Text-to-Image Generative Models
Despite the impressive text-to-image (T2I) synthesis capabilities of diffusion models, they often struggle to understand compositional relationships between objects and attributes, especially in complex settings. Existing solutions have tackled these challenges by optimizing the cross-attention mechanism or learning from the caption pairs with minimal semantic changes. However, can we generate high-quality complex contrastive images that diffusion models can directly discriminate based on visual representations? In this work, we leverage large-language models (LLMs) to compose realistic, complex scenarios and harness Visual-Question Answering (VQA) systems alongside diffusion models to automatically curate a contrastive dataset, ConPair, consisting of 15k pairs of high-quality contrastive images. These pairs feature minimal visual discrepancies and cover a wide range of attribute categories, especially complex and natural scenarios. To learn effectively from these error cases, i.e., hard negative images, we propose EvoGen, a new multi-stage curriculum for contrastive learning of diffusion models. Through extensive experiments across a wide range of compositional scenarios, we showcase the effectiveness of our proposed framework on compositional T2I benchmarks.
☆ Development of CNN Architectures using Transfer Learning Methods for Medical Image Classification
The application of deep learning-based architecture has seen a tremendous rise in recent years. For example, medical image classification using deep learning achieved breakthrough results. Convolutional Neural Networks (CNNs) are implemented predominantly in medical image classification and segmentation. On the other hand, transfer learning has emerged as a prominent supporting tool for enhancing the efficiency and accuracy of deep learning models. This paper investigates the development of CNN architectures using transfer learning techniques in the field of medical image classification using a timeline mapping model for key image classification challenges. Our findings help make an informed decision while selecting the optimum and state-of-the-art CNN architectures.
☆ DI-MaskDINO: A Joint Object Detection and Instance Segmentation Model
This paper is motivated by an interesting phenomenon: the performance of object detection lags behind that of instance segmentation (i.e., performance imbalance) when investigating the intermediate results from the beginning transformer decoder layer of MaskDINO (i.e., the SOTA model for joint detection and segmentation). This phenomenon inspires us to think about a question: will the performance imbalance at the beginning layer of transformer decoder constrain the upper bound of the final performance? With this question in mind, we further conduct qualitative and quantitative pre-experiments, which validate the negative impact of detection-segmentation imbalance issue on the model performance. To address this issue, this paper proposes DI-MaskDINO model, the core idea of which is to improve the final performance by alleviating the detection-segmentation imbalance. DI-MaskDINO is implemented by configuring our proposed De-Imbalance (DI) module and Balance-Aware Tokens Optimization (BATO) module to MaskDINO. DI is responsible for generating balance-aware query, and BATO uses the balance-aware query to guide the optimization of the initial feature tokens. The balance-aware query and optimized feature tokens are respectively taken as the Query and Key&Value of transformer decoder to perform joint object detection and instance segmentation. DI-MaskDINO outperforms existing joint object detection and instance segmentation models on COCO and BDD100K benchmarks, achieving +1.2 $AP^{box}$ and +0.9 $AP^{mask}$ improvements compared to SOTA joint detection and segmentation model MaskDINO. In addition, DI-MaskDINO also obtains +1.0 $AP^{box}$ improvement compared to SOTA object detection model DINO and +3.0 $AP^{mask}$ improvement compared to SOTA segmentation model Mask2Former.
comment: 16 pages, 3 figures, Conference on Neural Information Processing Systems
☆ MPT: A Large-scale Multi-Phytoplankton Tracking Benchmark
Phytoplankton are a crucial component of aquatic ecosystems, and effective monitoring of them can provide valuable insights into ocean environments and ecosystem changes. Traditional phytoplankton monitoring methods are often complex and lack timely analysis. Therefore, deep learning algorithms offer a promising approach for automated phytoplankton monitoring. However, the lack of large-scale, high-quality training samples has become a major bottleneck in advancing phytoplankton tracking. In this paper, we propose a challenging benchmark dataset, Multiple Phytoplankton Tracking (MPT), which covers diverse background information and variations in motion during observation. The dataset includes 27 species of phytoplankton and zooplankton, 14 different backgrounds to simulate diverse and complex underwater environments, and a total of 140 videos. To enable accurate real-time observation of phytoplankton, we introduce a multi-object tracking method, Deviation-Corrected Multi-Scale Feature Fusion Tracker(DSFT), which addresses issues such as focus shifts during tracking and the loss of small target information when computing frame-to-frame similarity. Specifically, we introduce an additional feature extractor to predict the residuals of the standard feature extractor's output, and compute multi-scale frame-to-frame similarity based on features from different layers of the extractor. Extensive experiments on the MPT have demonstrated the validity of the dataset and the superiority of DSFT in tracking phytoplankton, providing an effective solution for phytoplankton monitoring.
☆ NucleiMix: Realistic Data Augmentation for Nuclei Instance Segmentation
Nuclei instance segmentation is an essential task in pathology image analysis, serving as the foundation for many downstream applications. The release of several public datasets has significantly advanced research in this area, yet many existing methods struggle with data imbalance issues. To address this challenge, this study introduces a data augmentation method, called NucleiMix, which is designed to balance the distribution of nuclei types by increasing the number of rare-type nuclei within datasets. NucleiMix operates in two phases. In the first phase, it identifies candidate locations similar to the surroundings of rare-type nuclei and inserts rare-type nuclei into the candidate locations. In the second phase, it employs a progressive inpainting strategy using a pre-trained diffusion model to seamlessly integrate rare-type nuclei into their new environments in replacement of major-type nuclei or background locations. We systematically evaluate the effectiveness of NucleiMix on three public datasets using two popular nuclei instance segmentation models. The results demonstrate the superior ability of NucleiMix to synthesize realistic rare-type nuclei and to enhance the quality of nuclei segmentation and classification in an accurate and robust manner.
☆ Visual Question Answering in Ophthalmology: A Progressive and Practical Perspective
Accurate diagnosis of ophthalmic diseases relies heavily on the interpretation of multimodal ophthalmic images, a process often time-consuming and expertise-dependent. Visual Question Answering (VQA) presents a potential interdisciplinary solution by merging computer vision and natural language processing to comprehend and respond to queries about medical images. This review article explores the recent advancements and future prospects of VQA in ophthalmology from both theoretical and practical perspectives, aiming to provide eye care professionals with a deeper understanding and tools for leveraging the underlying models. Additionally, we discuss the promising trend of large language models (LLM) in enhancing various components of the VQA framework to adapt to multimodal ophthalmic tasks. Despite the promising outlook, ophthalmic VQA still faces several challenges, including the scarcity of annotated multimodal image datasets, the necessity of comprehensive and unified evaluation methods, and the obstacles to achieving effective real-world applications. This article highlights these challenges and clarifies future directions for advancing ophthalmic VQA with LLMs. The development of LLM-based ophthalmic VQA systems calls for collaborative efforts between medical professionals and AI experts to overcome existing obstacles and advance the diagnosis and care of eye diseases.
☆ Dual-Model Defense: Safeguarding Diffusion Models from Membership Inference Attacks through Disjoint Data Splitting
Diffusion models have demonstrated remarkable capabilities in image synthesis, but their recently proven vulnerability to Membership Inference Attacks (MIAs) poses a critical privacy concern. This paper introduces two novel and efficient approaches (DualMD and DistillMD) to protect diffusion models against MIAs while maintaining high utility. Both methods are based on training two separate diffusion models on disjoint subsets of the original dataset. DualMD then employs a private inference pipeline that utilizes both models. This strategy significantly reduces the risk of black-box MIAs by limiting the information any single model contains about individual training samples. The dual models can also generate "soft targets" to train a private student model in DistillMD, enhancing privacy guarantees against all types of MIAs. Extensive evaluations of DualMD and DistillMD against state-of-the-art MIAs across various datasets in white-box and black-box settings demonstrate their effectiveness in substantially reducing MIA success rates while preserving competitive image generation performance. Notably, our experiments reveal that DistillMD not only defends against MIAs but also mitigates model memorization, indicating that both vulnerabilities stem from overfitting and can be addressed simultaneously with our unified approach.
☆ TopoDiffusionNet: A Topology-aware Diffusion Model
Diffusion models excel at creating visually impressive images but often struggle to generate images with a specified topology. The Betti number, which represents the number of structures in an image, is a fundamental measure in topology. Yet, diffusion models fail to satisfy even this basic constraint. This limitation restricts their utility in applications requiring exact control, like robotics and environmental modeling. To address this, we propose TopoDiffusionNet (TDN), a novel approach that enforces diffusion models to maintain the desired topology. We leverage tools from topological data analysis, particularly persistent homology, to extract the topological structures within an image. We then design a topology-based objective function to guide the denoising process, preserving intended structures while suppressing noisy ones. Our experiments across four datasets demonstrate significant improvements in topological accuracy. TDN is the first to integrate topology with diffusion models, opening new avenues of research in this area.
comment: 20 pages, 11 figures, 7 tables
☆ Fire and Smoke Detection with Burning Intensity Representation
An effective Fire and Smoke Detection (FSD) and analysis system is of paramount importance due to the destructive potential of fire disasters. However, many existing FSD methods directly employ generic object detection techniques without considering the transparency of fire and smoke, which leads to imprecise localization and reduces detection performance. To address this issue, a new Attentive Fire and Smoke Detection Model (a-FSDM) is proposed. This model not only retains the robust feature extraction and fusion capabilities of conventional detection algorithms but also redesigns the detection head specifically for transparent targets in FSD, termed the Attentive Transparency Detection Head (ATDH). In addition, Burning Intensity (BI) is introduced as a pivotal feature for fire-related downstream risk assessments in traditional FSD methodologies. Extensive experiments on multiple FSD datasets showcase the effectiveness and versatility of the proposed FSD model. The project is available at \href{https://xiaoyihan6.github.io/FSD/}{https://xiaoyihan6.github.io/FSD/}.
☆ Benchmarking Multi-Scene Fire and Smoke Detection
The current irregularities in existing public Fire and Smoke Detection (FSD) datasets have become a bottleneck in the advancement of FSD technology. Upon in-depth analysis, we identify the core issue as the lack of standardized dataset construction, uniform evaluation systems, and clear performance benchmarks. To address this issue and drive innovation in FSD technology, we systematically gather diverse resources from public sources to create a more comprehensive and refined FSD benchmark. Additionally, recognizing the inadequate coverage of existing dataset scenes, we strategically expand scenes, relabel, and standardize existing public FSD datasets to ensure accuracy and consistency. We aim to establish a standardized, realistic, unified, and efficient FSD research platform that mirrors real-life scenes closely. Through our efforts, we aim to provide robust support for the breakthrough and development of FSD technology. The project is available at \href{https://xiaoyihan6.github.io/FSD/}{https://xiaoyihan6.github.io/FSD/}.
☆ EVC-MF: End-to-end Video Captioning Network with Multi-scale Features
Conventional approaches for video captioning leverage a variety of offline-extracted features to generate captions. Despite the availability of various offline-feature-extractors that offer diverse information from different perspectives, they have several limitations due to fixed parameters. Concretely, these extractors are solely pre-trained on image/video comprehension tasks, making them less adaptable to video caption datasets. Additionally, most of these extractors only capture features prior to the classifier of the pre-training task, ignoring a significant amount of valuable shallow information. Furthermore, employing multiple offline-features may introduce redundant information. To address these issues, we propose an end-to-end encoder-decoder-based network (EVC-MF) for video captioning, which efficiently utilizes multi-scale visual and textual features to generate video descriptions. Specifically, EVC-MF consists of three modules. Firstly, instead of relying on multiple feature extractors, we directly feed video frames into a transformer-based network to obtain multi-scale visual features and update feature extractor parameters. Secondly, we fuse the multi-scale features and input them into a masked encoder to reduce redundancy and encourage learning useful features. Finally, we utilize an enhanced transformer-based decoder, which can efficiently leverage shallow textual information, to generate video descriptions. To evaluate our proposed model, we conduct extensive experiments on benchmark datasets. The results demonstrate that EVC-MF yields competitive performance compared with the state-of-theart methods.
☆ Foundation Models for Remote Sensing and Earth Observation: A Survey
Remote Sensing (RS) is a crucial technology for observing, monitoring, and interpreting our planet, with broad applications across geoscience, economics, humanitarian fields, etc. While artificial intelligence (AI), particularly deep learning, has achieved significant advances in RS, unique challenges persist in developing more intelligent RS systems, including the complexity of Earth's environments, diverse sensor modalities, distinctive feature patterns, varying spatial and spectral resolutions, and temporal dynamics. Meanwhile, recent breakthroughs in large Foundation Models (FMs) have expanded AI's potential across many domains due to their exceptional generalizability and zero-shot transfer capabilities. However, their success has largely been confined to natural data like images and video, with degraded performance and even failures for RS data of various non-optical modalities. This has inspired growing interest in developing Remote Sensing Foundation Models (RSFMs) to address the complex demands of Earth Observation (EO) tasks, spanning the surface, atmosphere, and oceans. This survey systematically reviews the emerging field of RSFMs. It begins with an outline of their motivation and background, followed by an introduction of their foundational concepts. It then categorizes and reviews existing RSFM studies including their datasets and technical contributions across Visual Foundation Models (VFMs), Visual-Language Models (VLMs), Large Language Models (LLMs), and beyond. In addition, we benchmark these models against publicly available datasets, discuss existing challenges, and propose future research directions in this rapidly evolving field.
☆ LongVU: Spatiotemporal Adaptive Compression for Long Video-Language Understanding
Multimodal Large Language Models (MLLMs) have shown promising progress in understanding and analyzing video content. However, processing long videos remains a significant challenge constrained by LLM's context size. To address this limitation, we propose LongVU, a spatiotemporal adaptive compression mechanism thats reduces the number of video tokens while preserving visual details of long videos. Our idea is based on leveraging cross-modal query and inter-frame dependencies to adaptively reduce temporal and spatial redundancy in videos. Specifically, we leverage DINOv2 features to remove redundant frames that exhibit high similarity. Then we utilize text-guided cross-modal query for selective frame feature reduction. Further, we perform spatial token reduction across frames based on their temporal dependencies. Our adaptive compression strategy effectively processes a large number of frames with little visual information loss within given context length. Our LongVU consistently surpass existing methods across a variety of video understanding benchmarks, especially on hour-long video understanding tasks such as VideoMME and MLVU. Given a light-weight LLM, our LongVU also scales effectively into a smaller size with state-of-the-art video understanding performance.
comment: Project page: https://vision-cair.github.io/LongVU
☆ SigCLR: Sigmoid Contrastive Learning of Visual Representations
We propose SigCLR: Sigmoid Contrastive Learning of Visual Representations. SigCLR utilizes the logistic loss that only operates on pairs and does not require a global view as in the cross-entropy loss used in SimCLR. We show that logistic loss shows competitive performance on CIFAR-10, CIFAR-100, and Tiny-IN compared to other established SSL objectives. Our findings verify the importance of learnable bias as in the case of SigLUP, however, it requires a fixed temperature as in the SimCLR to excel. Overall, SigCLR is a promising replacement for the SimCLR which is ubiquitous and has shown tremendous success in various domains.
comment: Neurips 2024 SSL Workshop
☆ AG-SLAM: Active Gaussian Splatting SLAM
We present AG-SLAM, the first active SLAM system utilizing 3D Gaussian Splatting (3DGS) for online scene reconstruction. In recent years, radiance field scene representations, including 3DGS have been widely used in SLAM and exploration, but actively planning trajectories for robotic exploration is still unvisited. In particular, many exploration methods assume precise localization and thus do not mitigate the significant risk of constructing a trajectory, which is difficult for a SLAM system to operate on. This can cause camera tracking failure and lead to failures in real-world robotic applications. Our method leverages Fisher Information to balance the dual objectives of maximizing the information gain for the environment while minimizing the cost of localization errors. Experiments conducted on the Gibson and Habitat-Matterport 3D datasets demonstrate state-of-the-art results of the proposed method.
☆ Geometric Graph Neural Network Modeling of Human Interactions in Crowded Environments
Modeling human trajectories in crowded environments is challenging due to the complex nature of pedestrian behavior and interactions. This paper proposes a geometric graph neural network (GNN) architecture that integrates domain knowledge from psychological studies to model pedestrian interactions and predict future trajectories. Unlike prior studies using complete graphs, we define interaction neighborhoods using pedestrians' field of view, motion direction, and distance-based kernel functions to construct graph representations of crowds. Evaluations across multiple datasets demonstrate improved prediction accuracy through reduced average and final displacement error metrics. Our findings underscore the importance of integrating domain knowledge with data-driven approaches for effective modeling of human interactions in crowds.
comment: \c{opyright} 2024 the authors. This work has been accepted to IFAC for publication under a Creative Commons Licence CC-BY-NC-ND
☆ Efficient Feature Extraction Using Light-Weight CNN Attention-Based Deep Learning Architectures for Ultrasound Fetal Plane Classification
Ultrasound fetal imaging is beneficial to support prenatal development because it is affordable and non-intrusive. Nevertheless, fetal plane classification (FPC) remains challenging and time-consuming for obstetricians since it depends on nuanced clinical aspects, which increases the difficulty in identifying relevant features of the fetal anatomy. Thus, to assist with its accurate feature extraction, a lightweight artificial intelligence architecture leveraging convolutional neural networks and attention mechanisms is proposed to classify the largest benchmark ultrasound dataset. The approach fine-tunes from lightweight EfficientNet feature extraction backbones pre-trained on the ImageNet1k. to classify key fetal planes such as the brain, femur, thorax, cervix, and abdomen. Our methodology incorporates the attention mechanism to refine features and 3-layer perceptrons for classification, achieving superior performance with the highest Top-1 accuracy of 96.25%, Top-2 accuracy of 99.80% and F1-Score of 0.9576. Importantly, the model has 40x fewer trainable parameters than existing benchmark ensemble or transformer pipelines, facilitating easy deployment on edge devices to help clinical practitioners with real-time FPC. The findings are also interpreted using GradCAM to carry out clinical correlation to aid doctors with diagnostics and improve treatment plans for expectant mothers.
comment: Submitted to Computers in Biology and Medicine journal
☆ Denoise-I2W: Mapping Images to Denoising Words for Accurate Zero-Shot Composed Image Retrieval IJCAI 2024
Zero-Shot Composed Image Retrieval (ZS-CIR) supports diverse tasks with a broad range of visual content manipulation intentions that can be related to domain, scene, object, and attribute. A key challenge for ZS-CIR is to accurately map image representation to a pseudo-word token that captures the manipulation intention relevant image information for generalized CIR. However, existing methods between the retrieval and pre-training stages lead to significant redundancy in the pseudo-word tokens. In this paper, we propose a novel denoising image-to-word mapping approach, named Denoise-I2W, for mapping images into denoising pseudo-word tokens that, without intention-irrelevant visual information, enhance accurate ZS-CIR. Specifically, a pseudo triplet construction module first automatically constructs pseudo triples (\textit{i.e.,} a pseudo-reference image, a pseudo-manipulation text, and a target image) for pre-training the denoising mapping network. Then, a pseudo-composed mapping module maps the pseudo-reference image to a pseudo-word token and combines it with the pseudo-manipulation text with manipulation intention. This combination aligns with the target image, facilitating denoising intention-irrelevant visual information for mapping. Our proposed Denoise-I2W is a model-agnostic and annotation-free approach. It demonstrates strong generalization capabilities across three state-of-the-art ZS-CIR models on four benchmark datasets. By integrating Denoise-I2W with existing best models, we obtain consistent and significant performance boosts ranging from 1.45\% to 4.17\% over the best methods without increasing inference costs. and achieve new state-of-the-art results on ZS-CIR. Our code is available at \url{https://github.com/Pter61/denoise-i2w-tmm}.
comment: This work was submitted to IJCAI 2024, with a score of weak accept and borderline accept
☆ Do Vision-Language Models Represent Space and How? Evaluating Spatial Frame of Reference Under Ambiguities NeurIPS 2024
Spatial expressions in situated communication can be ambiguous, as their meanings vary depending on the frames of reference (FoR) adopted by speakers and listeners. While spatial language understanding and reasoning by vision-language models (VLMs) have gained increasing attention, potential ambiguities in these models are still under-explored. To address this issue, we present the COnsistent Multilingual Frame Of Reference Test (COMFORT), an evaluation protocol to systematically assess the spatial reasoning capabilities of VLMs. We evaluate nine state-of-the-art VLMs using COMFORT. Despite showing some alignment with English conventions in resolving ambiguities, our experiments reveal significant shortcomings of VLMs: notably, the models (1) exhibit poor robustness and consistency, (2) lack the flexibility to accommodate multiple FoRs, and (3) fail to adhere to language-specific or culture-specific conventions in cross-lingual tests, as English tends to dominate other languages. With a growing effort to align vision-language models with human cognitive intuitions, we call for more attention to the ambiguous nature and cross-cultural diversity of spatial reasoning.
comment: Accepted to Pluralistic Alignment @ NeurIPS 2024 | Project page: https://spatial-comfort.github.io/
☆ PtychoFormer: A Transformer-based Model for Ptychographic Phase Retrieval
Ptychography is a computational method of microscopy that recovers high-resolution transmission images of samples from a series of diffraction patterns. While conventional phase retrieval algorithms can iteratively recover the images, they require oversampled diffraction patterns, incur significant computational costs, and struggle to recover the absolute phase of the sample's transmission function. Deep learning algorithms for ptychography are a promising approach to resolving the limitations of iterative algorithms. We present PtychoFormer, a hierarchical transformer-based model for data-driven single-shot ptychographic phase retrieval. PtychoFormer processes subsets of diffraction patterns, generating local inferences that are seamlessly stitched together to produce a high-quality reconstruction. Our model exhibits tolerance to sparsely scanned diffraction patterns and achieves up to 3600 times faster imaging speed than the extended ptychographic iterative engine (ePIE). We also propose the extended-PtychoFormer (ePF), a hybrid approach that combines the benefits of PtychoFormer with the ePIE. ePF minimizes global phase shifts and significantly enhances reconstruction quality, achieving state-of-the-art phase retrieval in ptychography.
comment: 20 pages, 12 figures
☆ Image-aware Evaluation of Generated Medical Reports
The paper proposes a novel evaluation metric for automatic medical report generation from X-ray images, VLScore. It aims to overcome the limitations of existing evaluation methods, which either focus solely on textual similarities, ignoring clinical aspects, or concentrate only on a single clinical aspect, the pathology, neglecting all other factors. The key idea of our metric is to measure the similarity between radiology reports while considering the corresponding image. We demonstrate the benefit of our metric through evaluation on a dataset where radiologists marked errors in pairs of reports, showing notable alignment with radiologists' judgments. In addition, we provide a new dataset for evaluating metrics. This dataset includes well-designed perturbations that distinguish between significant modifications (e.g., removal of a diagnosis) and insignificant ones. It highlights the weaknesses in current evaluation metrics and provides a clear framework for analysis.
☆ Offline Evaluation of Set-Based Text-to-Image Generation
Text-to-Image (TTI) systems often support people during ideation, the early stages of a creative process when exposure to a broad set of relevant images can help explore the design space. Since ideation is an important subclass of TTI tasks, understanding how to quantitatively evaluate TTI systems according to how well they support ideation is crucial to promoting research and development for these users. However, existing evaluation metrics for TTI remain focused on distributional similarity metrics like Fr\'echet Inception Distance (FID). We take an alternative approach and, based on established methods from ranking evaluation, develop TTI evaluation metrics with explicit models of how users browse and interact with sets of spatially arranged generated images. Our proposed offline evaluation metrics for TTI not only capture how relevant generated images are with respect to the user's ideation need but also take into consideration the diversity and arrangement of the set of generated images. We analyze our proposed family of TTI metrics using human studies on image grids generated by three different TTI systems based on subsets of the widely used benchmarks such as MS-COCO captions and Localized Narratives as well as prompts used in naturalistic settings. Our results demonstrate that grounding metrics in how people use systems is an important and understudied area of benchmark design.
♻ ☆ Typography Leads Semantic Diversifying: Amplifying Adversarial Transferability across Multimodal Large Language Models
Recently, Multimodal Large Language Models (MLLMs) achieve remarkable performance in numerous zero-shot tasks due to their outstanding cross-modal interaction and comprehension abilities. However, MLLMs are found to still be vulnerable to human-imperceptible adversarial examples. In the exploration of security vulnerabilities in real-world scenarios, transferability, which can achieve cross-model impact, is considered the greatest threat posed by adversarial examples. However, there is currently no systematic research on the threat of cross-MLLMs adversarial transferability. Therefore, this paper as the first step to provide a comprehensive evaluation of the transferability of adversarial examples generated by various MLLMs. Furthermore, leveraging two key factors that influence transferability performance: 1) The strength of information diversity involved in the adversarial generation process; 2) Editing across vision-language modality information. We propose a boosting method called Typography Augment Transferability Method (TATM) to investigate the adversarial transferability performance across MLLMs further. Through extensive experimental validation, our TATM demonstrates exceptional performance in real-world applications of "Harmful Word Insertion" and "Important Information Protection".
♻ ☆ AIM 2024 Challenge on Compressed Video Quality Assessment: Methods and Results
Video quality assessment (VQA) is a crucial task in the development of video compression standards, as it directly impacts the viewer experience. This paper presents the results of the Compressed Video Quality Assessment challenge, held in conjunction with the Advances in Image Manipulation (AIM) workshop at ECCV 2024. The challenge aimed to evaluate the performance of VQA methods on a diverse dataset of 459 videos, encoded with 14 codecs of various compression standards (AVC/H.264, HEVC/H.265, AV1, and VVC/H.266) and containing a comprehensive collection of compression artifacts. To measure the methods performance, we employed traditional correlation coefficients between their predictions and subjective scores, which were collected via large-scale crowdsourced pairwise human comparisons. For training purposes, participants were provided with the Compressed Video Quality Assessment Dataset (CVQAD), a previously developed dataset of 1022 videos. Up to 30 participating teams registered for the challenge, while we report the results of 6 teams, which submitted valid final solutions and code for reproducing the results. Moreover, we calculated and present the performance of state-of-the-art VQA methods on the developed dataset, providing a comprehensive benchmark for future research. The dataset, results, and online leaderboard are publicly available at https://challenges.videoprocessing.ai/challenges/compressedvideo-quality-assessment.html.
♻ ☆ Oryx MLLM: On-Demand Spatial-Temporal Understanding at Arbitrary Resolution
Visual data comes in various forms, ranging from small icons of just a few pixels to long videos spanning hours. Existing multi-modal LLMs usually standardize these diverse visual inputs to a fixed resolution for visual encoders and yield similar numbers of tokens for LLMs. This approach is non-optimal for multimodal understanding and inefficient for processing inputs with long and short visual contents. To solve the problem, we propose Oryx, a unified multimodal architecture for the spatial-temporal understanding of images, videos, and multi-view 3D scenes. Oryx offers an on-demand solution to seamlessly and efficiently process visual inputs with arbitrary spatial sizes and temporal lengths through two core innovations: 1) a pre-trained OryxViT model that can encode images at any resolution into LLM-friendly visual representations; 2) a dynamic compressor module that supports 1x to 16x compression on visual tokens by request. These design features enable Oryx to accommodate extremely long visual contexts, such as videos, with lower resolution and high compression while maintaining high recognition precision for tasks like document understanding with native resolution and no compression. Beyond the architectural improvements, enhanced data curation and specialized training on long-context retrieval and spatial-aware data help Oryx achieve strong capabilities in image, video, and 3D multimodal understanding simultaneously. Our work is open-sourced at https://github.com/Oryx-mllm/Oryx.
♻ ☆ NaturalBench: Evaluating Vision-Language Models on Natural Adversarial Samples NeurIPS 24
Vision-language models (VLMs) have made significant progress in recent visual-question-answering (VQA) benchmarks that evaluate complex visio-linguistic reasoning. However, are these models truly effective? In this work, we show that VLMs still struggle with natural images and questions that humans can easily answer, which we term natural adversarial samples. We also find it surprisingly easy to generate these VQA samples from natural image-text corpora using off-the-shelf models like CLIP and ChatGPT. We propose a semi-automated approach to collect a new benchmark, NaturalBench, for reliably evaluating VLMs with 10,000 human-verified VQA samples. Crucially, we adopt a $\textbf{vision-centric}$ design by pairing each question with two images that yield different answers, preventing blind solutions from answering without using the images. This makes NaturalBench more challenging than previous benchmarks that can be solved with commonsense priors. We evaluate 53 state-of-the-art VLMs on NaturalBench, showing that models like LLaVA-OneVision, Cambrian-1, Llama3.2-Vision, Molmo, Qwen2-VL, and even GPT-4o lag 50%-70% behind human performance (over 90%). We analyze why NaturalBench is hard from two angles: (1) Compositionality: Solving NaturalBench requires diverse visio-linguistic skills, including understanding attribute bindings, object relationships, and advanced reasoning like logic and counting. To this end, unlike prior work that uses a single tag per sample, we tag each NaturalBench sample with 1 to 8 skill tags for fine-grained evaluation. (2) Biases: NaturalBench exposes severe biases in VLMs, as models often choose the same answer regardless of the image. Lastly, we apply our benchmark curation method to diverse data sources, including long captions (over 100 words) and non-English languages like Chinese and Hindi, highlighting its potential for dynamic evaluations of VLMs.
comment: Accepted to NeurIPS 24; We open-source our dataset at: https://huggingface.co/datasets/BaiqiL/NaturalBench ; Project page at: https://linzhiqiu.github.io/papers/naturalbench/
♻ ☆ ReCAP: Recursive Cross Attention Network for Pseudo-Label Generation in Robotic Surgical Skill Assessment
In surgical skill assessment, Objective Structured Assessments of Technical Skills (OSATS scores) and the Global Rating Scale (GRS) are established tools for evaluating the performance of surgeons during training. These metrics, coupled with feedback on their performance, enable surgeons to improve and achieve standards of practice. Recent studies on the open-source dataset JIGSAW, which contains both GRS and OSATS labels, have focused on regressing GRS scores from kinematic signals, video data, or a combination of both. In this paper, we argue that regressing the GRS score, a unitless value, by itself is too restrictive, and variations throughout the surgical trial do not hold significant clinical meaning. To address this gap, we developed a recurrent transformer model that outputs the surgeon's performance throughout their training session by relating the model's hidden states to five OSATS scores derived from kinematic signals. These scores are averaged and aggregated to produce a GRS prediction, enabling assessment of the model's performance against the state-of-the-art (SOTA). We report Spearman's Correlation Coefficient (SCC), demonstrating that our model outperforms SOTA models for all tasks, except for Suturing under the leave-one-subject-out (LOSO) scheme (SCC 0.68-0.89), while achieving comparable performance for suturing and across tasks under the leave-one-user-out (LOUO) scheme (SCC 0.45-0.68) and beating SOTA for Needle Passing (0.69). We argue that relating final OSATS scores to short instances throughout a surgeon's procedure is more clinically meaningful than a single GRS score. This approach also allows us to translate quantitative predictions into qualitative feedback, which is crucial for any automated surgical skill assessment pipeline. A senior surgeon validated our model's behaviour and agreed with the semi-supervised predictions 77 \% (p = 0.006) of the time.
♻ ☆ Context and Geometry Aware Voxel Transformer for Semantic Scene Completion NIPS 2024
Vision-based Semantic Scene Completion (SSC) has gained much attention due to its widespread applications in various 3D perception tasks. Existing sparse-to-dense approaches typically employ shared context-independent queries across various input images, which fails to capture distinctions among them as the focal regions of different inputs vary and may result in undirected feature aggregation of cross-attention. Additionally, the absence of depth information may lead to points projected onto the image plane sharing the same 2D position or similar sampling points in the feature map, resulting in depth ambiguity. In this paper, we present a novel context and geometry aware voxel transformer. It utilizes a context aware query generator to initialize context-dependent queries tailored to individual input images, effectively capturing their unique characteristics and aggregating information within the region of interest. Furthermore, it extend deformable cross-attention from 2D to 3D pixel space, enabling the differentiation of points with similar image coordinates based on their depth coordinates. Building upon this module, we introduce a neural network named CGFormer to achieve semantic scene completion. Simultaneously, CGFormer leverages multiple 3D representations (i.e., voxel and TPV) to boost the semantic and geometric representation abilities of the transformed 3D volume from both local and global perspectives. Experimental results demonstrate that CGFormer achieves state-of-the-art performance on the SemanticKITTI and SSCBench-KITTI-360 benchmarks, attaining a mIoU of 16.87 and 20.05, as well as an IoU of 45.99 and 48.07, respectively. Remarkably, CGFormer even outperforms approaches employing temporal images as inputs or much larger image backbone networks.
comment: NIPS 2024 Spotlight
♻ ☆ DCDepth: Progressive Monocular Depth Estimation in Discrete Cosine Domain NeurIPS-2024
In this paper, we introduce DCDepth, a novel framework for the long-standing monocular depth estimation task. Moving beyond conventional pixel-wise depth estimation in the spatial domain, our approach estimates the frequency coefficients of depth patches after transforming them into the discrete cosine domain. This unique formulation allows for the modeling of local depth correlations within each patch. Crucially, the frequency transformation segregates the depth information into various frequency components, with low-frequency components encapsulating the core scene structure and high-frequency components detailing the finer aspects. This decomposition forms the basis of our progressive strategy, which begins with the prediction of low-frequency components to establish a global scene context, followed by successive refinement of local details through the prediction of higher-frequency components. We conduct comprehensive experiments on NYU-Depth-V2, TOFDC, and KITTI datasets, and demonstrate the state-of-the-art performance of DCDepth. Code is available at https://github.com/w2kun/DCDepth.
comment: Accepted by NeurIPS-2024
♻ ☆ RectifID: Personalizing Rectified Flow with Anchored Classifier Guidance NeurIPS 2024
Customizing diffusion models to generate identity-preserving images from user-provided reference images is an intriguing new problem. The prevalent approaches typically require training on extensive domain-specific images to achieve identity preservation, which lacks flexibility across different use cases. To address this issue, we exploit classifier guidance, a training-free technique that steers diffusion models using an existing classifier, for personalized image generation. Our study shows that based on a recent rectified flow framework, the major limitation of vanilla classifier guidance in requiring a special classifier can be resolved with a simple fixed-point solution, allowing flexible personalization with off-the-shelf image discriminators. Moreover, its solving procedure proves to be stable when anchored to a reference flow trajectory, with a convergence guarantee. The derived method is implemented on rectified flow with different off-the-shelf image discriminators, delivering advantageous personalization results for human faces, live subjects, and certain objects. Code is available at https://github.com/feifeiobama/RectifID.
comment: NeurIPS 2024
♻ ☆ Developing a Thailand solar irradiance map using Himawari-8 satellite imageries and deep learning models
This paper presents an online platform that shows Thailand's solar irradiance map every 30 minutes. It is available at https://www.cusolarforecast.com. The methodology for estimating global horizontal irradiance (GHI) across Thailand relies on cloud index extracted from Himawari-8 satellite imagery, Ineichen clear-sky model with locally-tuned Linke turbidity, and machine learning models. The methods take clear-sky irradiance, cloud index, re-analyzed GHI and temperature data from the MERRA-2 database, and date-time as inputs for GHI estimation models, including LightGBM, LSTM, Informer, and Transformer. These are benchmarked with the estimate from a commercial service X by evaluating 15-minute ground GHI data from 53 ground stations over 1.5 years from 2022-2023. The results show that the four models have competitive performances and outperform the service X. The best model is LightGBM, with an MAE of 78.58 W/sqm and RMSE of 118.97 W/sqm. Obtaining re-analyzed MERRA-2 data for Thailand is not economically feasible for deployment. When removing these features, the Informer model has a winning performance of 78.67 W/sqm in MAE. The obtained performance aligns with existing literature by taking the climate zone and time granularity of data into consideration. As the map shows an estimate of GHI over 93,000 grids with a frequent update, the paper also describes a computational framework for displaying the entire map. It tests the runtime performance of deep learning models in the GHI estimation process.
comment: 23 pages, 14 figures
♻ ☆ Auxiliary CycleGAN-guidance for Task-Aware Domain Translation from Duplex to Monoplex IHC Images
Generative models enable the translation from a source image domain where readily trained models are available to a target domain unseen during training. While Cycle Generative Adversarial Networks (GANs) are well established, the associated cycle consistency constrain relies on that an invertible mapping exists between the two domains. This is, however, not the case for the translation between images stained with chromogenic monoplex and duplex immunohistochemistry (IHC) assays. Focusing on the translation from the latter to the first, we propose - through the introduction of a novel training design, an alternative constrain leveraging a set of immunofluorescence (IF) images as an auxiliary unpaired image domain. Quantitative and qualitative results on a downstream segmentation task show the benefit of the proposed method in comparison to baseline approaches.
comment: 5 pages
♻ ☆ Exploring Diversity-based Active Learning for 3D Object Detection in Autonomous Driving
3D object detection has recently received much attention due to its great potential in autonomous vehicle (AV). The success of deep learning based object detectors relies on the availability of large-scale annotated datasets, which is time-consuming and expensive to compile, especially for 3D bounding box annotation. In this work, we investigate diversity-based active learning (AL) as a potential solution to alleviate the annotation burden. Given limited annotation budget, only the most informative frames and objects are automatically selected for human to annotate. Technically, we take the advantage of the multimodal information provided in an AV dataset, and propose a novel acquisition function that enforces spatial and temporal diversity in the selected samples. We benchmark the proposed method against other AL strategies under realistic annotation cost measurement, where the realistic costs for annotating a frame and a 3D bounding box are both taken into consideration. We demonstrate the effectiveness of the proposed method on the nuScenes dataset and show that it outperforms existing AL strategies significantly. Code is available at https://github.com/Linkon87/Exploring-Diversity-based-Active-Learning-for-3D-Object-Detection-in-Autonomous-Driving
comment: IEEE Transactions on Intelligent Transportation Systems. Code is available at https://github.com/Linkon87/Exploring-Diversity-based-Active-Learning-for-3D-Object-Detection-in-Autonomous-Driving
♻ ☆ Slicing Through Bias: Explaining Performance Gaps in Medical Image Analysis using Slice Discovery Methods MICCAI 2024
Machine learning models have achieved high overall accuracy in medical image analysis. However, performance disparities on specific patient groups pose challenges to their clinical utility, safety, and fairness. This can affect known patient groups - such as those based on sex, age, or disease subtype - as well as previously unknown and unlabeled groups. Furthermore, the root cause of such observed performance disparities is often challenging to uncover, hindering mitigation efforts. In this paper, to address these issues, we leverage Slice Discovery Methods (SDMs) to identify interpretable underperforming subsets of data and formulate hypotheses regarding the cause of observed performance disparities. We introduce a novel SDM and apply it in a case study on the classification of pneumothorax and atelectasis from chest x-rays. Our study demonstrates the effectiveness of SDMs in hypothesis formulation and yields an explanation of previously observed but unexplained performance disparities between male and female patients in widely used chest X-ray datasets and models. Our findings indicate shortcut learning in both classification tasks, through the presence of chest drains and ECG wires, respectively. Sex-based differences in the prevalence of these shortcut features appear to cause the observed classification performance gap, representing a previously underappreciated interaction between shortcut learning and model fairness analyses.
comment: MICCAI 2024 Workshop on Fairness of AI in Medical Imaging
♻ ☆ Incremental Joint Learning of Depth, Pose and Implicit Scene Representation on Monocular Camera in Large-scale Scenes
Dense scene reconstruction for photo-realistic view synthesis has various applications, such as VR/AR, autonomous vehicles. However, most existing methods have difficulties in large-scale scenes due to three core challenges: \textit{(a) inaccurate depth input.} Accurate depth input is impossible to get in real-world large-scale scenes. \textit{(b) inaccurate pose estimation.} Most existing approaches rely on accurate pre-estimated camera poses. \textit{(c) insufficient scene representation capability.} A single global radiance field lacks the capacity to effectively scale to large-scale scenes. To this end, we propose an incremental joint learning framework, which can achieve accurate depth, pose estimation, and large-scale scene reconstruction. A vision transformer-based network is adopted as the backbone to enhance performance in scale information estimation. For pose estimation, a feature-metric bundle adjustment (FBA) method is designed for accurate and robust camera tracking in large-scale scenes. In terms of implicit scene representation, we propose an incremental scene representation method to construct the entire large-scale scene as multiple local radiance fields to enhance the scalability of 3D scene representation. Extended experiments have been conducted to demonstrate the effectiveness and accuracy of our method in depth estimation, pose estimation, and large-scale scene reconstruction.
♻ ☆ DPEC: Dual-Path Error Compensation Method for Enhanced Low-Light Image Clarity
For the task of low-light image enhancement, deep learning-based algorithms have demonstrated superiority and effectiveness compared to traditional methods. Existing deep learning algorithms are proposed mainly based on the Retinex theory but overlook the noise and color distortion present in the input, which frequently results in significant noise amplification and local color distortion in the final results. To address this, we propose a Dual-Path Error Compensation method (DPEC), which aims to improve image quality in low-light conditions. DPEC performs precise pixel-level error estimation, which accurately captures subtle pixels differences, and independent denoising, which effectively removes unnecessary noise. This method restores image brightness while preserving local texture details and avoiding noise amplification. Furthermore, to compensate for the traditional CNN's limited ability to capture long-range semantic information and considering both computational speed and resource efficiency, we integrated the VMamba architecture into the backbone of DPEC. In addition, we introduced the HIS-Retinex loss to constrain the training of DPEC, ensuring that the overall brightness distribution of the images more closely aligns with real-world conditions. Comprehensive quantitative and qualitative experimental results demonstrate that our algorithm significantly outperforms state-of-the-art methods across six benchmark tests.
♻ ☆ Toward Fairer Face Recognition Datasets
Face recognition and verification are two computer vision tasks whose performance has progressed with the introduction of deep representations. However, ethical, legal, and technical challenges due to the sensitive character of face data and biases in real training datasets hinder their development. Generative AI addresses privacy by creating fictitious identities, but fairness problems persist. We promote fairness by introducing a demographic attributes balancing mechanism in generated training datasets. We experiment with an existing real dataset, three generated training datasets, and the balanced versions of a diffusion-based dataset. We propose a comprehensive evaluation that considers accuracy and fairness equally and includes a rigorous regression-based statistical analysis of attributes. The analysis shows that balancing reduces demographic unfairness. Also, a performance gap persists despite generation becoming more accurate with time. The proposed balancing method and comprehensive verification evaluation promote fairer and transparent face recognition and verification.
♻ ☆ Feature Extraction for Generative Medical Imaging Evaluation: New Evidence Against an Evolving Trend
Fr\'echet Inception Distance (FID) is a widely used metric for assessing synthetic image quality. It relies on an ImageNet-based feature extractor, making its applicability to medical imaging unclear. A recent trend is to adapt FID to medical imaging through feature extractors trained on medical images. Our study challenges this practice by demonstrating that ImageNet-based extractors are more consistent and aligned with human judgment than their RadImageNet counterparts. We evaluated sixteen StyleGAN2 networks across four medical imaging modalities and four data augmentation techniques with Fr\'echet distances (FDs) computed using eleven ImageNet or RadImageNet-trained feature extractors. Comparison with human judgment via visual Turing tests revealed that ImageNet-based extractors produced rankings consistent with human judgment, with the FD derived from the ImageNet-trained SwAV extractor significantly correlating with expert evaluations. In contrast, RadImageNet-based rankings were volatile and inconsistent with human judgment. Our findings challenge prevailing assumptions, providing novel evidence that medical image-trained feature extractors do not inherently improve FDs and can even compromise their reliability. Our code is available at https://github.com/mckellwoodland/fid-med-eval.
comment: This preprint has not undergone peer review or any post-submission improvements or corrections. The Version of Record of this contribution is published in LNCS vol. 15012, and is available online at https://doi.org/10.1007/978-3-031-72390-2_9
♻ ☆ Taming Diffusion Models for Image Restoration: A Review
Diffusion models have achieved remarkable progress in generative modelling, particularly in enhancing image quality to conform to human preferences. Recently, these models have also been applied to low-level computer vision for photo-realistic image restoration (IR) in tasks such as image denoising, deblurring, dehazing, etc. In this review paper, we introduce key constructions in diffusion models and survey contemporary techniques that make use of diffusion models in solving general IR tasks. Furthermore, we point out the main challenges and limitations of existing diffusion-based IR frameworks and provide potential directions for future work.
comment: Review paper; any comments and suggestions are most welcome!
♻ ☆ Mitral Regurgitation Recognition based on Unsupervised Out-of-Distribution Detection with Residual Diffusion Amplification MICCAI
Mitral regurgitation (MR) is a serious heart valve disease. Early and accurate diagnosis of MR via ultrasound video is critical for timely clinical decision-making and surgical intervention. However, manual MR diagnosis heavily relies on the operator's experience, which may cause misdiagnosis and inter-observer variability. Since MR data is limited and has large intra-class variability, we propose an unsupervised out-of-distribution (OOD) detection method to identify MR rather than building a deep classifier. To our knowledge, we are the first to explore OOD in MR ultrasound videos. Our method consists of a feature extractor, a feature reconstruction model, and a residual accumulation amplification algorithm. The feature extractor obtains features from the video clips and feeds them into the feature reconstruction model to restore the original features. The residual accumulation amplification algorithm then iteratively performs noise feature reconstruction, amplifying the reconstructed error of OOD features. This algorithm is straightforward yet efficient and can seamlessly integrate as a plug-and-play component in reconstruction-based OOD detection methods. We validated the proposed method on a large ultrasound dataset containing 893 non-MR and 267 MR videos. Experimental results show that our OOD detection method can effectively identify MR samples.
comment: Accepted by MICCAI MLMI 2024, 11 pages, 3 figures
♻ ☆ Fully Explicit Dynamic Gaussian Splatting NeurIPS 2024
3D Gaussian Splatting has shown fast and high-quality rendering results in static scenes by leveraging dense 3D prior and explicit representations. Unfortunately, the benefits of the prior and representation do not involve novel view synthesis for dynamic motions. Ironically, this is because the main barrier is the reliance on them, which requires increasing training and rendering times to account for dynamic motions. In this paper, we design a Explicit 4D Gaussian Splatting(Ex4DGS). Our key idea is to firstly separate static and dynamic Gaussians during training, and to explicitly sample positions and rotations of the dynamic Gaussians at sparse timestamps. The sampled positions and rotations are then interpolated to represent both spatially and temporally continuous motions of objects in dynamic scenes as well as reducing computational cost. Additionally, we introduce a progressive training scheme and a point-backtracking technique that improves Ex4DGS's convergence. We initially train Ex4DGS using short timestamps and progressively extend timestamps, which makes it work well with a few point clouds. The point-backtracking is used to quantify the cumulative error of each Gaussian over time, enabling the detection and removal of erroneous Gaussians in dynamic scenes. Comprehensive experiments on various scenes demonstrate the state-of-the-art rendering quality from our method, achieving fast rendering of 62 fps on a single 2080Ti GPU.
comment: Accepted at NeurIPS 2024
♻ ☆ Towards Understanding the Working Mechanism of Text-to-Image Diffusion Model
Recently, the strong latent Diffusion Probabilistic Model (DPM) has been applied to high-quality Text-to-Image (T2I) generation (e.g., Stable Diffusion), by injecting the encoded target text prompt into the gradually denoised diffusion image generator. Despite the success of DPM in practice, the mechanism behind it remains to be explored. To fill this blank, we begin by examining the intermediate statuses during the gradual denoising generation process in DPM. The empirical observations indicate, the shape of image is reconstructed after the first few denoising steps, and then the image is filled with details (e.g., texture). The phenomenon is because the low-frequency signal (shape relevant) of the noisy image is not corrupted until the final stage in the forward process (initial stage of generation) of adding noise in DPM. Inspired by the observations, we proceed to explore the influence of each token in the text prompt during the two stages. After a series of experiments of T2I generations conditioned on a set of text prompts. We conclude that in the earlier generation stage, the image is mostly decided by the special token [\texttt{EOS}] in the text prompt, and the information in the text prompt is already conveyed in this stage. After that, the diffusion model completes the details of generated images by information from themselves. Finally, we propose to apply this observation to accelerate the process of T2I generation by properly removing text guidance, which finally accelerates the sampling up to 25\%+.
♻ ☆ IPDreamer: Appearance-Controllable 3D Object Generation with Complex Image Prompts
Recent advances in 3D generation have been remarkable, with methods such as DreamFusion leveraging large-scale text-to-image diffusion-based models to guide 3D object generation. These methods enable the synthesis of detailed and photorealistic textured objects. However, the appearance of 3D objects produced by such text-to-3D models is often unpredictable, and it is hard for single-image-to-3D methods to deal with images lacking a clear subject, complicating the generation of appearance-controllable 3D objects from complex images. To address these challenges, we present IPDreamer, a novel method that captures intricate appearance features from complex $\textbf{I}$mage $\textbf{P}$rompts and aligns the synthesized 3D object with these extracted features, enabling high-fidelity, appearance-controllable 3D object generation. Our experiments demonstrate that IPDreamer consistently generates high-quality 3D objects that align with both the textual and complex image prompts, highlighting its promising capability in appearance-controlled, complex 3D object generation. Our code is available at https://github.com/zengbohan0217/IPDreamer.
comment: 20 pages, 12 figures
♻ ☆ Point Cloud Matters: Rethinking the Impact of Different Observation Spaces on Robot Learning NeurIPS 2024
In robot learning, the observation space is crucial due to the distinct characteristics of different modalities, which can potentially become a bottleneck alongside policy design. In this study, we explore the influence of various observation spaces on robot learning, focusing on three predominant modalities: RGB, RGB-D, and point cloud. We introduce OBSBench, a benchmark comprising two simulators and 125 tasks, along with standardized pipelines for various encoders and policy baselines. Extensive experiments on diverse contact-rich manipulation tasks reveal a notable trend: point cloud-based methods, even those with the simplest designs, frequently outperform their RGB and RGB-D counterparts. This trend persists in both scenarios: training from scratch and utilizing pre-training. Furthermore, our findings demonstrate that point cloud observations often yield better policy performance and significantly stronger generalization capabilities across various geometric and visual conditions. These outcomes suggest that the 3D point cloud is a valuable observation modality for intricate robotic tasks. We also suggest that incorporating both appearance and coordinate information can enhance the performance of point cloud methods. We hope our work provides valuable insights and guidance for designing more generalizable and robust robotic models. Codes are available at https://github.com/HaoyiZhu/PointCloudMatters.
comment: 38th Conference on Neural Information Processing Systems (NeurIPS 2024) Track on Datasets and Benchmarks
♻ ☆ Selective Volume Mixup for Video Action Recognition
The recent advances in Convolutional Neural Networks (CNNs) and Vision Transformers have convincingly demonstrated high learning capability for video action recognition on large datasets. Nevertheless, deep models often suffer from the overfitting effect on small-scale datasets with a limited number of training videos. A common solution is to exploit the existing image augmentation strategies for each frame individually including Mixup, Cutmix, and RandAugment, which are not particularly optimized for video data. In this paper, we propose a novel video augmentation strategy named Selective Volume Mixup (SV-Mix) to improve the generalization ability of deep models with limited training videos. SV-Mix devises a learnable selective module to choose the most informative volumes from two videos and mixes the volumes up to achieve a new training video. Technically, we propose two new modules, i.e., a spatial selective module to select the local patches for each spatial position, and a temporal selective module to mix the entire frames for each timestamp and maintain the spatial pattern. At each time, we randomly choose one of the two modules to expand the diversity of training samples. The selective modules are jointly optimized with the video action recognition framework to find the optimal augmentation strategy. We empirically demonstrate the merits of the SV-Mix augmentation on a wide range of video action recognition benchmarks and consistently boot the performances of both CNN-based and transformer-based models.
♻ ☆ Meteor: Mamba-based Traversal of Rationale for Large Language and Vision Models
The rapid development of large language and vision models (LLVMs) has been driven by advances in visual instruction tuning. Recently, open-source LLVMs have curated high-quality visual instruction tuning datasets and utilized additional vision encoders or multiple computer vision models in order to narrow the performance gap with powerful closed-source LLVMs. These advancements are attributed to multifaceted information required for diverse capabilities, including fundamental image understanding, real-world knowledge about common-sense and non-object concepts (e.g., charts, diagrams, symbols, signs, and math problems), and step-by-step procedures for solving complex questions. Drawing from the multifaceted information, we present a new efficient LLVM, Mamba-based traversal of rationales (Meteor), which leverages multifaceted rationale to enhance understanding and answering capabilities. To embed lengthy rationales containing abundant information, we employ the Mamba architecture, capable of processing sequential data with linear time complexity. We introduce a new concept of traversal of rationale that facilitates efficient embedding of rationale. Subsequently, the backbone multimodal language model (MLM) is trained to generate answers with the aid of rationale. Through these steps, Meteor achieves significant improvements in vision language performances across multiple evaluation benchmarks requiring diverse capabilities, without scaling up the model size or employing additional vision encoders and computer vision models.
comment: Code is available in https://github.com/ByungKwanLee/Meteor
♻ ☆ 4Diffusion: Multi-view Video Diffusion Model for 4D Generation NeurIPS 2024
Current 4D generation methods have achieved noteworthy efficacy with the aid of advanced diffusion generative models. However, these methods lack multi-view spatial-temporal modeling and encounter challenges in integrating diverse prior knowledge from multiple diffusion models, resulting in inconsistent temporal appearance and flickers. In this paper, we propose a novel 4D generation pipeline, namely 4Diffusion, aimed at generating spatial-temporally consistent 4D content from a monocular video. We first design a unified diffusion model tailored for multi-view video generation by incorporating a learnable motion module into a frozen 3D-aware diffusion model to capture multi-view spatial-temporal correlations. After training on a curated dataset, our diffusion model acquires reasonable temporal consistency and inherently preserves the generalizability and spatial consistency of the 3D-aware diffusion model. Subsequently, we propose 4D-aware Score Distillation Sampling loss, which is based on our multi-view video diffusion model, to optimize 4D representation parameterized by dynamic NeRF. This aims to eliminate discrepancies arising from multiple diffusion models, allowing for generating spatial-temporally consistent 4D content. Moreover, we devise an anchor loss to enhance the appearance details and facilitate the learning of dynamic NeRF. Extensive qualitative and quantitative experiments demonstrate that our method achieves superior performance compared to previous methods.
comment: NeurIPS 2024. Project Page: https://aejion.github.io/4diffusion/
♻ ☆ UrbanWorld: An Urban World Model for 3D City Generation
Cities, as the essential environment of human life, encompass diverse physical elements such as buildings, roads and vegetation, which continuously interact with dynamic entities like people and vehicles. Crafting realistic, interactive 3D urban environments is essential for nurturing AGI systems and constructing AI agents capable of perceiving, decision-making, and acting like humans in real-world environments. However, creating high-fidelity 3D urban environments usually entails extensive manual labor from designers, involving intricate detailing and representation of complex urban elements. Therefore, accomplishing this automatically remains a longstanding challenge. Toward this problem, we propose UrbanWorld, the first generative urban world model that can automatically create a customized, realistic and interactive 3D urban world with flexible control conditions. UrbanWorld incorporates four key stages in the generation pipeline: flexible 3D layout generation from OSM data or urban layout with semantic and height maps, urban scene design with Urban MLLM, controllable urban asset rendering via progressive 3D diffusion, and MLLM-assisted scene refinement. We conduct extensive quantitative analysis on five visual metrics, demonstrating that UrbanWorld achieves SOTA generation realism. Next, we provide qualitative results about the controllable generation capabilities of UrbanWorld using both textual and image-based prompts. Lastly, we verify the interactive nature of these environments by showcasing the agent perception and navigation within the created environments. We contribute UrbanWorld as an open-source tool available at https://github.com/Urban-World/UrbanWorld.
comment: 14 pages
♻ ☆ Binarized Diffusion Model for Image Super-Resolution NeurIPS 2024
Advanced diffusion models (DMs) perform impressively in image super-resolution (SR), but the high memory and computational costs hinder their deployment. Binarization, an ultra-compression algorithm, offers the potential for effectively accelerating DMs. Nonetheless, due to the model structure and the multi-step iterative attribute of DMs, existing binarization methods result in significant performance degradation. In this paper, we introduce a novel binarized diffusion model, BI-DiffSR, for image SR. First, for the model structure, we design a UNet architecture optimized for binarization. We propose the consistent-pixel-downsample (CP-Down) and consistent-pixel-upsample (CP-Up) to maintain dimension consistent and facilitate the full-precision information transfer. Meanwhile, we design the channel-shuffle-fusion (CS-Fusion) to enhance feature fusion in skip connection. Second, for the activation difference across timestep, we design the timestep-aware redistribution (TaR) and activation function (TaA). The TaR and TaA dynamically adjust the distribution of activations based on different timesteps, improving the flexibility and representation alability of the binarized module. Comprehensive experiments demonstrate that our BI-DiffSR outperforms existing binarization methods. Code is released at: https://github.com/zhengchen1999/BI-DiffSR.
comment: Accepted to NeurIPS 2024. Code is available at https://github.com/zhengchen1999/BI-DiffSR
♻ ☆ Mini-InternVL: A Flexible-Transfer Pocket Multimodal Model with 5% Parameters and 90% Performance
Multimodal large language models (MLLMs) have demonstrated impressive performance in vision-language tasks across a broad spectrum of domains. However, the large model scale and associated high computational costs pose significant challenges for training and deploying MLLMs on consumer-grade GPUs or edge devices, thereby hindering their widespread application. In this work, we introduce Mini-InternVL, a series of MLLMs with parameters ranging from 1B to 4B, which achieves 90% of the performance with only 5% of the parameters. This significant improvement in efficiency and effectiveness makes our models more accessible and applicable in various real-world scenarios. To further promote the adoption of our models, we develop a unified adaptation framework for Mini-InternVL, which enables our models to transfer and outperform specialized models in downstream tasks, including autonomous driving, medical images, and remote sensing. We believe that our study can provide valuable insights and resources to advance the development of efficient and effective MLLMs. Code is available at https://github.com/OpenGVLab/InternVL.
comment: Technical report
♻ ☆ Localize, Understand, Collaborate: Semantic-Aware Dragging via Intention Reasoner NeurIPS 2024
Flexible and accurate drag-based editing is a challenging task that has recently garnered significant attention. Current methods typically model this problem as automatically learning "how to drag" through point dragging and often produce one deterministic estimation, which presents two key limitations: 1) Overlooking the inherently ill-posed nature of drag-based editing, where multiple results may correspond to a given input, as illustrated in Fig.1; 2) Ignoring the constraint of image quality, which may lead to unexpected distortion. To alleviate this, we propose LucidDrag, which shifts the focus from "how to drag" to "what-then-how" paradigm. LucidDrag comprises an intention reasoner and a collaborative guidance sampling mechanism. The former infers several optimal editing strategies, identifying what content and what semantic direction to be edited. Based on the former, the latter addresses "how to drag" by collaboratively integrating existing editing guidance with the newly proposed semantic guidance and quality guidance. Specifically, semantic guidance is derived by establishing a semantic editing direction based on reasoned intentions, while quality guidance is achieved through classifier guidance using an image fidelity discriminator. Both qualitative and quantitative comparisons demonstrate the superiority of LucidDrag over previous methods.
comment: Accepted by NeurIPS 2024
♻ ☆ Is Your HD Map Constructor Reliable under Sensor Corruptions? NeurIPS 2024
Driving systems often rely on high-definition (HD) maps for precise environmental information, which is crucial for planning and navigation. While current HD map constructors perform well under ideal conditions, their resilience to real-world challenges, \eg, adverse weather and sensor failures, is not well understood, raising safety concerns. This work introduces MapBench, the first comprehensive benchmark designed to evaluate the robustness of HD map construction methods against various sensor corruptions. Our benchmark encompasses a total of 29 types of corruptions that occur from cameras and LiDAR sensors. Extensive evaluations across 31 HD map constructors reveal significant performance degradation of existing methods under adverse weather conditions and sensor failures, underscoring critical safety concerns. We identify effective strategies for enhancing robustness, including innovative approaches that leverage multi-modal fusion, advanced data augmentation, and architectural techniques. These insights provide a pathway for developing more reliable HD map construction methods, which are essential for the advancement of autonomous driving technology. The benchmark toolkit and affiliated code and model checkpoints have been made publicly accessible.
comment: NeurIPS 2024; 40 pages, 17 figures, 23 tables; Code at https://mapbench.github.io/
♻ ☆ Evaluating Feature Attribution Methods for Electrocardiogram
The performance of cardiac arrhythmia detection with electrocardiograms(ECGs) has been considerably improved since the introduction of deep learning models. In practice, the high performance alone is not sufficient and a proper explanation is also required. Recently, researchers have started adopting feature attribution methods to address this requirement, but it has been unclear which of the methods are appropriate for ECG. In this work, we identify and customize three evaluation metrics for feature attribution methods based on the characteristics of ECG: localization score, pointing game, and degradation score. Using the three evaluation metrics, we evaluate and analyze eleven widely-used feature attribution methods. We find that some of the feature attribution methods are much more adequate for explaining ECG, where Grad-CAM outperforms the second-best method by a large margin.
comment: This is preliminary research related to https://www.sciencedirect.com/science/article/pii/S0010482524011739 . Code is available at https://github.com/SNU-DRL/Attribution-ECG
♻ ☆ A Lightweight Modular Framework for Low-Cost Open-Vocabulary Object Detection Training
Object detection is a fundamental challenge in computer vision, centered on recognizing objects within images, with diverse applications in areas like image analysis, robotics, and autonomous vehicles. Although existing methods have achieved great success, they are often constrained by a fixed vocabulary of objects. To overcome this limitation, approaches like MDETR have redefined object detection by incorporating region-level vision-language pre-training, enabling open-vocabulary object detectors. However, these methods are computationally heavy due to the simultaneous training of large models for both vision and language representations. To address this, we introduce a lightweight framework that significantly reduces the number of parameters while preserving, or even improving, performance. Our solution is applied to MDETR, resulting in the development of Lightweight MDETR (LightMDETR), an optimized version of MDETR designed to enhance computational efficiency without sacrificing accuracy. The core of our approach involves freezing the MDETR backbone and training only the Universal Projection module (UP), which bridges vision and language representations. A learnable modality token parameter allows the UP to seamlessly switch between modalities. Evaluations on tasks like phrase grounding, referring expression comprehension, and segmentation show that LightMDETR not only reduces computational costs but also outperforms several state-of-the-art methods in terms of accuracy.
♻ ☆ TensorProjection Layer: A Tensor-Based Dimension Reduction Method in Deep Neural Networks
In this paper, we propose a dimension reduction method specifically designed for tensor-structured feature data in deep neural networks. The method is implemented as a hidden layer, called the TensorProjection layer, which transforms input tensors into output tensors with reduced dimensions through mode-wise projections. The projection directions are treated as model parameters of the layer and are optimized during model training. Our method can serve as an alternative to pooling layers for summarizing image data, or to convolutional layers as a technique for reducing the number of channels. We conduct experiments on tasks such as medical image classification and segmentation, integrating the TensorProjection layer into commonly used baseline architectures to evaluate its effectiveness. Numerical experiments indicate that the proposed method can outperform traditional downsampling methods, such as pooling layers, in our tasks, suggesting it as a promising alternative for feature summarization.
♻ ☆ Boosting the Generalization Ability for Hyperspectral Image Classification using Spectral-spatial Axial Aggregation Transformer
In the hyperspectral image classification (HSIC) task, the most commonly used model validation paradigm is partitioning the training-test dataset through pixel-wise random sampling. By training on a small amount of data, the deep learning model can achieve almost perfect accuracy. However, in our experiments, we found that the high accuracy was reached because the training and test datasets share a lot of information. On non-overlapping dataset partitions, well-performing models suffer significant performance degradation. To this end, we propose a spectral-spatial axial aggregation transformer model, namely SaaFormer, that preserves generalization across dataset partitions. SaaFormer applies a multi-level spectral extraction structure to segment the spectrum into multiple spectrum clips, such that the wavelength continuity of the spectrum across the channel are preserved. For each spectrum clip, the axial aggregation attention mechanism, which integrates spatial features along multiple spectral axes is applied to mine the spectral characteristic. The multi-level spectral extraction and the axial aggregation attention emphasize spectral characteristic to improve the model generalization. The experimental results on five publicly available datasets demonstrate that our model exhibits comparable performance on the random partition, while significantly outperforming other methods on non-overlapping partitions. Moreover, SaaFormer shows excellent performance on background classification.
comment: arXiv admin note: text overlap with arXiv:2107.02988 by other authors
♻ ☆ Unsupervised Assessment of Landscape Shifts Based on Persistent Entropy and Topological Preservation KDD'2024
In Continual Learning (CL) contexts, concept drift typically refers to the analysis of changes in data distribution. A drift in the input data can have negative consequences on a learning predictor and the system's stability. The majority of concept drift methods emphasize the analysis of statistical changes in non-stationary data over time. In this context, we consider another perspective, where the concept drift also integrates substantial changes in the topological characteristics of the data stream. In this article, we introduce a novel framework for monitoring changes in multi-dimensional data streams. We explore variations in the topological structures of the data, presenting another angle on the standard concept drift. Our developed approach is based on persistent entropy and topology-preserving projections in a continual learning scenario. The framework operates in both unsupervised and supervised environments. To show the utility of the proposed framework, we analyze the model across three scenarios using data streams generated with MNIST samples. The obtained results reveal the potential of applying topological data analysis for shift detection and encourage further research in this area.
comment: KDD'2024. Workshop on Drift Detection and Landscape Shifts
♻ ☆ Transformer for Object Re-Identification: A Survey
Object Re-identification (Re-ID) aims to identify specific objects across different times and scenes, which is a widely researched task in computer vision. For a prolonged period, this field has been predominantly driven by deep learning technology based on convolutional neural networks. In recent years, the emergence of Vision Transformers has spurred a growing number of studies delving deeper into Transformer-based Re-ID, continuously breaking performance records and witnessing significant progress in the Re-ID field. Offering a powerful, flexible, and unified solution, Transformers cater to a wide array of Re-ID tasks with unparalleled efficacy. This paper provides a comprehensive review and in-depth analysis of the Transformer-based Re-ID. In categorizing existing works into Image/Video-Based Re-ID, Re-ID with limited data/annotations, Cross-Modal Re-ID, and Special Re-ID Scenarios, we thoroughly elucidate the advantages demonstrated by the Transformer in addressing a multitude of challenges across these domains. Considering the trending unsupervised Re-ID, we propose a new Transformer baseline, UntransReID, achieving state-of-the-art performance on both single/cross modal tasks. For the under-explored animal Re-ID, we devise a standardized experimental benchmark and conduct extensive experiments to explore the applicability of Transformer for this task and facilitate future research. Finally, we discuss some important yet under-investigated open issues in the large foundation model era, we believe it will serve as a new handbook for researchers in this field. A periodically updated website will be available at https://github.com/mangye16/ReID-Survey.
comment: Accepted by International Journal of Computer Vision (IJCV) in October 2024
♻ ☆ LucidFusion: Generating 3D Gaussians with Arbitrary Unposed Images
Recent large reconstruction models have made notable progress in generating high-quality 3D objects from single images. However, these methods often struggle with controllability, as they lack information from multiple views, leading to incomplete or inconsistent 3D reconstructions. To address this limitation, we introduce LucidFusion, a flexible end-to-end feed-forward framework that leverages the Relative Coordinate Map (RCM). Unlike traditional methods linking images to 3D world thorough pose, LucidFusion utilizes RCM to align geometric features coherently across different views, making it highly adaptable for 3D generation from arbitrary, unposed images. Furthermore, LucidFusion seamlessly integrates with the original single-image-to-3D pipeline, producing detailed 3D Gaussians at a resolution of $512 \times 512$, making it well-suited for a wide range of applications.
comment: 17 pages, 12 figures, [project page](https://heye0507.github.io/LucidFusion_page/)
♻ ☆ Dynamic Test-Time Augmentation via Differentiable Functions
Distribution shifts, which often occur in the real world, degrade the accuracy of deep learning systems, and thus improving robustness to distribution shifts is essential for practical applications. To improve robustness, we study an image enhancement method that generates recognition-friendly images without retraining the recognition model. We propose a novel image enhancement method, DynTTA, which is based on differentiable data augmentation techniques and generates a blended image from many augmented images to improve the recognition accuracy under distribution shifts. In addition to standard data augmentations, DynTTA also incorporates deep neural network-based image transformation, further improving the robustness. Because DynTTA is composed of differentiable functions, it can be directly trained with the classification loss of the recognition model. In experiments with widely used image recognition datasets using various classification models, DynTTA improves the robustness with almost no reduction in classification accuracy for clean images, thus outperforming the existing methods. Furthermore, the results show that robustness is significantly improved by estimating the training-time augmentations for distribution-shifted datasets using DynTTA and retraining the recognition model with the estimated augmentations. DynTTA is a promising approach for applications that require both clean accuracy and robustness. Our code is available at \url{https://github.com/s-enmt/DynTTA}.
comment: IEEE Access
♻ ☆ Mask-guided cross-image attention for zero-shot in-silico histopathologic image generation with a diffusion model
Creating in-silico data with generative AI promises a cost-effective alternative to staining, imaging, and annotating whole slide images in computational pathology. Diffusion models are the state-of-the-art solution for generating in-silico images, offering unparalleled fidelity and realism. Using appearance transfer diffusion models allows for zero-shot image generation, facilitating fast application and making model training unnecessary. However current appearance transfer diffusion models are designed for natural images, where the main task is to transfer the foreground object from an origin to a target domain, while the background is of insignificant importance. In computational pathology, specifically in oncology, it is however not straightforward to define which objects in an image should be classified as foreground and background, as all objects in an image may be of critical importance for the detailed understanding the tumor micro-environment. We contribute to the applicability of appearance transfer diffusion models to immunohistochemistry-stained images by modifying the appearance transfer guidance to alternate between class-specific AdaIN feature statistics matchings using existing segmentation masks. The performance of the proposed method is demonstrated on the downstream task of supervised epithelium segmentation, showing that the number of manual annotations required for model training can be reduced by 75%, outperforming the baseline approach. Additionally, we consulted with a certified pathologist to investigate future improvements. We anticipate this work to inspire the application of zero-shot diffusion models in computational pathology, providing an efficient method to generate in-silico images with unmatched fidelity and realism, which prove meaningful for downstream tasks, such as training existing deep learning models or finetuning foundation models.
comment: 5 pages
♻ ☆ Granularity Matters in Long-Tail Learning
Balancing training on long-tail data distributions remains a long-standing challenge in deep learning. While methods such as re-weighting and re-sampling help alleviate the imbalance issue, limited sample diversity continues to hinder models from learning robust and generalizable feature representations, particularly for tail classes. In contrast to existing methods, we offer a novel perspective on long-tail learning, inspired by an observation: datasets with finer granularity tend to be less affected by data imbalance. In this paper, we investigate this phenomenon through both quantitative and qualitative studies, showing that increased granularity enhances the generalization of learned features in tail categories. Motivated by these findings, we propose a method to increase dataset granularity through category extrapolation. Specifically, we introduce open-set auxiliary classes that are visually similar to existing ones, aiming to enhance representation learning for both head and tail classes. This forms the core contribution and insight of our approach. To automate the curation of auxiliary data, we leverage large language models (LLMs) as knowledge bases to search for auxiliary categories and retrieve relevant images through web crawling. To prevent the overwhelming presence of auxiliary classes from disrupting training, we introduce a neighbor-silencing loss that encourages the model to focus on class discrimination within the target dataset. During inference, the classifier weights for auxiliary categories are masked out, leaving only the target class weights for use. Extensive experiments and ablation studies on three standard long-tail benchmarks demonstrate the effectiveness of our approach, notably outperforming strong baseline methods that use the same amount of data. The code will be made publicly available.
♻ ☆ CamI2V: Camera-Controlled Image-to-Video Diffusion Model
Recently, camera pose, as a user-friendly and physics-related condition, has been introduced into text-to-video diffusion model for camera control. However, existing methods simply inject camera conditions through a side input. These approaches neglect the inherent physical knowledge of camera pose, resulting in imprecise camera control, inconsistencies, and also poor interpretability. In this paper, we emphasize the necessity of integrating explicit physical constraints into model design. Epipolar attention is proposed for modeling all cross-frame relationships from a novel perspective of noised condition. This ensures that features are aggregated from corresponding epipolar lines in all noised frames, overcoming the limitations of current attention mechanisms in tracking displaced features across frames, especially when features move significantly with the camera and become obscured by noise. Additionally, we introduce register tokens to handle cases without intersections between frames, commonly caused by rapid camera movements, dynamic objects, or occlusions. To support image-to-video, we propose the multiple guidance scale to allow for precise control for image, text, and camera, respectively. Furthermore, we establish a more robust and reproducible evaluation pipeline to solve the inaccuracy and instability of existing camera control measurement. We achieve a 25.5% improvement in camera controllability on RealEstate10K while maintaining strong generalization to out-of-domain images. Only 24GB and 12GB are required for training and inference, respectively. We plan to release checkpoints, along with training and evaluation codes. Dynamic videos are best viewed at https://zgctroy.github.io/CamI2V.
♻ ☆ Reducing Hallucinations in Vision-Language Models via Latent Space Steering
Hallucination poses a challenge to the deployment of large vision-language models (LVLMs) in applications. Unlike in large language models (LLMs), hallucination in LVLMs often arises from misalignments between visual inputs and textual outputs. This paper investigates the underlying mechanisms of hallucination, focusing on the unique structure of LVLMs that distinguishes them from large language models (LLMs). We identify that hallucinations often arise from the sensitivity of text decoders to vision inputs, a natural phenomenon when image encoders and text decoders are pre-trained separately. Inspired by this, we introduce Visual and Textual Intervention (VTI), a novel technique designed to reduce hallucinations by steering latent space representations during inference to enhance the stability of vision features. As a task-agnostic test-time intervention, VTI can be easily applied to any problem without additional cost. Extensive experiments demonstrate that it can effectively reduce hallucinations and outperform baseline methods across multiple metrics, highlighting the critical role of vision feature stability in LVLMs.
comment: 21 pages
♻ ☆ MotionClone: Training-Free Motion Cloning for Controllable Video Generation
Motion-based controllable video generation offers the potential for creating captivating visual content. Existing methods typically necessitate model training to encode particular motion cues or incorporate fine-tuning to inject certain motion patterns, resulting in limited flexibility and generalization. In this work, we propose MotionClone, a training-free framework that enables motion cloning from reference videos to versatile motion-controlled video generation, including text-to-video and image-to-video. Based on the observation that the dominant components in temporal-attention maps drive motion synthesis, while the rest mainly capture noisy or very subtle motions, MotionClone utilizes sparse temporal attention weights as motion representations for motion guidance, facilitating diverse motion transfer across varying scenarios. Meanwhile, MotionClone allows for the direct extraction of motion representation through a single denoising step, bypassing the cumbersome inversion processes and thus promoting both efficiency and flexibility. Extensive experiments demonstrate that MotionClone exhibits proficiency in both global camera motion and local object motion, with notable superiority in terms of motion fidelity, textual alignment, and temporal consistency.
comment: 18 pages, 14 figures, https://bujiazi.github.io/motionclone.github.io/
♻ ☆ Cross-Modality Perturbation Synergy Attack for Person Re-identification NeurIPS 2024
In recent years, there has been significant research focusing on addressing security concerns in single-modal person re-identification (ReID) systems that are based on RGB images. However, the safety of cross-modality scenarios, which are more commonly encountered in practical applications involving images captured by infrared cameras, has not received adequate attention. The main challenge in cross-modality ReID lies in effectively dealing with visual differences between different modalities. For instance, infrared images are typically grayscale, unlike visible images that contain color information. Existing attack methods have primarily focused on the characteristics of the visible image modality, overlooking the features of other modalities and the variations in data distribution among different modalities. This oversight can potentially undermine the effectiveness of these methods in image retrieval across diverse modalities. This study represents the first exploration into the security of cross-modality ReID models and proposes a universal perturbation attack specifically designed for cross-modality ReID. This attack optimizes perturbations by leveraging gradients from diverse modality data, thereby disrupting the discriminator and reinforcing the differences between modalities. We conducted experiments on three widely used cross-modality datasets, namely RegDB, SYSU, and LLCM. The results not only demonstrate the effectiveness of our method but also provide insights for future improvements in the robustness of cross-modality ReID systems.
comment: Accepted at the Thirty-eighth Annual Conference on Neural Information Processing Systems (NeurIPS 2024)
♻ ☆ 3D Adaptive Structural Convolution Network for Domain-Invariant Point Cloud Recognition
Adapting deep learning networks for point cloud data recognition in self-driving vehicles faces challenges due to the variability in datasets and sensor technologies, emphasizing the need for adaptive techniques to maintain accuracy across different conditions. In this paper, we introduce the 3D Adaptive Structural Convolution Network (3D-ASCN), a cutting-edge framework for 3D point cloud recognition. It combines 3D convolution kernels, a structural tree structure, and adaptive neighborhood sampling for effective geometric feature extraction. This method obtains domain-invariant features and demonstrates robust, adaptable performance on a variety of point cloud datasets, ensuring compatibility across diverse sensor configurations without the need for parameter adjustments. This highlights its potential to significantly enhance the reliability and efficiency of self-driving vehicle technology.
comment: 11 pages, 3 figures
♻ ☆ Warped Diffusion: Solving Video Inverse Problems with Image Diffusion Models NeurIPS 2024
Using image models naively for solving inverse video problems often suffers from flickering, texture-sticking, and temporal inconsistency in generated videos. To tackle these problems, in this paper, we view frames as continuous functions in the 2D space, and videos as a sequence of continuous warping transformations between different frames. This perspective allows us to train function space diffusion models only on images and utilize them to solve temporally correlated inverse problems. The function space diffusion models need to be equivariant with respect to the underlying spatial transformations. To ensure temporal consistency, we introduce a simple post-hoc test-time guidance towards (self)-equivariant solutions. Our method allows us to deploy state-of-the-art latent diffusion models such as Stable Diffusion XL to solve video inverse problems. We demonstrate the effectiveness of our method for video inpainting and $8\times$ video super-resolution, outperforming existing techniques based on noise transformations. We provide generated video results: https://giannisdaras.github.io/warped_diffusion.github.io/.
comment: Accepted in NeurIPS 2024
♻ ☆ SCMIL: Sparse Context-aware Multiple Instance Learning for Predicting Cancer Survival Probability Distribution in Whole Slide Images MICCAI2024
Cancer survival prediction is a challenging task that involves analyzing of the tumor microenvironment within Whole Slide Image (WSI). Previous methods cannot effectively capture the intricate interaction features among instances within the local area of WSI. Moreover, existing methods for cancer survival prediction based on WSI often fail to provide better clinically meaningful predictions. To overcome these challenges, we propose a Sparse Context-aware Multiple Instance Learning (SCMIL) framework for predicting cancer survival probability distributions. SCMIL innovatively segments patches into various clusters based on their morphological features and spatial location information, subsequently leveraging sparse self-attention to discern the relationships between these patches with a context-aware perspective. Considering many patches are irrelevant to the task, we introduce a learnable patch filtering module called SoftFilter, which ensures that only interactions between task-relevant patches are considered. To enhance the clinical relevance of our prediction, we propose a register-based mixture density network to forecast the survival probability distribution for individual patients. We evaluate SCMIL on two public WSI datasets from the The Cancer Genome Atlas (TCGA) specifically focusing on lung adenocarcinom (LUAD) and kidney renal clear cell carcinoma (KIRC). Our experimental results indicate that SCMIL outperforms current state-of-the-art methods for survival prediction, offering more clinically meaningful and interpretable outcomes. Our code is accessible at https://github.com/yang-ze-kang/SCMIL.
comment: Accepted by MICCAI2024
♻ ☆ Multi-Layer Feature Fusion with Cross-Channel Attention-Based U-Net for Kidney Tumor Segmentation
Renal tumors, especially renal cell carcinoma (RCC), show significant heterogeneity, posing challenges for diagnosis using radiology images such as MRI, echocardiograms, and CT scans. U-Net based deep learning techniques are emerging as a promising approach for automated medical image segmentation for minimally invasive diagnosis of renal tumors. However, current techniques need further improvements in accuracy to become clinically useful to radiologists. In this study, we present an improved U-Net based model for end-to-end automated semantic segmentation of CT scan images to identify renal tumors. The model uses residual connections across convolution layers, integrates a multi-layer feature fusion (MFF) and cross-channel attention (CCA) within encoder blocks, and incorporates skip connections augmented with additional information derived using MFF and CCA. We evaluated our model on the KiTS19 dataset, which contains data from 210 patients. For kidney segmentation, our model achieves a Dice Similarity Coefficient (DSC) of 0.97 and a Jaccard index (JI) of 0.95. For renal tumor segmentation, our model achieves a DSC of 0.96 and a JI of 0.91. Based on a comparison of available DSC scores, our model outperforms the current leading models.
comment: 8 pages
♻ ☆ CHITNet: A Complementary to Harmonious Information Transfer Network for Infrared and Visible Image Fusion
Current infrared and visible image fusion (IVIF) methods go to great lengths to excavate complementary features and design complex fusion strategies, which is extremely challenging. To this end, we rethink the IVIF outside the box, proposing a complementary to harmonious information transfer network (CHITNet). It reasonably transfers complementary information into harmonious one, which integrates both the shared and complementary features from two modalities. Specifically, to skillfully sidestep aggregating complementary information in IVIF, we design a mutual information transfer (MIT) module to mutually represent features from two modalities, roughly transferring complementary information into harmonious one. Then, a harmonious information acquisition supervised by source image (HIASSI) module is devised to further ensure the complementary to harmonious information transfer after MIT. Meanwhile, we also propose a structure information preservation (SIP) module to guarantee that the edge structure information of the source images can be transferred to the fusion results. Moreover, a mutual promotion training paradigm with interaction loss is adopted to facilitate better collaboration among MIT, HIASSI and SIP. In this way, the proposed method is able to generate fused images with higher qualities. Extensive experimental results demonstrate the superiority of CHITNet over state-of-the-art algorithms in terms of visual quality and quantitative evaluations.
♻ ☆ Forgettable Federated Linear Learning with Certified Data Unlearning
The advent of Federated Learning (FL) has revolutionized the way distributed systems handle collaborative model training while preserving user privacy. Recently, Federated Unlearning (FU) has emerged to address demands for the "right to be forgotten"" and unlearning of the impact of poisoned clients without requiring retraining in FL. Most FU algorithms require the cooperation of retained or target clients (clients to be unlearned), introducing additional communication overhead and potential security risks. In addition, some FU methods need to store historical models to execute the unlearning process. These challenges hinder the efficiency and memory constraints of the current FU methods. Moreover, due to the complexity of nonlinear models and their training strategies, most existing FU methods for deep neural networks (DNN) lack theoretical certification. In this work, we introduce a novel FL training and unlearning strategy in DNN, termed Forgettable Federated Linear Learning (F^2L^2). F^2L^2 considers a common practice of using pre-trained models to approximate DNN linearly, allowing them to achieve similar performance as the original networks via Federated Linear Training (FLT). We then present FedRemoval, a certified, efficient, and secure unlearning strategy that enables the server to unlearn a target client without requiring client communication or adding additional storage. We have conducted extensive empirical validation on small- to large-scale datasets, using both convolutional neural networks and modern foundation models. These experiments demonstrate the effectiveness of F^2L^2 in balancing model accuracy with the successful unlearning of target clients. F^2L^2 represents a promising pipeline for efficient and trustworthy FU. The code is available here.
comment: 17 pages, 6 figures
♻ ☆ Towards Domain Adaptive Neural Contextual Bandits
Contextual bandit algorithms are essential for solving real-world decision making problems. In practice, collecting a contextual bandit's feedback from different domains may involve different costs. For example, measuring drug reaction from mice (as a source domain) and humans (as a target domain). Unfortunately, adapting a contextual bandit algorithm from a source domain to a target domain with distribution shift still remains a major challenge and largely unexplored. In this paper, we introduce the first general domain adaptation method for contextual bandits. Our approach learns a bandit model for the target domain by collecting feedback from the source domain. Our theoretical analysis shows that our algorithm maintains a sub-linear regret bound even adapting across domains. Empirical results show that our approach outperforms the state-of-the-art contextual bandit algorithms on real-world datasets.
♻ ☆ EP-SAM: Weakly Supervised Histopathology Segmentation via Enhanced Prompt with Segment Anything
This work proposes a novel approach beyond supervised learning for effective pathological image analysis, addressing the challenge of limited robust labeled data. Pathological diagnosis of diseases like cancer has conventionally relied on the evaluation of morphological features by physicians and pathologists. However, recent advancements in compute-aided diagnosis (CAD) systems are gaining significant attention as diagnostic support tools. Although the advancement of deep learning has improved CAD significantly, segmentation models typically require large pixel-level annotated dataset, and such labeling is expensive. Existing studies not based on supervised approaches still struggle with limited generalization, and no practical approach has emerged yet. To address this issue, we present a weakly supervised semantic segmentation (WSSS) model by combining class activation map and Segment Anything Model (SAM)-based pseudo-labeling. For effective pretraining, we adopt the SAM-a foundation model that is pretrained on large datasets and operates in zero-shot configurations using only coarse prompts. The proposed approach transfer enhanced Attention Dropout Layer's knowledge to SAM, thereby generating pseudo-labels. To demonstrate the superiority of the proposed method, experimental studies are conducted on histopathological breast cancer datasets. The proposed method outperformed other WSSS methods across three datasets, demonstrating its efficiency by achieving this with only 12GB of GPU memory during training. Our code is available at : https://github.com/QI-NemoSong/EP-SAM
comment: 10 pages, 7 figures
♻ ☆ Hierarchical Simplicity Bias of Neural Networks
Neural networks often exhibit simplicity bias, favoring simpler features over more complex ones, even when both are equally predictive. We introduce a novel method called imbalanced label coupling to explore and extend this simplicity bias across multiple hierarchical levels. Our approach demonstrates that trained networks sequentially consider features of increasing complexity based on their correlation with labels in the training set, regardless of their actual predictive power. For example, in CIFAR-10, simple spurious features can cause misclassifications where most cats are predicted as dogs and most trucks as automobiles. We empirically show that last-layer retraining with target data distribution \citep{kirichenko2022last} is insufficient to fully recover core features when spurious features perfectly correlate with target labels in our synthetic datasets. Our findings deepen the understanding of the implicit biases inherent in neural networks.
comment: 20 pages, 21 figures, revised version, accepted at OPT2024: 16th Annual Workshop on Optimization for Machine Learning
♻ ☆ Compress Guidance in Conditional Diffusion Sampling
We found that enforcing guidance throughout the sampling process is often counterproductive due to the model-fitting issue, where samples are 'tuned' to match the classifier's parameters rather than generalizing the expected condition. This work identifies and quantifies the problem, demonstrating that reducing or excluding guidance at numerous timesteps can mitigate this issue. By distributing a small amount of guidance over a large number of sampling timesteps, we observe a significant improvement in image quality and diversity while also reducing the required guidance timesteps by nearly 40%. This approach addresses a major challenge in applying guidance effectively to generative tasks. Consequently, our proposed method, termed Compress Guidance, allows for the exclusion of a substantial number of guidance timesteps while still surpassing baseline models in image quality. We validate our approach through benchmarks on label-conditional and text-to-image generative tasks across various datasets and models.
comment: 10 pages, 5 figures, Computer Vision and Machine Learning
♻ ☆ NVLM: Open Frontier-Class Multimodal LLMs
We introduce NVLM 1.0, a family of frontier-class multimodal large language models (LLMs) that achieve state-of-the-art results on vision-language tasks, rivaling the leading proprietary models (e.g., GPT-4o) and open-access models (e.g., Llama 3-V 405B and InternVL 2). Remarkably, NVLM 1.0 shows improved text-only performance over its LLM backbone after multimodal training. In terms of model design, we perform a comprehensive comparison between decoder-only multimodal LLMs (e.g., LLaVA) and cross-attention-based models (e.g., Flamingo). Based on the strengths and weaknesses of both approaches, we propose a novel architecture that enhances both training efficiency and multimodal reasoning capabilities. Furthermore, we introduce a 1-D tile-tagging design for tile-based dynamic high-resolution images, which significantly boosts performance on multimodal reasoning and OCR-related tasks. Regarding training data, we meticulously curate and provide detailed information on our multimodal pretraining and supervised fine-tuning datasets. Our findings indicate that dataset quality and task diversity are more important than scale, even during the pretraining phase, across all architectures. Notably, we develop production-grade multimodality for the NVLM-1.0 models, enabling them to excel in vision-language tasks while maintaining and even improving text-only performance compared to their LLM backbones. To achieve this, we craft and integrate a high-quality text-only dataset into multimodal training, alongside a substantial amount of multimodal math and reasoning data, leading to enhanced math and coding capabilities across modalities. To advance research in the field, we release the model weights at https://huggingface.co/nvidia/NVLM-D-72B and will open-source the training code for the community soon.
comment: Fixed the typos. For more information, please visit our project page at: https://research.nvidia.com/labs/adlr/NVLM-1
♻ ☆ MoRE: Multi-Modal Contrastive Pre-training with Transformers on X-Rays, ECGs, and Diagnostic Report
In this paper, we introduce a novel Multi-Modal Contrastive Pre-training Framework that synergistically combines X-rays, electrocardiograms (ECGs), and radiology/cardiology reports. Our approach leverages transformers to encode these diverse modalities into a unified representation space, aiming to enhance diagnostic accuracy and facilitate comprehensive patient assessments. We utilize LoRA-Peft to significantly reduce trainable parameters in the LLM and incorporate recent linear attention dropping strategy in the Vision Transformer(ViT) for smoother attention. Furthermore, we provide novel multimodal attention explanations and retrieval for our model. To the best of our knowledge, we are the first to propose an integrated model that combines X-ray, ECG, and Radiology/Cardiology Report with this approach. By utilizing contrastive loss, MoRE effectively aligns modality-specific features into a coherent embedding, which supports various downstream tasks such as zero-shot classification and multimodal retrieval. Employing our proposed methodology, we achieve state-of-the-art (SOTA) on the Mimic-IV, CheXpert, Edema Severity, and PtbXl downstream datasets, surpassing existing multimodal approaches. Our proposed framework shows significant improvements in capturing intricate inter-modal relationships and its robustness in medical diagnosis that establishes a framework for future research in multimodal learning in the healthcare sector.
comment: 10 pages, 5 figures, 9 tables. Supplementary detail in Appendix. Code made available in Github for reproducibility
♻ ☆ RetriBooru: Leakage-Free Retrieval of Conditions from Reference Images for Subject-Driven Generation
Diffusion-based methods have demonstrated remarkable capabilities in generating a diverse array of high-quality images, sparking interests for styled avatars, virtual try-on, and more. Previous methods use the same reference image as the target. An overlooked aspect is the leakage of the target's spatial information, style, etc. from the reference, harming the generated diversity and causing shortcuts. However, this approach continues as widely available datasets usually consist of single images not grouped by identities, and it is expensive to recollect large-scale same-identity data. Moreover, existing metrics adopt decoupled evaluation on text alignment and identity preservation, which fail at distinguishing between balanced outputs and those that over-fit to one aspect. In this paper, we propose a multi-level, same-identity dataset RetriBooru, which groups anime characters by both face and cloth identities. RetriBooru enables adopting reference images of the same character and outfits as the target, while keeping flexible gestures and actions. We benchmark previous methods on our dataset, and demonstrate the effectiveness of training with a reference image different from target (but same identity). We introduce a new concept composition task, where the conditioning encoder learns to retrieve different concepts from several reference images, and modify a baseline network RetriNet for the new task. Finally, we introduce a novel class of metrics named Similarity Weighted Diversity (SWD), to measure the overlooked diversity and better evaluate the alignment between similarity and diversity.
♻ ☆ AUPIMO: Redefining Visual Anomaly Detection Benchmarks with High Speed and Low Tolerance BMVC 2024
Recent advances in visual anomaly detection research have seen AUROC and AUPRO scores on public benchmark datasets such as MVTec and VisA converge towards perfect recall, giving the impression that these benchmarks are near-solved. However, high AUROC and AUPRO scores do not always reflect qualitative performance, which limits the validity of these metrics in real-world applications. We argue that the artificial ceiling imposed by the lack of an adequate evaluation metric restrains progression of the field, and it is crucial that we revisit the evaluation metrics used to rate our algorithms. In response, we introduce Per-IMage Overlap (PIMO), a novel metric that addresses the shortcomings of AUROC and AUPRO. PIMO retains the recall-based nature of the existing metrics but introduces two distinctions: the assignment of curves (and respective area under the curve) is per-image, and its X-axis relies solely on normal images. Measuring recall per image simplifies instance score indexing and is more robust to noisy annotations. As we show, it also accelerates computation and enables the usage of statistical tests to compare models. By imposing low tolerance for false positives on normal images, PIMO provides an enhanced model validation procedure and highlights performance variations across datasets. Our experiments demonstrate that PIMO offers practical advantages and nuanced performance insights that redefine anomaly detection benchmarks -- notably challenging the perception that MVTec AD and VisA datasets have been solved by contemporary models. Available on GitHub: https://github.com/jpcbertoldo/aupimo.
comment: Accepted to BMVC 2024. Official implementation: https://github.com/jpcbertoldo/aupimo. Integrated in anomalib https://github.com/openvinotoolkit/anomalib. This research was conducted during Google Summer of Code 2023 (GSoC 2023) with the anomalib team from Intel's OpenVINO Toolkit
♻ ☆ XReal: Realistic Anatomy and Pathology-Aware X-ray Generation via Controllable Diffusion Model
Large-scale generative models have demonstrated impressive capabilities in producing visually compelling images, with increasing applications in medical imaging. However, they continue to grapple with hallucination challenges and the generation of anatomically inaccurate outputs. These limitations are mainly due to the reliance on textual inputs and lack of spatial control over the generated images, hindering the potential usefulness of such models in real-life settings. In this work, we present XReal, a novel controllable diffusion model for generating realistic chest X-ray images through precise anatomy and pathology location control. Our lightweight method comprises an Anatomy Controller and a Pathology Controller to introduce spatial control over anatomy and pathology in a pre-trained Text-to-Image Diffusion Model, respectively, without fine-tuning the model. XReal outperforms state-of-the-art X-ray diffusion models in quantitative metrics and radiologists' ratings, showing significant gains in anatomy and pathology realism. Our model holds promise for advancing generative models in medical imaging, offering greater precision and adaptability while inviting further exploration in this evolving field. The code and pre-trained model weights are publicly available at https://github.com/BioMedIA-MBZUAI/XReal.
♻ ☆ Fine-tuning with Very Large Dropout
It is impossible today to pretend that the practice of machine learning is compatible with the idea that training and testing data follow the same distribution. Several authors have recently used ensemble techniques to show how scenarios involving multiple data distributions are best served by representations that are both richer than those obtained by regularizing for the best in-distribution performance, and richer than those obtained under the influence of the implicit sparsity bias of common stochastic gradient procedures. This contribution investigates the use of very high dropout rates instead of ensembles to obtain such rich representations. Although training a deep network from scratch using such dropout rates is virtually impossible, fine-tuning a large pre-trained model under such conditions is not only possible but also achieves out-of-distribution performances that exceed those of both ensembles and weight averaging methods such as model soups. This result has practical significance because the importance of the fine-tuning scenario has considerably grown in recent years. This result also provides interesting insights on the nature of rich representations and on the intrinsically linear nature of fine-tuning a large network using a comparatively small dataset.
comment: Fine-tuning with very large dropout outperforms weight-averaging and ensemble on ResNet and large vision transformer
♻ ☆ D2S: Representing sparse descriptors and 3D coordinates for camera relocalization
State-of-the-art visual localization methods mostly rely on complex procedures to match local descriptors and 3D point clouds. However, these procedures can incur significant costs in terms of inference, storage, and updates over time. In this study, we propose a direct learning-based approach that utilizes a simple network named D2S to represent complex local descriptors and their scene coordinates. Our method is characterized by its simplicity and cost-effectiveness. It solely leverages a single RGB image for localization during the testing phase and only requires a lightweight model to encode a complex sparse scene. The proposed D2S employs a combination of a simple loss function and graph attention to selectively focus on robust descriptors while disregarding areas such as clouds, trees, and several dynamic objects. This selective attention enables D2S to effectively perform a binary-semantic classification for sparse descriptors. Additionally, we propose a simple outdoor dataset to evaluate the capabilities of visual localization methods in scene-specific generalization and self-updating from unlabeled observations. Our approach outperforms the previous regression-based methods in both indoor and outdoor environments. It demonstrates the ability to generalize beyond training data, including scenarios involving transitions from day to night and adapting to domain shifts. The source code, trained models, dataset, and demo videos are available at the following link: https://thpjp.github.io/d2s.
comment: Accepted to IEEE Robotics and Automation Letters
♻ ☆ DEAR: Disentangled Environment and Agent Representations for Reinforcement Learning without Reconstruction IROS 2024
Reinforcement Learning (RL) algorithms can learn robotic control tasks from visual observations, but they often require a large amount of data, especially when the visual scene is complex and unstructured. In this paper, we explore how the agent's knowledge of its shape can improve the sample efficiency of visual RL methods. We propose a novel method, Disentangled Environment and Agent Representations (DEAR), that uses the segmentation mask of the agent as supervision to learn disentangled representations of the environment and the agent through feature separation constraints. Unlike previous approaches, DEAR does not require reconstruction of visual observations. These representations are then used as an auxiliary loss to the RL objective, encouraging the agent to focus on the relevant features of the environment. We evaluate DEAR on two challenging benchmarks: Distracting DeepMind control suite and Franka Kitchen manipulation tasks. Our findings demonstrate that DEAR surpasses state-of-the-art methods in sample efficiency, achieving comparable or superior performance with reduced parameters. Our results indicate that integrating agent knowledge into visual RL methods has the potential to enhance their learning efficiency and robustness.
comment: 6 pages, 7 figures, 2 tables. Accepted at 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2024)
♻ ☆ Counter-Hypothetical Particle Filters for Single Object Pose Tracking ICRA
Particle filtering is a common technique for six degrees of freedom (6D) pose estimation due to its ability to tractably represent belief over object pose. However, the particle filter is prone to particle deprivation due to the high-dimensional nature of 6D pose. When particle deprivation occurs, it can cause mode collapse of the underlying belief distribution during importance sampling. If the region surrounding the true state suffers from mode collapse, recovering its belief is challenging since the area is no longer represented in the probability mass formed by the particles. Previous methods mitigate this problem by randomizing and resetting particles in the belief distribution, but determining the frequency of reinvigoration has relied on hand-tuning abstract heuristics. In this paper, we estimate the necessary reinvigoration rate at each time step by introducing a Counter-Hypothetical likelihood function, which is used alongside the standard likelihood. Inspired by the notions of plausibility and implausibility from Evidential Reasoning, the addition of our Counter-Hypothetical likelihood function assigns a level of doubt to each particle. The competing cumulative values of confidence and doubt across the particle set are used to estimate the level of failure within the filter, in order to determine the portion of particles to be reinvigorated. We demonstrate the effectiveness of our method on the rigid body object 6D pose tracking task.
comment: International Conference on Robotics and Automation (ICRA) 2023
♻ ☆ EF-3DGS: Event-Aided Free-Trajectory 3D Gaussian Splatting
Scene reconstruction from casually captured videos has wide applications in real-world scenarios. With recent advancements in differentiable rendering techniques, several methods have attempted to simultaneously optimize scene representations (NeRF or 3DGS) and camera poses. Despite recent progress, existing methods relying on traditional camera input tend to fail in high-speed (or equivalently low-frame-rate) scenarios. Event cameras, inspired by biological vision, record pixel-wise intensity changes asynchronously with high temporal resolution, providing valuable scene and motion information in blind inter-frame intervals. In this paper, we introduce the event camera to aid scene construction from a casually captured video for the first time, and propose Event-Aided Free-Trajectory 3DGS, called EF-3DGS, which seamlessly integrates the advantages of event cameras into 3DGS through three key components. First, we leverage the Event Generation Model (EGM) to fuse events and frames, supervising the rendered views observed by the event stream. Second, we adopt the Contrast Maximization (CMax) framework in a piece-wise manner to extract motion information by maximizing the contrast of the Image of Warped Events (IWE), thereby calibrating the estimated poses. Besides, based on the Linear Event Generation Model (LEGM), the brightness information encoded in the IWE is also utilized to constrain the 3DGS in the gradient domain. Third, to mitigate the absence of color information of events, we introduce photometric bundle adjustment (PBA) to ensure view consistency across events and frames. We evaluate our method on the public Tanks and Temples benchmark and a newly collected real-world dataset, RealEv-DAVIS. Our project page is https://lbh666.github.io/ef-3dgs/.
comment: Project Page: https://lbh666.github.io/ef-3dgs/
♻ ☆ AI-Driven Approaches for Glaucoma Detection -- A Comprehensive Review
The diagnosis of glaucoma plays a critical role in the management and treatment of this vision-threatening disease. Glaucoma is a group of eye diseases that cause blindness by damaging the optic nerve at the back of the eye. Often called "silent thief of sight", it exhibits no symptoms during the early stages. Therefore, early detection is crucial to prevent vision loss. With the rise of Artificial Intelligence (AI), particularly Deep Learning (DL) techniques, Computer-Aided Diagnosis (CADx) systems have emerged as promising tools to assist clinicians in accurately diagnosing glaucoma early. This paper aims to provide a comprehensive overview of AI techniques utilized in CADx systems for glaucoma diagnosis. Through a detailed analysis of current literature, we identify key gaps and challenges in these systems, emphasizing the need for improved safety, reliability, interpretability, and explainability. By identifying research gaps, we aim to advance the field of CADx systems especially for the early diagnosis of glaucoma, in order to prevent any potential loss of vision.
♻ ☆ ARBEx: Attentive Feature Extraction with Reliability Balancing for Robust Facial Expression Learning ACCV 2024
In this paper, we introduce a framework ARBEx, a novel attentive feature extraction framework driven by Vision Transformer with reliability balancing to cope against poor class distributions, bias, and uncertainty in the facial expression learning (FEL) task. We reinforce several data pre-processing and refinement methods along with a window-based cross-attention ViT to squeeze the best of the data. We also employ learnable anchor points in the embedding space with label distributions and multi-head self-attention mechanism to optimize performance against weak predictions with reliability balancing, which is a strategy that leverages anchor points, attention scores, and confidence values to enhance the resilience of label predictions. To ensure correct label classification and improve the models' discriminative power, we introduce anchor loss, which encourages large margins between anchor points. Additionally, the multi-head self-attention mechanism, which is also trainable, plays an integral role in identifying accurate labels. This approach provides critical elements for improving the reliability of predictions and has a substantial positive effect on final prediction capabilities. Our adaptive model can be integrated with any deep neural network to forestall challenges in various recognition tasks. Our strategy outperforms current state-of-the-art methodologies, according to extensive experiments conducted in a variety of contexts.
comment: Extended version is accepted in ACCV 2024 as GReFEL (arXiv:2410.15927)
♻ ☆ Explaining Chest X-ray Pathology Models using Textual Concepts NeurIPS'24
Deep learning models have revolutionized medical imaging and diagnostics, yet their opaque nature poses challenges for clinical adoption and trust. Amongst approaches to improve model interpretability, concept-based explanations aim to provide concise and human-understandable explanations of any arbitrary classifier. However, such methods usually require a large amount of manually collected data with concept annotation, which is often scarce in the medical domain. In this paper, we propose Conceptual Counterfactual Explanations for Chest X-ray (CoCoX), which leverages the joint embedding space of an existing vision-language model (VLM) to explain black-box classifier outcomes without the need for annotated datasets. Specifically, we utilize textual concepts derived from chest radiography reports and a pre-trained chest radiography-based VLM to explain three common cardiothoracic pathologies. We demonstrate that the explanations generated by our method are semantically meaningful and faithful to underlying pathologies.
comment: Accepted at NeurIPS'24 workshop on Advancements In Medical Foundation Models: Explainability, Robustness, Security, and Beyond (AIM-FM)
Information Retrieval 22
☆ Large Language Models Empowered Personalized Web Agents
Web agents have emerged as a promising direction to automate Web task completion based on user instructions, significantly enhancing user experience. Recently, Web agents have evolved from traditional agents to Large Language Models (LLMs)-based Web agents. Despite their success, existing LLM-based Web agents overlook the importance of personalized data (e.g., user profiles and historical Web behaviors) in assisting the understanding of users' personalized instructions and executing customized actions. To overcome the limitation, we first formulate the task of LLM-empowered personalized Web agents, which integrate personalized data and user instructions to personalize instruction comprehension and action execution. To address the absence of a comprehensive evaluation benchmark, we construct a Personalized Web Agent Benchmark (PersonalWAB), featuring user instructions, personalized user data, Web functions, and two evaluation paradigms across three personalized Web tasks. Moreover, we propose a Personalized User Memory-enhanced Alignment (PUMA) framework to adapt LLMs to the personalized Web agent task. PUMA utilizes a memory bank with a task-specific retrieval strategy to filter relevant historical Web behaviors. Based on the behaviors, PUMA then aligns LLMs for personalized action execution through fine-tuning and direct preference optimization. Extensive experiments validate the superiority of PUMA over existing Web agents on PersonalWAB.
comment: The code and data are available on the project website https://hongrucai.github.io/PersonalWAB/
☆ Improving Pinterest Search Relevance Using Large Language Models CIKM 2024
To improve relevance scoring on Pinterest Search, we integrate Large Language Models (LLMs) into our search relevance model, leveraging carefully designed text representations to predict the relevance of Pins effectively. Our approach uses search queries alongside content representations that include captions extracted from a generative visual language model. These are further enriched with link-based text data, historically high-quality engaged queries, user-curated boards, Pin titles and Pin descriptions, creating robust models for predicting search relevance. We use a semi-supervised learning approach to efficiently scale up the amount of training data, expanding beyond the expensive human labeled data available. By utilizing multilingual LLMs, our system extends training data to include unseen languages and domains, despite initial data and annotator expertise being confined to English. Furthermore, we distill from the LLM-based model into real-time servable model architectures and features. We provide comprehensive offline experimental validation for our proposed techniques and demonstrate the gains achieved through the final deployed system at scale.
comment: CIKM 2024 Workshop on Industrial Recommendation Systems
☆ TELII: Temporal Event Level Inverted Indexing for Cohort Discovery on a Large Covid-19 EHR Dataset
Cohort discovery is a crucial step in clinical research on Electronic Health Record (EHR) data. Temporal queries, which are common in cohort discovery, can be time-consuming and prone to errors when processed on large EHR datasets. In this work, we introduce TELII, a temporal event level inverted indexing method designed for cohort discovery on large EHR datasets. TELII is engineered to pre-compute and store the relations along with the time difference between events, thereby providing fast and accurate temporal query capabilities. We implemented TELII for the OPTUM de-identified COVID-19 EHR dataset, which contains data from 8.87 million patients. We demonstrate four common temporal query tasks and their implementation using TELII with a MongoDB backend. Our results show that the temporal query speed for TELII is up to 2000 times faster than that of existing non-temporal inverted indexes. TELII achieves millisecond-level response times, enabling users to quickly explore event relations and find preliminary evidence for their research questions. Not only is TELII practical and straightforward to implement, but it also offers easy adaptability to other EHR datasets. These advantages underscore TELII's potential to serve as the query engine for EHR-based applications, ensuring fast, accurate, and user-friendly query responses.
☆ Enhancing Answer Attribution for Faithful Text Generation with Large Language Models
The increasing popularity of Large Language Models (LLMs) in recent years has changed the way users interact with and pose questions to AI-based conversational systems. An essential aspect for increasing the trustworthiness of generated LLM answers is the ability to trace the individual claims from responses back to relevant sources that support them, the process known as answer attribution. While recent work has started exploring the task of answer attribution in LLMs, some challenges still remain. In this work, we first perform a case study analyzing the effectiveness of existing answer attribution methods, with a focus on subtasks of answer segmentation and evidence retrieval. Based on the observed shortcomings, we propose new methods for producing more independent and contextualized claims for better retrieval and attribution. The new methods are evaluated and shown to improve the performance of answer attribution components. We end with a discussion and outline of future directions for the task.
comment: Accepted to KDIR 2024 (part of IC3K 2024)
☆ Bridging the Modality Gap: Dimension Information Alignment and Sparse Spatial Constraint for Image-Text Matching
Many contrastive learning based models have achieved advanced performance in image-text matching tasks. The key of these models lies in analyzing the correlation between image-text pairs, which involves cross-modal interaction of embeddings in corresponding dimensions. However, the embeddings of different modalities are from different models or modules, and there is a significant modality gap. Directly interacting such embeddings lacks rationality and may capture inaccurate correlation. Therefore, we propose a novel method called DIAS to bridge the modality gap from two aspects: (1) We align the information representation of embeddings from different modalities in corresponding dimension to ensure the correlation calculation is based on interactions of similar information. (2) The spatial constraints of inter- and intra-modalities unmatched pairs are introduced to ensure the effectiveness of semantic alignment of the model. Besides, a sparse correlation algorithm is proposed to select strong correlated spatial relationships, enabling the model to learn more significant features and avoid being misled by weak correlation. Extensive experiments demonstrate the superiority of DIAS, achieving 4.3\%-10.2\% rSum improvements on Flickr30k and MSCOCO benchmarks.
☆ Neural Collaborative Filtering Classification Model to Obtain Prediction Reliabilities
Neural collaborative filtering is the state of art field in the recommender systems area; it provides some models that obtain accurate predictions and recommendations. These models are regression-based, and they just return rating predictions. This paper proposes the use of a classification-based approach, returning both rating predictions and their reliabilities. The extra information (prediction reliabilities) can be used in a variety of relevant collaborative filtering areas such as detection of shilling attacks, recommendations explanation or navigational tools to show users and items dependences. Additionally, recommendation reliabilities can be gracefully provided to users: "probably you will like this film", "almost certainly you will like this song", etc. This paper provides the proposed neural architecture; it also tests that the quality of its recommendation results is as good as the state of art baselines. Remarkably, individual rating predictions are improved by using the proposed architecture compared to baselines. Experiments have been performed making use of four popular public datasets, showing generalizable quality results. Overall, the proposed architecture improves individual rating predictions quality, maintains recommendation results and opens the doors to a set of relevant collaborative filtering fields.
comment: 9 pages, 7 figures
☆ Bridging Search and Recommendation in Generative Retrieval: Does One Task Help the Other? RecSys'24
Generative retrieval for search and recommendation is a promising paradigm for retrieving items, offering an alternative to traditional methods that depend on external indexes and nearest-neighbor searches. Instead, generative models directly associate inputs with item IDs. Given the breakthroughs of Large Language Models (LLMs), these generative systems can play a crucial role in centralizing a variety of Information Retrieval (IR) tasks in a single model that performs tasks such as query understanding, retrieval, recommendation, explanation, re-ranking, and response generation. Despite the growing interest in such a unified generative approach for IR systems, the advantages of using a single, multi-task model over multiple specialized models are not well established in the literature. This paper investigates whether and when such a unified approach can outperform task-specific models in the IR tasks of search and recommendation, broadly co-existing in multiple industrial online platforms, such as Spotify, YouTube, and Netflix. Previous work shows that (1) the latent representations of items learned by generative recommenders are biased towards popularity, and (2) content-based and collaborative-filtering-based information can improve an item's representations. Motivated by this, our study is guided by two hypotheses: [H1] the joint training regularizes the estimation of each item's popularity, and [H2] the joint training regularizes the item's latent representations, where search captures content-based aspects of an item and recommendation captures collaborative-filtering aspects. Our extensive experiments with both simulated and real-world data support both [H1] and [H2] as key contributors to the effectiveness improvements observed in the unified search and recommendation generative models over the single-task approaches.
comment: Accepted for publication in the 18th ACM Conference on Recommender Systems (RecSys'24)
☆ Beyond Retrieval: Generating Narratives in Conversational Recommender Systems
The recent advances in Large Language Model's generation and reasoning capabilities present an opportunity to develop truly conversational recommendation systems. However, effectively integrating recommender system knowledge into LLMs for natural language generation which is tailored towards recommendation tasks remains a challenge. This paper addresses this challenge by making two key contributions. First, we introduce a new dataset (REGEN) for natural language generation tasks in conversational recommendations. REGEN (Reviews Enhanced with GEnerative Narratives) extends the Amazon Product Reviews dataset with rich user narratives, including personalized explanations of product preferences, product endorsements for recommended items, and summaries of user purchase history. REGEN is made publicly available to facilitate further research. Furthermore, we establish benchmarks using well-known generative metrics, and perform an automated evaluation of the new dataset using a rater LLM. Second, the paper introduces a fusion architecture (CF model with an LLM) which serves as a baseline for REGEN. And to the best of our knowledge, represents the first attempt to analyze the capabilities of LLMs in understanding recommender signals and generating rich narratives. We demonstrate that LLMs can effectively learn from simple fusion architectures utilizing interaction-based CF embeddings, and this can be further enhanced using the metadata and personalization data associated with items. Our experiments show that combining CF and content embeddings leads to improvements of 4-12% in key language metrics compared to using either type of embedding individually. We also provide an analysis to interpret how CF and content embeddings contribute to this new generative task.
☆ Coarse-to-fine Dynamic Uplift Modeling for Real-time Video Recommendation
With the rise of short video platforms, video recommendation technology faces more complex challenges. Currently, there are multiple non-personalized modules in the video recommendation pipeline that urgently need personalized modeling techniques for improvement. Inspired by the success of uplift modeling in online marketing, we attempt to implement uplift modeling in the video recommendation scenario. However, we face two main challenges: 1) Design and utilization of treatments, and 2) Capture of user real-time interest. To address them, we design adjusting the distribution of videos with varying durations as the treatment and propose Coarse-to-fine Dynamic Uplift Modeling (CDUM) for real-time video recommendation. CDUM consists of two modules, CPM and FIC. The former module fully utilizes the offline features of users to model their long-term preferences, while the latter module leverages online real-time contextual features and request-level candidates to model users' real-time interests. These two modules work together to dynamically identify and targeting specific user groups and applying treatments effectively. Further, we conduct comprehensive experiments on the offline public and industrial datasets and online A/B test, demonstrating the superiority and effectiveness of our proposed CDUM. Our proposed CDUM is eventually fully deployed on the Kuaishou platform, serving hundreds of millions of users every day. The source code will be provided after the paper is accepted.
comment: 9 pages, 4 figures, 5 tables
☆ Cutting Through the Confusion and Hype: Understanding the True Potential of Generative AI
This paper explores the nuanced landscape of generative AI (genAI), particularly focusing on neural network-based models like Large Language Models (LLMs). While genAI garners both optimistic enthusiasm and sceptical criticism, this work seeks to provide a balanced examination of its capabilities, limitations, and the profound impact it may have on societal functions and personal interactions. The first section demystifies language-based genAI through detailed discussions on how LLMs learn, their computational needs, distinguishing features from supporting technologies, and the inherent limitations in their accuracy and reliability. Real-world examples illustrate the practical applications and implications of these technologies. The latter part of the paper adopts a systems perspective, evaluating how the integration of LLMs with existing technologies can enhance productivity and address emerging concerns. It highlights the need for significant investment to understand the implications of recent advancements, advocating for a well-informed dialogue to ethically and responsibly integrate genAI into diverse sectors. The paper concludes with prospective developments and recommendations, emphasizing a forward-looking approach to harnessing genAI`s potential while mitigating its risks.
☆ Distill-SynthKG: Distilling Knowledge Graph Synthesis Workflow for Improved Coverage and Efficiency
Knowledge graphs (KGs) generated by large language models (LLMs) are becoming increasingly valuable for Retrieval-Augmented Generation (RAG) applications that require knowledge-intensive reasoning. However, existing KG extraction methods predominantly rely on prompt-based approaches, which are inefficient for processing large-scale corpora. These approaches often suffer from information loss, particularly with long documents, due to the lack of specialized design for KG construction. Additionally, there is a gap in evaluation datasets and methodologies for ontology-free KG construction. To overcome these limitations, we propose SynthKG, a multi-step, document-level ontology-free KG synthesis workflow based on LLMs. By fine-tuning a smaller LLM on the synthesized document-KG pairs, we streamline the multi-step process into a single-step KG generation approach called Distill-SynthKG, substantially reducing the number of LLM inference calls. Furthermore, we re-purpose existing question-answering datasets to establish KG evaluation datasets and introduce new evaluation metrics. Using KGs produced by Distill-SynthKG, we also design a novel graph-based retrieval framework for RAG. Experimental results demonstrate that Distill-SynthKG not only surpasses all baseline models in KG quality -- including models up to eight times larger -- but also consistently excels in retrieval and question-answering tasks. Our proposed graph retrieval framework also outperforms all KG-retrieval methods across multiple benchmark datasets. We release the SynthKG dataset and Distill-SynthKG model publicly to support further research and development.
☆ Optimizing LLMs with Direct Preferences: A Data Efficiency Perspective
Aligning the output of Large Language Models (LLMs) with human preferences (e.g., by means of reinforcement learning with human feedback, or RLHF) is essential for ensuring their effectiveness in real-world scenarios. Despite significant advancements in LLM alignment techniques, the impact of different type of preference data on model performance has yet to be systematically explored. In this study, we investigate the scalability, data efficiency, and effectiveness of Direct Preference Optimization (DPO) in fine-tuning pre-trained LLMs, aiming to reduce their dependency on extensive amounts of preference data, which is expensive to collect. We (1) systematically compare the performance of models fine-tuned with varying percentages of a combined preference judgement dataset to define the improvement curve of DPO and assess its effectiveness in data-constrained environments; and (2) provide insights for the development of an optimal approach for selective preference data usage. Our study reveals that increasing the amount of data used for training generally enhances and stabilizes model performance. Moreover, the use of a combination of diverse datasets significantly improves model effectiveness. Furthermore, when models are trained separately using different types of prompts, models trained with conversational prompts outperformed those trained with question answering prompts.
☆ ProveRAG: Provenance-Driven Vulnerability Analysis with Automated Retrieval-Augmented LLMs
In cybersecurity, security analysts face the challenge of mitigating newly discovered vulnerabilities in real-time, with over 300,000 Common Vulnerabilities and Exposures (CVEs) identified since 1999. The sheer volume of known vulnerabilities complicates the detection of patterns for unknown threats. While LLMs can assist, they often hallucinate and lack alignment with recent threats. Over 25,000 vulnerabilities have been identified so far in 2024, which are introduced after popular LLMs' (e.g., GPT-4) training data cutoff. This raises a major challenge of leveraging LLMs in cybersecurity, where accuracy and up-to-date information are paramount. In this work, we aim to improve the adaptation of LLMs in vulnerability analysis by mimicking how analysts perform such tasks. We propose ProveRAG, an LLM-powered system designed to assist in rapidly analyzing CVEs with automated retrieval augmentation of web data while self-evaluating its responses with verifiable evidence. ProveRAG incorporates a self-critique mechanism to help alleviate omission and hallucination common in the output of LLMs applied in cybersecurity applications. The system cross-references data from verifiable sources (NVD and CWE), giving analysts confidence in the actionable insights provided. Our results indicate that ProveRAG excels in delivering verifiable evidence to the user with over 99% and 97% accuracy in exploitation and mitigation strategies, respectively. This system outperforms direct prompting and chunking retrieval in vulnerability analysis by overcoming temporal and context-window limitations. ProveRAG guides analysts to secure their systems more effectively while documenting the process for future audits.
☆ Captions Speak Louder than Images (CASLIE): Generalizing Foundation Models for E-commerce from High-quality Multimodal Instruction Data
Leveraging multimodal data to drive breakthroughs in e-commerce applications through Multimodal Foundation Models (MFMs) is gaining increasing attention from the research community. However, there are significant challenges that hinder the optimal use of multimodal e-commerce data by foundation models: (1) the scarcity of large-scale, high-quality multimodal benchmark datasets; and (2) the lack of effective multimodal information integration methods. To address these challenges, in this paper, we introduce MMECInstruct, the first-ever, large-scale, and high-quality multimodal instruction dataset for e-commerce. We also develop CASLIE, a simple, lightweight, yet effective framework for integrating multimodal information for e-commerce. Leveraging MMECInstruct, we fine-tune a series of e-commerce MFMs within CASLIE, denoted as CASLIE models. Our comprehensive evaluation demonstrates that CASLIE models substantially outperform 5 categories of advanced baseline models in the in-domain evaluation. Moreover, CASLIE models show strong generalizability to out-of-domain settings. MMECInstruct and CASLIE models are publicly accessible through https://ninglab.github.io/CASLIE/.
comment: Xinyi Ling and Bo Peng contributed equally to this paper
☆ Offline Evaluation of Set-Based Text-to-Image Generation
Text-to-Image (TTI) systems often support people during ideation, the early stages of a creative process when exposure to a broad set of relevant images can help explore the design space. Since ideation is an important subclass of TTI tasks, understanding how to quantitatively evaluate TTI systems according to how well they support ideation is crucial to promoting research and development for these users. However, existing evaluation metrics for TTI remain focused on distributional similarity metrics like Fr\'echet Inception Distance (FID). We take an alternative approach and, based on established methods from ranking evaluation, develop TTI evaluation metrics with explicit models of how users browse and interact with sets of spatially arranged generated images. Our proposed offline evaluation metrics for TTI not only capture how relevant generated images are with respect to the user's ideation need but also take into consideration the diversity and arrangement of the set of generated images. We analyze our proposed family of TTI metrics using human studies on image grids generated by three different TTI systems based on subsets of the widely used benchmarks such as MS-COCO captions and Localized Narratives as well as prompts used in naturalistic settings. Our results demonstrate that grounding metrics in how people use systems is an important and understudied area of benchmark design.
♻ ☆ BRIGHT: A Realistic and Challenging Benchmark for Reasoning-Intensive Retrieval
Existing retrieval benchmarks primarily consist of information-seeking queries (e.g., aggregated questions from search engines) where keyword or semantic-based retrieval is usually sufficient. However, many complex real-world queries require in-depth reasoning to identify relevant documents that go beyond surface form matching. For example, finding documentation for a coding question requires understanding the logic and syntax of the functions involved. To better benchmark retrieval on such challenging queries, we introduce BRIGHT, the first text retrieval benchmark that requires intensive reasoning to retrieve relevant documents. Our dataset consists of 1,384 real-world queries spanning diverse domains, such as economics, psychology, mathematics, and coding. These queries are drawn from naturally occurring and carefully curated human data. Extensive evaluation reveals that even state-of-the-art retrieval models perform poorly on BRIGHT. The leading model on the MTEB leaderboard (Muennighoff et al., 2023), which achieves a score of 59.0 nDCG@10, produces a score of nDCG@10 of 18.3 on BRIGHT. We show that incorporating explicit reasoning about the query improves retrieval performance by up to 12.2 points. Moreover, incorporating retrieved documents from the top-performing retriever boosts question-answering performance by over 6.6 points. We believe that BRIGHT paves the way for future research on retrieval systems in more realistic and challenging settings.
comment: 48 pages
♻ ☆ RecPrompt: A Self-tuning Prompting Framework for News Recommendation Using Large Language Models
News recommendations heavily rely on Natural Language Processing (NLP) methods to analyze, understand, and categorize content, enabling personalized suggestions based on user interests and reading behaviors. Large Language Models (LLMs) like GPT-4 have shown promising performance in understanding natural language. However, the extent of their applicability to news recommendation systems remains to be validated. This paper introduces RecPrompt, the first self-tuning prompting framework for news recommendation, leveraging the capabilities of LLMs to perform complex news recommendation tasks. This framework incorporates a news recommender and a prompt optimizer that applies an iterative bootstrapping process to enhance recommendations through automatic prompt engineering. Extensive experimental results with 400 users show that RecPrompt can achieve an improvement of 3.36% in AUC, 10.49% in MRR, 9.64% in nDCG@5, and 6.20% in nDCG@10 compared to deep neural models. Additionally, we introduce TopicScore, a novel metric to assess explainability by evaluating LLM's ability to summarize topics of interest for users. The results show LLM's effectiveness in accurately identifying topics of interest and delivering comprehensive topic-based explanations.
comment: 5 pages, 2 figures, and 2 tables
♻ ☆ Research on Travel Route Planing Problems Based on Greedy Algorithm
The route planning problem based on the greedy algorithm represents a method of identifying the optimal or near-optimal route between a given start point and end point. In this paper, the PCA method is employed initially to downscale the city evaluation indexes, extract the key principal components, and then downscale the data using the KMO and TOPSIS algorithms, all of which are based on the MindSpore framework. Secondly, for the dataset that does not pass the KMO test, the entropy weight method and TOPSIS method will be employed for comprehensive evaluation. Finally, a route planning algorithm is proposed and optimised based on the greedy algorithm, which provides personalised route customisation according to the different needs of tourists. In addition, the local travelling efficiency, the time required to visit tourist attractions and the necessary daily breaks are considered in order to reduce the cost and avoid falling into the locally optimal solution.
comment: 5 pages, 3 figures
♻ ☆ Generate and Instantiate What You Prefer: Text-Guided Diffusion for Sequential Recommendation
Recent advancements in generative recommendation systems, particularly in the realm of sequential recommendation tasks, have shown promise in enhancing generalization to new items. Among these approaches, diffusion-based generative recommendation has emerged as an effective tool, leveraging its ability to capture data distributions and generate high-quality samples. Despite effectiveness, two primary challenges have been identified: 1) the lack of consistent modeling of data distribution for oracle items; and 2) the difficulty in scaling to more informative control signals beyond historical interactions. These issues stem from the uninformative nature of ID embeddings, which necessitate random initialization and limit the incorporation of additional control signals. To address these limitations, we propose iDreamRec to involve more concrete prior knowledge to establish item embeddings, particularly through detailed item text descriptions and advanced Text Embedding Models (TEM). More importantly, by converting item descriptions into embeddings aligned with TEM, we enable the integration of intention instructions as control signals to guide the generation of oracle items. Experimental results on four datasets demonstrate that iDreamRec not only outperforms existing diffusion-based generative recommenders but also facilitates the incorporation of intention instructions for more precise and effective recommendation generation.
♻ ☆ A Text is Worth Several Tokens: Text Embedding from LLMs Secretly Aligns Well with The Key Tokens
Text embeddings from large language models (LLMs) have achieved excellent results in tasks such as information retrieval, semantic textual similarity, etc. In this work, we show an interesting finding: when feeding a text into the embedding LLMs, the obtained text embedding will be able to be aligned with the key tokens in the input text. We first fully analyze this phenomenon on eight embedding LLMs and show that this phenomenon is universal and is not affected by model architecture, training strategy, and embedding method. With a deeper analysis, we then find that the main change in embedding space between the embedding LLMs and their original generative LLMs is in the first principal component. By adjusting the first principal component, we can align text embedding with the key tokens. Finally, we give several examples to demonstrate the vast application potential of this finding: (1) we propose a simple and practical sparse retrieval method based on the aligned tokens, which can achieve 80\% of the dense retrieval effect of the same model while reducing the computation significantly; (2) we show that our findings provide a fresh perspective to help understand fuzzy concepts (e.g., semantic relatedness vs. semantic similarity) and emerging technologies (e.g., instruction-following embedding) in this field.
comment: Work in Progress
♻ ☆ Automatic Search of Multiword Place Names on Historical Maps SP
Historical maps are invaluable sources of information about the past, and scanned historical maps are increasingly accessible in online libraries. To retrieve maps from these large libraries that contain specific places of interest, previous work has applied computer vision techniques to recognize words on historical maps, enabling searches for maps that contain specific place names. However, searching for multiword place names is challenging due to complex layouts of text labels on historical maps. This paper proposes an efficient query method for searching a given multiword place name on historical maps. Using existing methods to recognize words on historical maps, we link single-word text labels into potential multiword phrases by constructing minimum spanning trees. These trees aim to link pairs of text labels that are spatially close and have similar height, angle, and capitalization. We then query these trees for the given multiword place name. We evaluate the proposed method in two experiments: 1) to evaluate the accuracy of the minimum spanning tree approach at linking multiword place names and 2) to evaluate the number and time range of maps retrieved by the query approach. The resulting maps reveal how places using multiword names have changed on a large number of maps from across history.
comment: 4 pages, 4 figures, and 2 tables. To be published in proceedings ACM SIGSPATIAL 2024 GeoSearch Workshop
♻ ☆ Token-wise Influential Training Data Retrieval for Large Language Models ACL 2024
Given a Large Language Model (LLM) generation, how can we identify which training data led to this generation? In this paper, we proposed RapidIn, a scalable framework adapting to LLMs for estimating the influence of each training data. The proposed framework consists of two stages: caching and retrieval. First, we compress the gradient vectors by over 200,000x, allowing them to be cached on disk or in GPU/CPU memory. Then, given a generation, RapidIn efficiently traverses the cached gradients to estimate the influence within minutes, achieving over a 6,326x speedup. Moreover, RapidIn supports multi-GPU parallelization to substantially accelerate caching and retrieval. Our empirical result confirms the efficiency and effectiveness of RapidIn.
comment: Accepted to ACL 2024. Keywords: Influence Function, Influence Estimation, Training Data Attribution
Machine Learning 208
☆ LVSM: A Large View Synthesis Model with Minimal 3D Inductive Bias
We propose the Large View Synthesis Model (LVSM), a novel transformer-based approach for scalable and generalizable novel view synthesis from sparse-view inputs. We introduce two architectures: (1) an encoder-decoder LVSM, which encodes input image tokens into a fixed number of 1D latent tokens, functioning as a fully learned scene representation, and decodes novel-view images from them; and (2) a decoder-only LVSM, which directly maps input images to novel-view outputs, completely eliminating intermediate scene representations. Both models bypass the 3D inductive biases used in previous methods -- from 3D representations (e.g., NeRF, 3DGS) to network designs (e.g., epipolar projections, plane sweeps) -- addressing novel view synthesis with a fully data-driven approach. While the encoder-decoder model offers faster inference due to its independent latent representation, the decoder-only LVSM achieves superior quality, scalability, and zero-shot generalization, outperforming previous state-of-the-art methods by 1.5 to 3.5 dB PSNR. Comprehensive evaluations across multiple datasets demonstrate that both LVSM variants achieve state-of-the-art novel view synthesis quality. Notably, our models surpass all previous methods even with reduced computational resources (1-2 GPUs). Please see our website for more details: https://haian-jin.github.io/projects/LVSM/ .
comment: project page: https://haian-jin.github.io/projects/LVSM/
☆ SELA: Tree-Search Enhanced LLM Agents for Automated Machine Learning
Automated Machine Learning (AutoML) approaches encompass traditional methods that optimize fixed pipelines for model selection and ensembling, as well as newer LLM-based frameworks that autonomously build pipelines. While LLM-based agents have shown promise in automating machine learning tasks, they often generate low-diversity and suboptimal code, even after multiple iterations. To overcome these limitations, we introduce Tree-Search Enhanced LLM Agents (SELA), an innovative agent-based system that leverages Monte Carlo Tree Search (MCTS) to optimize the AutoML process. By representing pipeline configurations as trees, our framework enables agents to conduct experiments intelligently and iteratively refine their strategies, facilitating a more effective exploration of the machine learning solution space. This novel approach allows SELA to discover optimal pathways based on experimental feedback, improving the overall quality of the solutions. In an extensive evaluation across 20 machine learning datasets, we compare the performance of traditional and agent-based AutoML methods, demonstrating that SELA achieves a win rate of 65% to 80% against each baseline across all datasets. These results underscore the significant potential of agent-based strategies in AutoML, offering a fresh perspective on tackling complex machine learning challenges.
comment: The code is available at https://github.com/geekan/MetaGPT
☆ Fine-Tuning Large Language Models to Appropriately Abstain with Semantic Entropy NeurIPS
Large Language Models (LLMs) are known to hallucinate, whereby they generate plausible but inaccurate text. This phenomenon poses significant risks in critical applications, such as medicine or law, necessitating robust hallucination mitigation strategies. While recent works have proposed fine-tuning methods to teach LLMs to abstain from answering questions beyond their knowledge or capabilities, these methods rely on the existence of ground-truth labels or are limited to short-form responses. To address these limitations, we propose fine-tuning using semantic entropy, an uncertainty measure derived from introspection into the model which does not require external labels. We demonstrate that our approach matches or outperforms models fine-tuned using prior work and achieves strong performance for both short and long-form generations on a range of datasets.
comment: Accepted to NeurIPS Safe Generative AI Workshop 2024
☆ Few-shot In-Context Preference Learning Using Large Language Models
Designing reward functions is a core component of reinforcement learning but can be challenging for truly complex behavior. Reinforcement Learning from Human Feedback (RLHF) has been used to alleviate this challenge by replacing a hand-coded reward function with a reward function learned from preferences. However, it can be exceedingly inefficient to learn these rewards as they are often learned tabula rasa. We investigate whether Large Language Models (LLMs) can reduce this query inefficiency by converting an iterative series of human preferences into code representing the rewards. We propose In-Context Preference Learning (ICPL), a method that uses the grounding of an LLM to accelerate learning reward functions from preferences. ICPL takes the environment context and task description, synthesizes a set of reward functions, and then repeatedly updates the reward functions using human rankings of videos of the resultant policies. Using synthetic preferences, we demonstrate that ICPL is orders of magnitude more efficient than RLHF and is even competitive with methods that use ground-truth reward functions instead of preferences. Finally, we perform a series of human preference-learning trials and observe that ICPL extends beyond synthetic settings and can work effectively with humans-in-the-loop. Additional information and videos are provided at https://sites.google.com/view/few-shot-icpl/home.
☆ Optimal Robust Estimation under Local and Global Corruptions: Stronger Adversary and Smaller Error
Algorithmic robust statistics has traditionally focused on the contamination model where a small fraction of the samples are arbitrarily corrupted. We consider a recent contamination model that combines two kinds of corruptions: (i) small fraction of arbitrary outliers, as in classical robust statistics, and (ii) local perturbations, where samples may undergo bounded shifts on average. While each noise model is well understood individually, the combined contamination model poses new algorithmic challenges, with only partial results known. Existing efficient algorithms are limited in two ways: (i) they work only for a weak notion of local perturbations, and (ii) they obtain suboptimal error for isotropic subgaussian distributions (among others). The latter limitation led [NGS24, COLT'24] to hypothesize that improving the error might, in fact, be computationally hard. Perhaps surprisingly, we show that information theoretically optimal error can indeed be achieved in polynomial time, under an even \emph{stronger} local perturbation model (the sliced-Wasserstein metric as opposed to the Wasserstein metric). Notably, our analysis reveals that the entire family of stability-based robust mean estimators continues to work optimally in a black-box manner for the combined contamination model. This generalization is particularly useful in real-world scenarios where the specific form of data corruption is not known in advance. We also present efficient algorithms for distribution learning and principal component analysis in the combined contamination model.
☆ Dhoroni: Exploring Bengali Climate Change and Environmental Views with a Multi-Perspective News Dataset and Natural Language Processing
Climate change poses critical challenges globally, disproportionately affecting low-income countries that often lack resources and linguistic representation on the international stage. Despite Bangladesh's status as one of the most vulnerable nations to climate impacts, research gaps persist in Bengali-language studies related to climate change and NLP. To address this disparity, we introduce Dhoroni, a novel Bengali (Bangla) climate change and environmental news dataset, comprising a 2300 annotated Bangla news articles, offering multiple perspectives such as political influence, scientific/statistical data, authenticity, stance detection, and stakeholder involvement. Furthermore, we present an in-depth exploratory analysis of Dhoroni and introduce BanglaBERT-Dhoroni family, a novel baseline model family for climate and environmental opinion detection in Bangla, fine-tuned on our dataset. This research contributes significantly to enhancing accessibility and analysis of climate discourse in Bengali (Bangla), addressing crucial communication and research gaps in climate-impacted regions like Bangladesh with 180 million people.
comment: In Review
☆ Scalable spectral representations for network multiagent control
Network Markov Decision Processes (MDPs), a popular model for multi-agent control, pose a significant challenge to efficient learning due to the exponential growth of the global state-action space with the number of agents. In this work, utilizing the exponential decay property of network dynamics, we first derive scalable spectral local representations for network MDPs, which induces a network linear subspace for the local $Q$-function of each agent. Building on these local spectral representations, we design a scalable algorithmic framework for continuous state-action network MDPs, and provide end-to-end guarantees for the convergence of our algorithm. Empirically, we validate the effectiveness of our scalable representation-based approach on two benchmark problems, and demonstrate the advantages of our approach over generic function approximation approaches to representing the local $Q$-functions.
☆ Hierarchical Upper Confidence Bounds for Constrained Online Learning
The multi-armed bandit (MAB) problem is a foundational framework in sequential decision-making under uncertainty, extensively studied for its applications in areas such as clinical trials, online advertising, and resource allocation. Traditional MAB formulations, however, do not adequately capture scenarios where decisions are structured hierarchically, involve multi-level constraints, or feature context-dependent action spaces. In this paper, we introduce the hierarchical constrained bandits (HCB) framework, which extends the contextual bandit problem to incorporate hierarchical decision structures and multi-level constraints. We propose the hierarchical constrained upper confidence bound (HC-UCB) algorithm, designed to address the complexities of the HCB problem by leveraging confidence bounds within a hierarchical setting. Our theoretical analysis establishes sublinear regret bounds for HC-UCB and provides high-probability guarantees for constraint satisfaction at all hierarchical levels. Furthermore, we derive a minimax lower bound on the regret for the HCB problem, demonstrating the near-optimality of our algorithm. The results are significant for real-world applications where decision-making processes are inherently hierarchical and constrained, offering a robust and efficient solution that balances exploration and exploitation across multiple levels of decision-making.
☆ Neuroevolution Neural Architecture Search for Evolving RNNs in Stock Return Prediction and Portfolio Trading
Stock return forecasting is a major component of numerous finance applications. Predicted stock returns can be incorporated into portfolio trading algorithms to make informed buy or sell decisions which can optimize returns. In such portfolio trading applications, the predictive performance of a time series forecasting model is crucial. In this work, we propose the use of the Evolutionary eXploration of Augmenting Memory Models (EXAMM) algorithm to progressively evolve recurrent neural networks (RNNs) for stock return predictions. RNNs are evolved independently for each stocks and portfolio trading decisions are made based on the predicted stock returns. The portfolio used for testing consists of the 30 companies in the Dow-Jones Index (DJI) with each stock have the same weight. Results show that using these evolved RNNs and a simple daily long-short strategy can generate higher returns than both the DJI index and the S&P 500 Index for both 2022 (bear market) and 2023 (bull market).
☆ Audio-to-Score Conversion Model Based on Whisper methodology
This thesis develops a Transformer model based on Whisper, which extracts melodies and chords from music audio and records them into ABC notation. A comprehensive data processing workflow is customized for ABC notation, including data cleansing, formatting, and conversion, and a mutation mechanism is implemented to increase the diversity and quality of training data. This thesis innovatively introduces the "Orpheus' Score", a custom notation system that converts music information into tokens, designs a custom vocabulary library, and trains a corresponding custom tokenizer. Experiments show that compared to traditional algorithms, the model has significantly improved accuracy and performance. While providing a convenient audio-to-score tool for music enthusiasts, this work also provides new ideas and tools for research in music information processing.
comment: 5 pages, 7 figures
☆ Representation Shattering in Transformers: A Synthetic Study with Knowledge Editing
Knowledge Editing (KE) algorithms alter models' internal weights to perform targeted updates to incorrect, outdated, or otherwise unwanted factual associations. In order to better define the possibilities and limitations of these approaches, recent work has shown that applying KE can adversely affect models' factual recall accuracy and diminish their general reasoning abilities. While these studies give broad insights into the potential harms of KE algorithms, e.g., via performance evaluations on benchmarks, we argue little is understood as to why such destructive failures occur. Is it possible KE methods distort representations of concepts beyond the targeted fact, hence hampering abilities at broad? If so, what is the extent of this distortion? To take a step towards addressing such questions, we define a novel synthetic task wherein a Transformer is trained from scratch to internalize a ``structured'' knowledge graph. The structure enforces relationships between entities of the graph, such that editing a factual association has "trickling effects" on other entities in the graph (e.g., altering X's parent is Y to Z affects who X's siblings' parent is). Through evaluations of edited models and analysis of extracted representations, we show that KE inadvertently affects representations of entities beyond the targeted one, distorting relevant structures that allow a model to infer unseen knowledge about an entity. We call this phenomenon representation shattering and demonstrate that it results in degradation of factual recall and reasoning performance more broadly. To corroborate our findings in a more naturalistic setup, we perform preliminary experiments with a pretrained GPT-2-XL model and reproduce the representation shattering effect therein as well. Overall, our work yields a precise mechanistic hypothesis to explain why KE has adverse effects on model capabilities.
comment: Under review
☆ On Functional Dimension and Persistent Pseudodimension
For any fixed feedforward ReLU neural network architecture, it is well-known that many different parameter settings can determine the same function. It is less well-known that the degree of this redundancy is inhomogeneous across parameter space. In this work, we discuss two locally applicable complexity measures for ReLU network classes and what we know about the relationship between them: (1) the local functional dimension [14, 18], and (2) a local version of VC dimension that we call persistent pseudodimension. The former is easy to compute on finite batches of points; the latter should give local bounds on the generalization gap, which would inform an understanding of the mechanics of the double descent phenomenon [7].
comment: 41 pages
☆ Remote Timing Attacks on Efficient Language Model Inference
Scaling up language models has significantly increased their capabilities. But larger models are slower models, and so there is now an extensive body of work (e.g., speculative sampling or parallel decoding) that improves the (average case) efficiency of language model generation. But these techniques introduce data-dependent timing characteristics. We show it is possible to exploit these timing differences to mount a timing attack. By monitoring the (encrypted) network traffic between a victim user and a remote language model, we can learn information about the content of messages by noting when responses are faster or slower. With complete black-box access, on open source systems we show how it is possible to learn the topic of a user's conversation (e.g., medical advice vs. coding assistance) with 90%+ precision, and on production systems like OpenAI's ChatGPT and Anthropic's Claude we can distinguish between specific messages or infer the user's language. We further show that an active adversary can leverage a boosting attack to recover PII placed in messages (e.g., phone numbers or credit card numbers) for open source systems. We conclude with potential defenses and directions for future work.
☆ Interchangeable Token Embeddings for Extendable Vocabulary and Alpha-Equivalence
We propose a novel approach for learning interchangeable tokens in language models to obtain an extendable vocabulary that can generalize to new tokens. Our method is designed to address alpha-equivalence, the principle that renaming bound variables in a syntactic expression preserves semantics. This property arises in many formal languages such as temporal logics, in which all proposition symbols represent the same concept but are distinguishable from each other. To handle such tokens, we develop a dual-part embedding approach. The first part is shared across all interchangeable tokens, thereby enforcing that they represent the same core concept. The second part is randomly generated for each token, which enables distinguishability. We evaluate our method in a Transformer encoder-decoder model on two tasks: solving linear temporal logic formulae and copying with extendable vocabulary. Our method demonstrates promising generalization capabilities in addition to introducing a favorable inductive bias for alpha-equivalence.
comment: 14 pages, 5 figures
☆ LiNo: Advancing Recursive Residual Decomposition of Linear and Nonlinear Patterns for Robust Time Series Forecasting
Forecasting models are pivotal in a data-driven world with vast volumes of time series data that appear as a compound of vast Linear and Nonlinear patterns. Recent deep time series forecasting models struggle to utilize seasonal and trend decomposition to separate the entangled components. Such a strategy only explicitly extracts simple linear patterns like trends, leaving the other linear modes and vast unexplored nonlinear patterns to the residual. Their flawed linear and nonlinear feature extraction models and shallow-level decomposition limit their adaptation to the diverse patterns present in real-world scenarios. Given this, we innovate Recursive Residual Decomposition by introducing explicit extraction of both linear and nonlinear patterns. This deeper-level decomposition framework, which is named LiNo, captures linear patterns using a Li block which can be a moving average kernel, and models nonlinear patterns using a No block which can be a Transformer encoder. The extraction of these two patterns is performed alternatively and recursively. To achieve the full potential of LiNo, we develop the current simple linear pattern extractor to a general learnable autoregressive model, and design a novel No block that can handle all essential nonlinear patterns. Remarkably, the proposed LiNo achieves state-of-the-art on thirteen real-world benchmarks under univariate and multivariate forecasting scenarios. Experiments show that current forecasting models can deliver more robust and precise results through this advanced Recursive Residual Decomposition. We hope this work could offer insight into designing more effective forecasting models. Code is available at this Repository: https://github.com/Levi-Ackman/LiNo.
☆ Covariance estimation using Markov chain Monte Carlo
We investigate the complexity of covariance matrix estimation for Gibbs distributions based on dependent samples from a Markov chain. We show that when $\pi$ satisfies a Poincar\'e inequality and the chain possesses a spectral gap, we can achieve similar sample complexity using MCMC as compared to an estimator constructed using i.i.d. samples, with potentially much better query complexity. As an application of our methods, we show improvements for the query complexity in both constrained and unconstrained settings for concrete instances of MCMC. In particular, we provide guarantees regarding isotropic rounding procedures for sampling uniformly on convex bodies.
comment: 30 pages
☆ LiNeS: Post-training Layer Scaling Prevents Forgetting and Enhances Model Merging
Large pre-trained models exhibit impressive zero-shot performance across diverse tasks, but fine-tuning often leads to catastrophic forgetting, where improvements on a target domain degrade generalization on other tasks. To address this challenge, we introduce LiNeS, Layer-increasing Network Scaling, a post-training editing technique designed to preserve pre-trained generalization while enhancing fine-tuned task performance. LiNeS scales parameter updates linearly based on their layer depth within the network, maintaining shallow layers close to their pre-trained values to preserve general features while allowing deeper layers to retain task-specific representations. We further extend this approach to multi-task model merging scenarios, where layer-wise scaling of merged parameters reduces negative task interference. LiNeS demonstrates significant improvements in both single-task and multi-task settings across various benchmarks in vision and natural language processing. It mitigates forgetting, enhances out-of-distribution generalization, integrates seamlessly with existing multi-task model merging baselines improving their performance across benchmarks and model sizes, and can boost generalization when merging LLM policies aligned with different rewards via RLHF. Importantly, our method is simple to implement and complementary to many existing techniques.
comment: The first two authors contributed equally to this work; Project website: \url{https://lines-merging.github.io/}
☆ Can General-Purpose Large Language Models Generalize to English-Thai Machine Translation ? EMNLP 2024
Large language models (LLMs) perform well on common tasks but struggle with generalization in low-resource and low-computation settings. We examine this limitation by testing various LLMs and specialized translation models on English-Thai machine translation and code-switching datasets. Our findings reveal that under more strict computational constraints, such as 4-bit quantization, LLMs fail to translate effectively. In contrast, specialized models, with comparable or lower computational requirements, consistently outperform LLMs. This underscores the importance of specialized models for maintaining performance under resource constraints.
comment: Accepted in GenBench EMNLP 2024
☆ Coniferest: a complete active anomaly detection framework
We present coniferest, an open source generic purpose active anomaly detection framework written in Python. The package design and implemented algorithms are described. Currently, static outlier detection analysis is supported via the Isolation forest algorithm. Moreover, Active Anomaly Discovery (AAD) and Pineforest algorithms are available to tackle active anomaly detection problems. The algorithms and package performance are evaluated on a series of synthetic datasets. We also describe a few success cases which resulted from applying the package to real astronomical data in active anomaly detection tasks within the SNAD project.
comment: 13 pages, 1 figure
☆ Reinforcement Learning for Data-Driven Workflows in Radio Interferometry. I. Principal Demonstration in Calibration
Radio interferometry is an observational technique used to study astrophysical phenomena. Data gathered by an interferometer requires substantial processing before astronomers can extract the scientific information from it. Data processing consists of a sequence of calibration and analysis procedures where choices must be made about the sequence of procedures as well as the specific configuration of the procedure itself. These choices are typically based on a combination of measurable data characteristics, an understanding of the instrument itself, an appreciation of the trade-offs between compute cost and accuracy, and a learned understanding of what is considered "best practice". A metric of absolute correctness is not always available and validity is often subject to human judgment. The underlying principles and software configurations to discern a reasonable workflow for a given dataset is the subject of training workshops for students and scientists. Our goal is to use objective metrics that quantify best practice, and numerically map out the decision space with respect to our metrics. With these objective metrics we demonstrate an automated, data-driven, decision system that is capable of sequencing the optimal action(s) for processing interferometric data. This paper introduces a simplified description of the principles behind interferometry and the procedures required for data processing. We highlight the issues with current automation approaches and propose our ideas for solving these bottlenecks. A prototype is demonstrated and the results are discussed.
comment: 22 pages, 13 figures; accepted for publication in The Astronomical Journal October 18, 2024
☆ Understanding Transfer Learning via Mean-field Analysis
We propose a novel framework for exploring generalization errors of transfer learning through the lens of differential calculus on the space of probability measures. In particular, we consider two main transfer learning scenarios, $\alpha$-ERM and fine-tuning with the KL-regularized empirical risk minimization and establish generic conditions under which the generalization error and the population risk convergence rates for these scenarios are studied. Based on our theoretical results, we show the benefits of transfer learning with a one-hidden-layer neural network in the mean-field regime under some suitable integrability and regularity assumptions on the loss and activation functions.
comment: Under review
☆ Exploring RL-based LLM Training for Formal Language Tasks with Programmed Rewards
Proximal Policy Optimization (PPO) is commonly used in Reinforcement Learning from Human Feedback to align large language models (LLMs) with downstream tasks. This paper investigates the feasibility of using PPO for direct reinforcement learning (RL) from explicitly programmed reward signals, as opposed to indirect learning from human feedback via an intermediary reward model. We focus on tasks expressed through formal languages, such as mathematics and programming, where explicit reward functions can be programmed to automatically assess the quality of generated outputs. We apply this approach to a sentiment alignment task, a simple arithmetic task, and a more complex game synthesis task. The sentiment alignment task replicates prior research and serves to validate our experimental setup. Our results show that pure RL-based training for the two formal language tasks is challenging, with success being limited even for the simple arithmetic task. We propose a novel batch-entropy regularization term to aid exploration, although training is not yet entirely stable. Our findings suggest that direct RL training of LLMs may be more suitable for relatively minor changes, such as alignment, than for learning new tasks altogether, even if an informative reward signal can be expressed programmatically.
comment: Accepted at BNAIC 2024
☆ Learning Load Balancing with GNN in MPTCP-Enabled Heterogeneous Networks
Hybrid light fidelity (LiFi) and wireless fidelity (WiFi) networks are a promising paradigm of heterogeneous network (HetNet), attributed to the complementary physical properties of optical spectra and radio frequency. However, the current development of such HetNets is mostly bottlenecked by the existing transmission control protocol (TCP), which restricts the user equipment (UE) to connecting one access point (AP) at a time. While the ongoing investigation on multipath TCP (MPTCP) can bring significant benefits, it complicates the network topology of HetNets, making the existing load balancing (LB) learning models less effective. Driven by this, we propose a graph neural network (GNN)-based model to tackle the LB problem for MPTCP-enabled HetNets, which results in a partial mesh topology. Such a topology can be modeled as a graph, with the channel state information and data rate requirement embedded as node features, while the LB solutions are deemed as edge labels. Compared to the conventional deep neural network (DNN), the proposed GNN-based model exhibits two key strengths: i) it can better interpret a complex network topology; and ii) it can handle various numbers of APs and UEs with a single trained model. Simulation results show that against the traditional optimisation method, the proposed learning model can achieve near-optimal throughput within a gap of 11.5%, while reducing the inference time by 4 orders of magnitude. In contrast to the DNN model, the new method can improve the network throughput by up to 21.7%, at a similar inference time level.
☆ Permutation Picture of Graph Combinatorial Optimization Problems
This paper proposes a framework that formulates a wide range of graph combinatorial optimization problems using permutation-based representations. These problems include the travelling salesman problem, maximum independent set, maximum cut, and various other related problems. This work potentially opens up new avenues for algorithm design in neural combinatorial optimization, bridging the gap between discrete and continuous optimization techniques.
comment: 15 pages, 2 figures
☆ Human-LLM Hybrid Text Answer Aggregation for Crowd Annotations EMNLP 2024
The quality is a crucial issue for crowd annotations. Answer aggregation is an important type of solution. The aggregated answers estimated from multiple crowd answers to the same instance are the eventually collected annotations, rather than the individual crowd answers themselves. Recently, the capability of Large Language Models (LLMs) on data annotation tasks has attracted interest from researchers. Most of the existing studies mainly focus on the average performance of individual crowd workers; several recent works studied the scenarios of aggregation on categorical labels and LLMs used as label creators. However, the scenario of aggregation on text answers and the role of LLMs as aggregators are not yet well-studied. In this paper, we investigate the capability of LLMs as aggregators in the scenario of close-ended crowd text answer aggregation. We propose a human-LLM hybrid text answer aggregation method with a Creator-Aggregator Multi-Stage (CAMS) crowdsourcing framework. We make the experiments based on public crowdsourcing datasets. The results show the effectiveness of our approach based on the collaboration of crowd workers and LLMs.
comment: Accepted in EMNLP 2024
☆ Exploration and Persuasion
How to incentivize self-interested agents to explore when they prefer to exploit? Consider a population of self-interested agents that make decisions under uncertainty. They "explore" to acquire new information and "exploit" this information to make good decisions. Collectively they need to balance these two objectives, but their incentives are skewed toward exploitation. This is because exploration is costly, but its benefits are spread over many agents in the future. "Incentivized Exploration" addresses this issue via strategic communication. Consider a benign ``principal" which can communicate with the agents and make recommendations, but cannot force the agents to comply. Moreover, suppose the principal can observe the agents' decisions and the outcomes of these decisions. The goal is to design a communication and recommendation policy which (i) achieves a desirable balance between exploration and exploitation, and (ii) incentivizes the agents to follow recommendations. What makes it feasible is "information asymmetry": the principal knows more than any one agent, as it collects information from many. It is essential that the principal does not fully reveal all its knowledge to the agents. Incentivized exploration combines two important problems in, resp., machine learning and theoretical economics. First, if agents always follow recommendations, the principal faces a multi-armed bandit problem: essentially, design an algorithm that balances exploration and exploitation. Second, interaction with a single agent corresponds to "Bayesian persuasion", where a principal leverages information asymmetry to convince an agent to take a particular action. We provide a brief but self-contained introduction to each problem through the lens of incentivized exploration, solving a key special case of the former as a sub-problem of the latter.
comment: This is a chapter published in "Online and Matching-Based Markets", Cambridge University Press, 2023. It has been available from the author's website since 2021
☆ Combinatorial Logistic Bandits
We introduce a novel framework called combinatorial logistic bandits (CLogB), where in each round, a subset of base arms (called the super arm) is selected, with the outcome of each base arm being binary and its expectation following a logistic parametric model. The feedback is governed by a general arm triggering process. Our study covers CLogB with reward functions satisfying two smoothness conditions, capturing application scenarios such as online content delivery, online learning to rank, and dynamic channel allocation. We first propose a simple yet efficient algorithm, CLogUCB, utilizing a variance-agnostic exploration bonus. Under the 1-norm triggering probability modulated (TPM) smoothness condition, CLogUCB achieves a regret bound of $\tilde{O}(d\sqrt{\kappa KT})$, where $\tilde{O}$ ignores logarithmic factors, $d$ is the dimension of the feature vector, $\kappa$ represents the nonlinearity of the logistic model, and $K$ is the maximum number of base arms a super arm can trigger. This result improves on prior work by a factor of $\tilde{O}(\sqrt{\kappa})$. We then enhance CLogUCB with a variance-adaptive version, VA-CLogUCB, which attains a regret bound of $\tilde{O}(d\sqrt{KT})$ under the same 1-norm TPM condition, improving another $\tilde{O}(\sqrt{\kappa})$ factor. VA-CLogUCB shows even greater promise under the stronger triggering probability and variance modulated (TPVM) condition, achieving a leading $\tilde{O}(d\sqrt{T})$ regret, thus removing the additional dependency on the action-size $K$. Furthermore, we enhance the computational efficiency of VA-CLogUCB by eliminating the nonconvex optimization process when the context feature map is time-invariant while maintaining the tight $\tilde{O}(d\sqrt{T})$ regret. Finally, experiments on synthetic and real-world datasets demonstrate the superior performance of our algorithms compared to benchmark algorithms.
comment: Accepted to ACM SIGMETRICS 2025
☆ Neuronal Competition Groups with Supervised STDP for Spike-Based Classification
Spike Timing-Dependent Plasticity (STDP) is a promising substitute to backpropagation for local training of Spiking Neural Networks (SNNs) on neuromorphic hardware. STDP allows SNNs to address classification tasks by combining unsupervised STDP for feature extraction and supervised STDP for classification. Unsupervised STDP is usually employed with Winner-Takes-All (WTA) competition to learn distinct patterns. However, WTA for supervised STDP classification faces unbalanced competition challenges. In this paper, we propose a method to effectively implement WTA competition in a spiking classification layer employing first-spike coding and supervised STDP training. We introduce the Neuronal Competition Group (NCG), an architecture that improves classification capabilities by promoting the learning of various patterns per class. An NCG is a group of neurons mapped to a specific class, implementing intra-class WTA and a novel competition regulation mechanism based on two-compartment thresholds. We incorporate our proposed architecture into spiking classification layers trained with state-of-the-art supervised STDP rules. On top of two different unsupervised feature extractors, we obtain significant accuracy improvements on image recognition datasets such as CIFAR-10 and CIFAR-100. We show that our competition regulation mechanism is crucial for ensuring balanced competition and improved class separation.
☆ Optimal Design for Reward Modeling in RLHF
Reinforcement Learning from Human Feedback (RLHF) has become a popular approach to align language models (LMs) with human preferences. This method involves collecting a large dataset of human pairwise preferences across various text generations and using it to infer (implicitly or explicitly) a reward model. Numerous methods have been proposed to learn the reward model and align a LM with it. However, the costly process of collecting human preferences has received little attention and could benefit from theoretical insights. This paper addresses this issue and aims to formalize the reward training model in RLHF. We frame the selection of an effective dataset as a simple regret minimization task, using a linear contextual dueling bandit method. Given the potentially large number of arms, this approach is more coherent than the best-arm identification setting. We then propose an offline framework for solving this problem. Under appropriate assumptions - linearity of the reward model in the embedding space, and boundedness of the reward parameter - we derive bounds on the simple regret. Finally, we provide a lower bound that matches our upper bound up to constant and logarithmic terms. To our knowledge, this is the first theoretical contribution in this area to provide an offline approach as well as worst-case guarantees.
☆ UnStar: Unlearning with Self-Taught Anti-Sample Reasoning for LLMs
The key components of machine learning are data samples for training, model for learning patterns, and loss function for optimizing accuracy. Analogously, unlearning can potentially be achieved through anti-data samples (or anti-samples), unlearning method, and reversed loss function. While prior research has explored unlearning methods and reversed loss functions, the potential of anti-samples remains largely untapped. In this paper, we introduce UnSTAR: Unlearning with Self-Taught Anti-Sample Reasoning for large language models (LLMs). Our contributions are threefold; first, we propose a novel concept of anti-sample-induced unlearning; second, we generate anti-samples by leveraging misleading rationales, which help reverse learned associations and accelerate the unlearning process; and third, we enable fine-grained targeted unlearning, allowing for the selective removal of specific associations without impacting related knowledge - something not achievable by previous works. Results demonstrate that anti-samples offer an efficient, targeted unlearning strategy for LLMs, opening new avenues for privacy-preserving machine learning and model modification.
☆ A Comparison of Baseline Models and a Transformer Network for SOC Prediction in Lithium-Ion Batteries
Accurately predicting the state of charge of Lithium-ion batteries is essential to the performance of battery management systems of electric vehicles. One of the main reasons for the slow global adoption of electric cars is driving range anxiety. The ability of a battery management system to accurately estimate the state of charge can help alleviate this problem. In this paper, a comparison between data-driven state-of-charge estimation methods is conducted. The paper compares different neural network-based models and common regression models for SOC estimation. These models include several ablated transformer networks, a neural network, a lasso regression model, a linear regression model and a decision tree. Results of various experiments conducted on data obtained from natural driving cycles of the BMW i3 battery show that the decision tree outperformed all other models including the more complex transformer network with self-attention and positional encoding.
☆ Optimizing Mixture-of-Experts Inference Time Combining Model Deployment and Communication Scheduling
As machine learning models scale in size and complexity, their computational requirements become a significant barrier. Mixture-of-Experts (MoE) models alleviate this issue by selectively activating relevant experts. Despite this, MoE models are hindered by high communication overhead from all-to-all operations, low GPU utilization due to the synchronous communication constraint, and complications from heterogeneous GPU environments. This paper presents Aurora, which optimizes both model deployment and all-to-all communication scheduling to address these challenges in MoE inference. Aurora achieves minimal communication times by strategically ordering token transmissions in all-to-all communications. It improves GPU utilization by colocating experts from different models on the same device, avoiding the limitations of synchronous all-to-all communication. We analyze Aurora's optimization strategies theoretically across four common GPU cluster settings: exclusive vs. colocated models on GPUs, and homogeneous vs. heterogeneous GPUs. Aurora provides optimal solutions for three cases, and for the remaining NP-hard scenario, it offers a polynomial-time sub-optimal solution with only a 1.07x degradation from the optimal. Aurora is the first approach to minimize MoE inference time via optimal model deployment and communication scheduling across various scenarios. Evaluations demonstrate that Aurora significantly accelerates inference, achieving speedups of up to 2.38x in homogeneous clusters and 3.54x in heterogeneous environments. Moreover, Aurora enhances GPU utilization by up to 1.5x compared to existing methods.
☆ Deep Memory Search: A Metaheuristic Approach for Optimizing Heuristic Search
Metaheuristic search methods have proven to be essential tools for tackling complex optimization challenges, but their full potential is often constrained by conventional algorithmic frameworks. In this paper, we introduce a novel approach called Deep Heuristic Search (DHS), which models metaheuristic search as a memory-driven process. DHS employs multiple search layers and memory-based exploration-exploitation mechanisms to navigate large, dynamic search spaces. By utilizing model-free memory representations, DHS enhances the ability to traverse temporal trajectories without relying on probabilistic transition models. The proposed method demonstrates significant improvements in search efficiency and performance across a range of heuristic optimization problems.
comment: 10 pages, 6 figures
☆ Can a Machine Distinguish High and Low Amount of Social Creak in Speech?
Objectives: ncreased prevalence of social creak particularly among female speakers has been reported in several studies. The study of social creak has been previously conducted by combining perceptual evaluation of speech with conventional acoustical parameters such as the harmonic-to-noise ratio and cepstral peak prominence. In the current study, machine learning (ML) was used to automatically distinguish speech of low amount of social creak from speech of high amount of social creak. Methods: The amount of creak in continuous speech samples produced in Finnish by 90 female speakers was first perceptually assessed by two voice specialists. Based on their assessments, the speech samples were divided into two categories (low $vs$. high amount of creak). Using the speech signals and their creak labels, seven different ML models were trained. Three spectral representations were used as feature for each model. Results: The results show that the best performance (accuracy of 71.1\%) was obtained by the following two systems: an Adaboost classifier using the mel-spectrogram feature and a decision tree classifier using the mel-frequency cepstral coefficient feature. Conclusions: The study of social creak is becoming increasingly popular in sociolinguistic and vocological research. The conventional human perceptual assessment of the amount of creak is laborious and therefore ML technology could be used to assist researchers studying social creak. The classification systems reported in this study could be considered as baselines in future ML-based studies on social creak.
comment: Accepted in Journal of Voice
☆ LFME: A Simple Framework for Learning from Multiple Experts in Domain Generalization NeurIPS 2024
Domain generalization (DG) methods aim to maintain good performance in an unseen target domain by using training data from multiple source domains. While success on certain occasions are observed, enhancing the baseline across most scenarios remains challenging. This work introduces a simple yet effective framework, dubbed learning from multiple experts (LFME), that aims to make the target model an expert in all source domains to improve DG. Specifically, besides learning the target model used in inference, LFME will also train multiple experts specialized in different domains, whose output probabilities provide professional guidance by simply regularizing the logit of the target model. Delving deep into the framework, we reveal that the introduced logit regularization term implicitly provides effects of enabling the target model to harness more information, and mining hard samples from the experts during training. Extensive experiments on benchmarks from different DG tasks demonstrate that LFME is consistently beneficial to the baseline and can achieve comparable performance to existing arts. Code is available at~\url{https://github.com/liangchen527/LFME}.
comment: Accepted by NeurIPS 2024
☆ Sample-Efficient Geometry Reconstruction from Euclidean Distances using Non-Convex Optimization
The problem of finding suitable point embedding or geometric configurations given only Euclidean distance information of point pairs arises both as a core task and as a sub-problem in a variety of machine learning applications. In this paper, we aim to solve this problem given a minimal number of distance samples. To this end, we leverage continuous and non-convex rank minimization formulations of the problem and establish a local convergence guarantee for a variant of iteratively reweighted least squares (IRLS), which applies if a minimal random set of observed distances is provided. As a technical tool, we establish a restricted isometry property (RIP) restricted to a tangent space of the manifold of symmetric rank-$r$ matrices given random Euclidean distance measurements, which might be of independent interest for the analysis of other non-convex approaches. Furthermore, we assess data efficiency, scalability and generalizability of different reconstruction algorithms through numerical experiments with simulated data as well as real-world data, demonstrating the proposed algorithm's ability to identify the underlying geometry from fewer distance samples compared to the state-of-the-art.
☆ Publishing Neural Networks in Drug Discovery Might Compromise Training Data Privacy
This study investigates the risks of exposing confidential chemical structures when machine learning models trained on these structures are made publicly available. We use membership inference attacks, a common method to assess privacy that is largely unexplored in the context of drug discovery, to examine neural networks for molecular property prediction in a black-box setting. Our results reveal significant privacy risks across all evaluated datasets and neural network architectures. Combining multiple attacks increases these risks. Molecules from minority classes, often the most valuable in drug discovery, are particularly vulnerable. We also found that representing molecules as graphs and using message-passing neural networks may mitigate these risks. We provide a framework to assess privacy risks of classification models and molecular representations. Our findings highlight the need for careful consideration when sharing neural networks trained on proprietary chemical structures, informing organisations and researchers about the trade-offs between data confidentiality and model openness.
☆ Learning Mathematical Rules with Large Language Models NeurIPS'24
In this paper, we study the ability of large language models to learn specific mathematical rules such as distributivity or simplifying equations. We present an empirical analysis of their ability to generalize these rules, as well as to reuse them in the context of word problems. For this purpose, we provide a rigorous methodology to build synthetic data incorporating such rules, and perform fine-tuning of large language models on such data. Our experiments show that our model can learn and generalize these rules to some extent, as well as suitably reuse them in the context of word problems.
comment: 4th MATH-AI Workshop at NeurIPS'24
☆ Sample-efficient Bayesian Optimisation Using Known Invariances NeurIPS 2024
Bayesian optimisation (BO) is a powerful framework for global optimisation of costly functions, using predictions from Gaussian process models (GPs). In this work, we apply BO to functions that exhibit invariance to a known group of transformations. We show that vanilla and constrained BO algorithms are inefficient when optimising such invariant objectives, and provide a method for incorporating group invariances into the kernel of the GP to produce invariance-aware algorithms that achieve significant improvements in sample efficiency. We derive a bound on the maximum information gain of these invariant kernels, and provide novel upper and lower bounds on the number of observations required for invariance-aware BO algorithms to achieve $\epsilon$-optimality. We demonstrate our method's improved performance on a range of synthetic invariant and quasi-invariant functions. We also apply our method in the case where only some of the invariance is incorporated into the kernel, and find that these kernels achieve similar gains in sample efficiency at significantly reduced computational cost. Finally, we use invariant BO to design a current drive system for a nuclear fusion reactor, finding a high-performance solution where non-invariant methods failed.
comment: Accepted as a poster at NeurIPS 2024
☆ ISImed: A Framework for Self-Supervised Learning using Intrinsic Spatial Information in Medical Images
This paper demonstrates that spatial information can be used to learn interpretable representations in medical images using Self-Supervised Learning (SSL). Our proposed method, ISImed, is based on the observation that medical images exhibit a much lower variability among different images compared to classic data vision benchmarks. By leveraging this resemblance of human body structures across multiple images, we establish a self-supervised objective that creates a latent representation capable of capturing its location in the physical realm. More specifically, our method involves sampling image crops and creating a distance matrix that compares the learned representation vectors of all possible combinations of these crops to the true distance between them. The intuition is, that the learned latent space is a positional encoding for a given image crop. We hypothesize, that by learning these positional encodings, comprehensive image representations have to be generated. To test this hypothesis and evaluate our method, we compare our learned representation with two state-of-the-art SSL benchmarking methods on two publicly available medical imaging datasets. We show that our method can efficiently learn representations that capture the underlying structure of the data and can be used to transfer to a downstream classification task.
comment: 11 pages, 4 figures
☆ Business Process Simulation: Probabilistic Modeling of Intermittent Resource Availability and Multitasking Behavior
In business process simulation, resource availability is typically modeled by assigning a calendar to each resource, e.g., Monday-Friday, 9:00-18:00. Resources are assumed to be always available during each time slot in their availability calendar. This assumption often becomes invalid due to interruptions, breaks, or time-sharing across processes. In other words, existing approaches fail to capture intermittent availability. Another limitation of existing approaches is that they either do not consider multitasking behavior, or if they do, they assume that resources always multitask (up to a maximum capacity) whenever available. However, studies have shown that the multitasking patterns vary across days. This paper introduces a probabilistic approach to model resource availability and multitasking behavior for business process simulation. In this approach, each time slot in a resource calendar has an associated availability probability and a multitasking probability per multitasking level. For example, a resource may be available on Fridays between 14:00-15:00 with 90\% probability, and given that they are performing one task during this slot, they may take on a second concurrent task with 60\% probability. We propose algorithms to discover probabilistic calendars and probabilistic multitasking capacities from event logs. An evaluation shows that, with these enhancements, simulation models discovered from event logs better replicate the distribution of activities and cycle times, relative to approaches with crisp calendars and monotasking assumptions.
☆ Graph Neural Networks for Edge Signals: Orientation Equivariance and Invariance
Many applications in traffic, civil engineering, or electrical engineering revolve around edge-level signals. Such signals can be categorized as inherently directed, for example, the water flow in a pipe network, and undirected, like the diameter of a pipe. Topological methods model edge signals with inherent direction by representing them relative to a so-called orientation assigned to each edge. These approaches can neither model undirected edge signals nor distinguish if an edge itself is directed or undirected. We address these shortcomings by (i) revising the notion of orientation equivariance to enable edge direction-aware topological models, (ii) proposing orientation invariance as an additional requirement to describe signals without inherent direction, and (iii) developing EIGN, an architecture composed of novel direction-aware edge-level graph shift operators, that provably fulfills the aforementioned desiderata. It is the first general-purpose topological GNN for edge-level signals that can model directed and undirected signals while distinguishing between directed and undirected edges. A comprehensive evaluation shows that EIGN outperforms prior work in edge-level tasks, for example, improving in RMSE on flow simulation tasks by up to 43.5%.
☆ xLSTM-Mixer: Multivariate Time Series Forecasting by Mixing via Scalar Memories
Time series data is prevalent across numerous fields, necessitating the development of robust and accurate forecasting models. Capturing patterns both within and between temporal and multivariate components is crucial for reliable predictions. We introduce xLSTM-Mixer, a model designed to effectively integrate temporal sequences, joint time-variate information, and multiple perspectives for robust forecasting. Our approach begins with a linear forecast shared across variates, which is then refined by xLSTM blocks. These blocks serve as key elements for modeling the complex dynamics of challenging time series data. xLSTM-Mixer ultimately reconciles two distinct views to produce the final forecast. Our extensive evaluations demonstrate xLSTM-Mixer's superior long-term forecasting performance compared to recent state-of-the-art methods. A thorough model analysis provides further insights into its key components and confirms its robustness and effectiveness. This work contributes to the resurgence of recurrent models in time series forecasting.
☆ Pyramid Vector Quantization for LLMs
Recent works on compression of large language models (LLM) using quantization considered reparameterizing the architecture such that weights are distributed on the sphere. This demonstratively improves the ability to quantize by increasing the mathematical notion of coherence, resulting in fewer weight outliers without affecting the network output. In this work, we aim to further exploit this spherical geometry of the weights when performing quantization by considering Pyramid Vector Quantization (PVQ) for large language models. Arranging points evenly on the sphere is notoriously difficult, especially in high dimensions, and in case approximate solutions exists, representing points explicitly in a codebook is typically not feasible due to its additional memory cost. Instead, PVQ uses a fixed integer lattice on the sphere by projecting points onto the 1-sphere, which allows for efficient encoding and decoding without requiring an explicit codebook in memory. To obtain a practical algorithm, we propose to combine PVQ with scale quantization for which we derive theoretically optimal quantizations, under empirically verified assumptions. Further, we extend pyramid vector quantization to use Hessian information to minimize quantization error under expected feature activations, instead of only relying on weight magnitudes. Experimentally, we achieves state-of-the-art quantization performance with pareto-optimal trade-off between performance and bits per weight and bits per activation, compared to compared methods. On weight-only, we find that we can quantize a Llama-3 70B model to 3.25 bits per weight and retain 98\% accuracy on downstream tasks.
☆ EnvBridge: Bridging Diverse Environments with Cross-Environment Knowledge Transfer for Embodied AI
In recent years, Large Language Models (LLMs) have demonstrated high reasoning capabilities, drawing attention for their applications as agents in various decision-making processes. One notably promising application of LLM agents is robotic manipulation. Recent research has shown that LLMs can generate text planning or control code for robots, providing substantial flexibility and interaction capabilities. However, these methods still face challenges in terms of flexibility and applicability across different environments, limiting their ability to adapt autonomously. Current approaches typically fall into two categories: those relying on environment-specific policy training, which restricts their transferability, and those generating code actions based on fixed prompts, which leads to diminished performance when confronted with new environments. These limitations significantly constrain the generalizability of agents in robotic manipulation. To address these limitations, we propose a novel method called EnvBridge. This approach involves the retention and transfer of successful robot control codes from source environments to target environments. EnvBridge enhances the agent's adaptability and performance across diverse settings by leveraging insights from multiple environments. Notably, our approach alleviates environmental constraints, offering a more flexible and generalizable solution for robotic manipulation tasks. We validated the effectiveness of our method using robotic manipulation benchmarks: RLBench, MetaWorld, and CALVIN. Our experiments demonstrate that LLM agents can successfully leverage diverse knowledge sources to solve complex tasks. Consequently, our approach significantly enhances the adaptability and robustness of robotic manipulation agents in planning across diverse environments.
☆ DNAHLM -- DNA sequence and Human Language mixed large language Model
There are already many DNA large language models, but most of them still follow traditional uses, such as extracting sequence features for classification tasks. More innovative applications of large language models, such as prompt engineering, RAG, and zero-shot or few-shot prediction, remain challenging for DNA-based models. The key issue lies in the fact that DNA models and human natural language models are entirely separate; however, techniques like prompt engineering require the use of natural language, thereby significantly limiting the application of DNA large language models. This paper introduces a hybrid model trained on the GPT-2 network, combining DNA sequences and English text to explore the potential of using prompts and fine-tuning in DNA models. The model has demonstrated its effectiveness in DNA related zero-shot prediction and multitask application.
comment: 10 pages, 7 figures
☆ Bayes without Underfitting: Fully Correlated Deep Learning Posteriors via Alternating Projections
Bayesian deep learning all too often underfits so that the Bayesian prediction is less accurate than a simple point estimate. Uncertainty quantification then comes at the cost of accuracy. For linearized models, the null space of the generalized Gauss-Newton matrix corresponds to parameters that preserve the training predictions of the point estimate. We propose to build Bayesian approximations in this null space, thereby guaranteeing that the Bayesian predictive does not underfit. We suggest a matrix-free algorithm for projecting onto this null space, which scales linearly with the number of parameters and quadratically with the number of output dimensions. We further propose an approximation that only scales linearly with parameters to make the method applicable to generative models. An extensive empirical evaluation shows that the approach scales to large models, including vision transformers with 28 million parameters.
☆ MBD: Multi b-value Denoising of Diffusion Magnetic Resonance Images
We propose a novel approach to denoising diffusion magnetic resonance images (dMRI) using convolutional neural networks, that exploits the benefits of data acquired at multiple b-values to offset the need for many redundant observations. Denoising is especially relevant in dMRI since noise can have a deleterious impact on both quantification accuracy and image preprocessing. The most successful methods proposed to date, like Marchenko-Pastur Principal Component Analysis (MPPCA) denoising, are tailored to diffusion-weighting repeated for many encoding directions. They exploit high redundancy of the dataset that oversamples the diffusion-encoding direction space, since many directions have collinear components. However, there are many dMRI techniques that do not entail a large number of encoding directions or repetitions, and are therefore less suited to this approach. For example, clinical dMRI exams may include as few as three encoding directions, with low or negligible data redundancy across directions. Moreover, promising new dMRI approaches, like spherical b-tensor encoding (STE), benefit from high b-values while sensitizing the signal to diffusion along all directions in just a single shot. We introduce a convolutional neural network approach that we call multi-b-value-based denoising (MBD). MBD exploits the similarity in diffusion-weighted images (DWI) across different b-values but along the same diffusion encoding direction. It allows denoising of diffusion images with high noise variance while avoiding blurring, and using just a small number input images.
comment: this is a biomedical engineering work using machine learning to enhance medical images
☆ Global Optimization of Gaussian Process Acquisition Functions Using a Piecewise-Linear Kernel Approximation
Bayesian optimization relies on iteratively constructing and optimizing an acquisition function. The latter turns out to be a challenging, non-convex optimization problem itself. Despite the relative importance of this step, most algorithms employ sampling- or gradient-based methods, which do not provably converge to global optima. This work investigates mixed-integer programming (MIP) as a paradigm for \textit{global} acquisition function optimization. Specifically, our Piecewise-linear Kernel Mixed Integer Quadratic Programming (PK-MIQP) formulation introduces a piecewise-linear approximation for Gaussian process kernels and admits a corresponding MIQP representation for acquisition functions. We analyze the theoretical regret bounds of the proposed approximation, and empirically demonstrate the framework on synthetic functions, constrained benchmarks, and a hyperparameter tuning task.
comment: 16 pages, 3 figures, 2 tables
☆ Unsupervised Time Series Anomaly Prediction with Importance-based Generative Contrastive Learning
Time series anomaly prediction plays an essential role in many real-world scenarios, such as environmental prevention and prompt maintenance of cyber-physical systems. However, existing time series anomaly prediction methods mainly require supervised training with plenty of manually labeled data, which are difficult to obtain in practice. Besides, unseen anomalies can occur during inference, which could differ from the labeled training data and make these models fail to predict such new anomalies. In this paper, we study a novel problem of unsupervised time series anomaly prediction. We provide a theoretical analysis and propose Importance-based Generative Contrastive Learning (IGCL) to address the aforementioned problems. IGCL distinguishes between normal and anomaly precursors, which are generated by our anomaly precursor pattern generation module. To address the efficiency issues caused by the potential complex anomaly precursor combinations, we propose a memory bank with importance-based scores to adaptively store representative anomaly precursors and generate more complicated anomaly precursors. Extensive experiments on seven benchmark datasets show our method outperforms state-of-the-art baselines on unsupervised time series anomaly prediction problems.
comment: 16 pages
☆ Large Language Model-based Augmentation for Imbalanced Node Classification on Text-Attributed Graphs
Node classification on graphs frequently encounters the challenge of class imbalance, leading to biased performance and posing significant risks in real-world applications. Although several data-centric solutions have been proposed, none of them focus on Text-Attributed Graphs (TAGs), and therefore overlook the potential of leveraging the rich semantics encoded in textual features for boosting the classification of minority nodes. Given this crucial gap, we investigate the possibility of augmenting graph data in the text space, leveraging the textual generation power of Large Language Models (LLMs) to handle imbalanced node classification on TAGs. Specifically, we propose a novel approach called LA-TAG (LLM-based Augmentation on Text-Attributed Graphs), which prompts LLMs to generate synthetic texts based on existing node texts in the graph. Furthermore, to integrate these synthetic text-attributed nodes into the graph, we introduce a text-based link predictor to connect the synthesized nodes with the existing nodes. Our experiments across multiple datasets and evaluation metrics show that our framework significantly outperforms traditional non-textual-based data augmentation strategies and specific node imbalance solutions. This highlights the promise of using LLMs to resolve imbalance issues on TAGs.
comment: 11 pages, 4 figures
☆ Just In Time Transformers
Precise energy load forecasting in residential households is crucial for mitigating carbon emissions and enhancing energy efficiency; indeed, accurate forecasting enables utility companies and policymakers, who advocate sustainable energy practices, to optimize resource utilization. Moreover, smart meters provide valuable information by allowing for granular insights into consumption patterns. Building upon available smart meter data, our study aims to cluster consumers into distinct groups according to their energy usage behaviours, effectively capturing a diverse spectrum of consumption patterns. Next, we design JITtrans (Just In Time transformer), a novel transformer deep learning model that significantly improves energy consumption forecasting accuracy, with respect to traditional forecasting methods. Extensive experimental results validate our claims using proprietary smart meter data. Our findings highlight the potential of advanced predictive technologies to revolutionize energy management and advance sustainable power systems: the development of efficient and eco-friendly energy solutions critically depends on such technologies.
☆ Contrasting Attitudes Towards Current and Future AI Applications for Computerised Interpretation of ECG: A Clinical Stakeholder Interview Study
Objectives: To investigate clinicians' attitudes towards current automated interpretation of ECG and novel AI technologies and their perception of computer-assisted interpretation. Materials and Methods: We conducted a series of interviews with clinicians in the UK. Our study: (i) explores the potential for AI, specifically future 'human-like' computing approaches, to facilitate ECG interpretation and support clinical decision making, and (ii) elicits their opinions about the importance of explainability and trustworthiness of AI algorithms. Results: We performed inductive thematic analysis on interview transcriptions from 23 clinicians and identified the following themes: (i) a lack of trust in current systems, (ii) positive attitudes towards future AI applications and requirements for these, (iii) the relationship between the accuracy and explainability of algorithms, and (iv) opinions on education, possible deskilling, and the impact of AI on clinical competencies. Discussion: Clinicians do not trust current computerised methods, but welcome future 'AI' technologies. Where clinicians trust future AI interpretation to be accurate, they are less concerned that it is explainable. They also preferred ECG interpretation that demonstrated the results of the algorithm visually. Whilst clinicians do not fear job losses, they are concerned about deskilling and the need to educate the workforce to use AI responsibly. Conclusion: Clinicians are positive about the future application of AI in clinical decision-making. Accuracy is a key factor of uptake and visualisations are preferred over current computerised methods. This is viewed as a potential means of training and upskilling, in contrast to the deskilling that automation might be perceived to bring.
☆ CK4Gen: A Knowledge Distillation Framework for Generating High-Utility Synthetic Survival Datasets in Healthcare
Access to real clinical data is heavily restricted by privacy regulations, hindering both healthcare research and education. These constraints slow progress in developing new treatments and data-driven healthcare solutions, while also limiting students' access to real-world datasets, leaving them without essential practical skills. High-utility synthetic datasets are therefore critical for advancing research and providing meaningful training material. However, current generative models -- such as Variational Autoencoders (VAEs) and Generative Adversarial Networks (GANs) -- produce surface-level realism at the expense of healthcare utility, blending distinct patient profiles and producing synthetic data of limited practical relevance. To overcome these limitations, we introduce CK4Gen (Cox Knowledge for Generation), a novel framework that leverages knowledge distillation from Cox Proportional Hazards (CoxPH) models to create synthetic survival datasets that preserve key clinical characteristics, including hazard ratios and survival curves. CK4Gen avoids the interpolation issues seen in VAEs and GANs by maintaining distinct patient risk profiles, ensuring realistic and reliable outputs for research and educational use. Validated across four benchmark datasets -- GBSG2, ACTG320, WHAS500, and FLChain -- CK4Gen outperforms competing techniques by better aligning real and synthetic data, enhancing survival model performance in both discrimination and calibration via data augmentation. As CK4Gen is scalable across clinical conditions, and with code to be made publicly available, future researchers can apply it to their own datasets to generate synthetic versions suitable for open sharing.
☆ Error Feedback under $(L_0,L_1)$-Smoothness: Normalization and Momentum
We provide the first proof of convergence for normalized error feedback algorithms across a wide range of machine learning problems. Despite their popularity and efficiency in training deep neural networks, traditional analyses of error feedback algorithms rely on the smoothness assumption that does not capture the properties of objective functions in these problems. Rather, these problems have recently been shown to satisfy generalized smoothness assumptions, and the theoretical understanding of error feedback algorithms under these assumptions remains largely unexplored. Moreover, to the best of our knowledge, all existing analyses under generalized smoothness either i) focus on single-node settings or ii) make unrealistically strong assumptions for distributed settings, such as requiring data heterogeneity, and almost surely bounded stochastic gradient noise variance. In this paper, we propose distributed error feedback algorithms that utilize normalization to achieve the $O(1/\sqrt{K})$ convergence rate for nonconvex problems under generalized smoothness. Our analyses apply for distributed settings without data heterogeneity conditions, and enable stepsize tuning that is independent of problem parameters. Additionally, we provide strong convergence guarantees of normalized error feedback algorithms for stochastic settings. Finally, we show that due to their larger allowable stepsizes, our new normalized error feedback algorithms outperform their non-normalized counterparts on various tasks, including the minimization of polynomial functions, logistic regression, and ResNet-20 training.
☆ Federated Causal Inference: Multi-Centric ATE Estimation beyond Meta-Analysis
We study Federated Causal Inference, an approach to estimate treatment effects from decentralized data across centers. We compare three classes of Average Treatment Effect (ATE) estimators derived from the Plug-in G-Formula, ranging from simple meta-analysis to one-shot and multi-shot federated learning, the latter leveraging the full data to learn the outcome model (albeit requiring more communication). Focusing on Randomized Controlled Trials (RCTs), we derive the asymptotic variance of these estimators for linear models. Our results provide practical guidance on selecting the appropriate estimator for various scenarios, including heterogeneity in sample sizes, covariate distributions, treatment assignment schemes, and center effects. We validate these findings with a simulation study.
☆ Rethinking generalization of classifiers in separable classes scenarios and over-parameterized regimes
We investigate the learning dynamics of classifiers in scenarios where classes are separable or classifiers are over-parameterized. In both cases, Empirical Risk Minimization (ERM) results in zero training error. However, there are many global minima with a training error of zero, some of which generalize well and some of which do not. We show that in separable classes scenarios the proportion of "bad" global minima diminishes exponentially with the number of training data n. Our analysis provides bounds and learning curves dependent solely on the density distribution of the true error for the given classifier function set, irrespective of the set's size or complexity (e.g., number of parameters). This observation may shed light on the unexpectedly good generalization of over-parameterized Neural Networks. For the over-parameterized scenario, we propose a model for the density distribution of the true error, yielding learning curves that align with experiments on MNIST and CIFAR-10.
☆ Dynamic graph neural networks for enhanced volatility prediction in financial markets
Volatility forecasting is essential for risk management and decision-making in financial markets. Traditional models like Generalized Autoregressive Conditional Heteroskedasticity (GARCH) effectively capture volatility clustering but often fail to model complex, non-linear interdependencies between multiple indices. This paper proposes a novel approach using Graph Neural Networks (GNNs) to represent global financial markets as dynamic graphs. The Temporal Graph Attention Network (Temporal GAT) combines Graph Convolutional Networks (GCNs) and Graph Attention Networks (GATs) to capture the temporal and structural dynamics of volatility spillovers. By utilizing correlation-based and volatility spillover indices, the Temporal GAT constructs directed graphs that enhance the accuracy of volatility predictions. Empirical results from a 15-year study of eight major global indices show that the Temporal GAT outperforms traditional GARCH models and other machine learning methods, particularly in short- to mid-term forecasts. The sensitivity and scenario-based analysis over a range of parameters and hyperparameters further demonstrate the significance of the proposed technique. Hence, this work highlights the potential of GNNs in modeling complex market behaviors, providing valuable insights for financial analysts and investors.
☆ Polyak's Heavy Ball Method Achieves Accelerated Local Rate of Convergence under Polyak-Lojasiewicz Inequality
In this work, we consider the convergence of Polyak's heavy ball method, both in continuous and discrete time, on a non-convex objective function. We recover the convergence rates derived in [Polyak, U.S.S.R. Comput. Math. and Math. Phys., 1964] for strongly convex objective functions, assuming only validity of the Polyak-Lojasiewicz inequality. In continuous time our result holds for all initializations, whereas in the discrete time setting we conduct a local analysis around the global minima. Our results demonstrate that the heavy ball method does, in fact, accelerate on the class of objective functions satisfying the Polyak-Lojasiewicz inequality. This holds even in the discrete time setting, provided the method reaches a neighborhood of the global minima. Instead of the usually employed Lyapunov-type arguments, our approach leverages a new differential geometric perspective of the Polyak-Lojasiewicz inequality proposed in [Rebjock and Boumal, Math. Program., 2024].
☆ Safe Load Balancing in Software-Defined-Networking
High performance, reliability and safety are crucial properties of any Software-Defined-Networking (SDN) system. Although the use of Deep Reinforcement Learning (DRL) algorithms has been widely studied to improve performance, their practical applications are still limited as they fail to ensure safe operations in exploration and decision-making. To fill this gap, we explore the design of a Control Barrier Function (CBF) on top of Deep Reinforcement Learning (DRL) algorithms for load-balancing. We show that our DRL-CBF approach is capable of meeting safety requirements during training and testing while achieving near-optimal performance in testing. We provide results using two simulators: a flow-based simulator, which is used for proof-of-concept and benchmarking, and a packet-based simulator that implements real protocols and scheduling. Thanks to the flow-based simulator, we compared the performance against the optimal policy, solving a Non Linear Programming (NLP) problem with the SCIP solver. Furthermore, we showed that pre-trained models in the flow-based simulator, which is faster, can be transferred to the packet simulator, which is slower but more accurate, with some fine-tuning. Overall, the results suggest that near-optimal Quality-of-Service (QoS) performance in terms of end-to-end delay can be achieved while safety requirements related to link capacity constraints are guaranteed. In the packet-based simulator, we also show that our DRL-CBF algorithms outperform non-RL baseline algorithms. When the models are fine-tuned over a few episodes, we achieved smoother QoS and safety in training, and similar performance in testing compared to the case where models have been trained from scratch.
comment: Accepted to Computer Communications 2024. arXiv admin note: text overlap with arXiv:2401.05525
☆ Fast Graph Sharpness-Aware Minimization for Enhancing and Accelerating Few-Shot Node Classification NeurIPS24
Graph Neural Networks (GNNs) have shown superior performance in node classification. However, GNNs perform poorly in the Few-Shot Node Classification (FSNC) task that requires robust generalization to make accurate predictions for unseen classes with limited labels. To tackle the challenge, we propose the integration of Sharpness-Aware Minimization (SAM)--a technique designed to enhance model generalization by finding a flat minimum of the loss landscape--into GNN training. The standard SAM approach, however, consists of two forward-backward steps in each training iteration, doubling the computational cost compared to the base optimizer (e.g., Adam). To mitigate this drawback, we introduce a novel algorithm, Fast Graph Sharpness-Aware Minimization (FGSAM), that integrates the rapid training of Multi-Layer Perceptrons (MLPs) with the superior performance of GNNs. Specifically, we utilize GNNs for parameter perturbation while employing MLPs to minimize the perturbed loss so that we can find a flat minimum with good generalization more efficiently. Moreover, our method reutilizes the gradient from the perturbation phase to incorporate graph topology into the minimization process at almost zero additional cost. To further enhance training efficiency, we develop FGSAM+ that executes exact perturbations periodically. Extensive experiments demonstrate that our proposed algorithm outperforms the standard SAM with lower computational costs in FSNC tasks. In particular, our FGSAM+ as a SAM variant offers a faster optimization than the base optimizer in most cases. In addition to FSNC, our proposed methods also demonstrate competitive performance in the standard node classification task for heterophilic graphs, highlighting the broad applicability. The code is available at https://github.com/draym28/FGSAM_NeurIPS24.
comment: NeurIPS24; The first two authors contributed equally to this work
☆ Guarantees of a Preconditioned Subgradient Algorithm for Overparameterized Asymmetric Low-rank Matrix Recovery
In this paper, we focus on a matrix factorization-based approach for robust low-rank and asymmetric matrix recovery from corrupted measurements. We address the challenging scenario where the rank of the sought matrix is unknown and employ an overparameterized approach using the variational form of the nuclear norm as a regularizer. We propose a subgradient algorithm that inherits the merits of preconditioned algorithms, whose rate of convergence does not depend on the condition number of the sought matrix, and addresses their current limitation, i.e., the lack of convergence guarantees in the case of asymmetric matrices with unknown rank. In this setting, we provide, for the first time in the literature, linear convergence guarantees for the derived overparameterized preconditioned subgradient algorithm in the presence of gross corruptions. Additionally, by applying our approach to matrix sensing, we highlight its merits when the measurement operator satisfies the mixed-norm restricted isometry properties. Lastly, we present numerical experiments that validate our theoretical results and demonstrate the effectiveness of our approach.
☆ Klein Model for Hyperbolic Neural Networks NeurIPS 2024
Hyperbolic neural networks (HNNs) have been proved effective in modeling complex data structures. However, previous works mainly focused on the Poincar\'e ball model and the hyperboloid model as coordinate representations of the hyperbolic space, often neglecting the Klein model. Despite this, the Klein model offers its distinct advantages thanks to its straight-line geodesics, which facilitates the well-known Einstein midpoint construction, previously leveraged to accompany HNNs in other models. In this work, we introduce a framework for hyperbolic neural networks based on the Klein model. We provide detailed formulation for representing useful operations using the Klein model. We further study the Klein linear layer and prove that the "tangent space construction" of the scalar multiplication and parallel transport are exactly the Einstein scalar multiplication and the Einstein addition, analogous to the M\"obius operations used in the Poincar\'e ball model. We show numerically that the Klein HNN performs on par with the Poincar\'e ball model, providing a third option for HNN that works as a building block for more complicated architectures.
comment: Accepted to NeurIPS 2024 Symmetry and Geometry in Neural Representations Workshop
☆ Masked Clinical Modelling: A Framework for Synthetic and Augmented Survival Data Generation
Access to real clinical data is often restricted due to privacy obligations, creating significant barriers for healthcare research. Synthetic datasets provide a promising solution, enabling secure data sharing and model development. However, most existing approaches focus on data realism rather than utility -- ensuring that models trained on synthetic data yield clinically meaningful insights comparable to those trained on real data. In this paper, we present Masked Clinical Modelling (MCM), a framework inspired by masked language modelling, designed for both data synthesis and conditional data augmentation. We evaluate this prototype on the WHAS500 dataset using Cox Proportional Hazards models, focusing on the preservation of hazard ratios as key clinical metrics. Our results show that data generated using the MCM framework improves both discrimination and calibration in survival analysis, outperforming existing methods. MCM demonstrates strong potential to support survival data analysis and broader healthcare applications.
☆ Test-time Adversarial Defense with Opposite Adversarial Path and High Attack Time Cost
Deep learning models are known to be vulnerable to adversarial attacks by injecting sophisticated designed perturbations to input data. Training-time defenses still exhibit a significant performance gap between natural accuracy and robust accuracy. In this paper, we investigate a new test-time adversarial defense method via diffusion-based recovery along opposite adversarial paths (OAPs). We present a purifier that can be plugged into a pre-trained model to resist adversarial attacks. Different from prior arts, the key idea is excessive denoising or purification by integrating the opposite adversarial direction with reverse diffusion to push the input image further toward the opposite adversarial direction. For the first time, we also exemplify the pitfall of conducting AutoAttack (Rand) for diffusion-based defense methods. Through the lens of time complexity, we examine the trade-off between the effectiveness of adaptive attack and its computation complexity against our defense. Experimental evaluation along with time cost analysis verifies the effectiveness of the proposed method.
☆ Evaluating the Effectiveness of Attack-Agnostic Features for Morphing Attack Detection
Morphing attacks have diversified significantly over the past years, with new methods based on generative adversarial networks (GANs) and diffusion models posing substantial threats to face recognition systems. Recent research has demonstrated the effectiveness of features extracted from large vision models pretrained on bonafide data only (attack-agnostic features) for detecting deep generative images. Building on this, we investigate the potential of these image representations for morphing attack detection (MAD). We develop supervised detectors by training a simple binary linear SVM on the extracted features and one-class detectors by modeling the distribution of bonafide features with a Gaussian Mixture Model (GMM). Our method is evaluated across a comprehensive set of attacks and various scenarios, including generalization to unseen attacks, different source datasets, and print-scan data. Our results indicate that attack-agnostic features can effectively detect morphing attacks, outperforming traditional supervised and one-class detectors from the literature in most scenarios. Additionally, we provide insights into the strengths and limitations of each considered representation and discuss potential future research directions to further enhance the robustness and generalizability of our approach.
comment: Published in the 2024 IEEE International Joint Conference on Biometrics (IJCB)
☆ One-Step Diffusion Distillation through Score Implicit Matching NeurIPS 2024
Despite their strong performances on many generative tasks, diffusion models require a large number of sampling steps in order to generate realistic samples. This has motivated the community to develop effective methods to distill pre-trained diffusion models into more efficient models, but these methods still typically require few-step inference or perform substantially worse than the underlying model. In this paper, we present Score Implicit Matching (SIM) a new approach to distilling pre-trained diffusion models into single-step generator models, while maintaining almost the same sample generation ability as the original model as well as being data-free with no need of training samples for distillation. The method rests upon the fact that, although the traditional score-based loss is intractable to minimize for generator models, under certain conditions we can efficiently compute the gradients for a wide class of score-based divergences between a diffusion model and a generator. SIM shows strong empirical performances for one-step generators: on the CIFAR10 dataset, it achieves an FID of 2.06 for unconditional generation and 1.96 for class-conditional generation. Moreover, by applying SIM to a leading transformer-based diffusion model, we distill a single-step generator for text-to-image (T2I) generation that attains an aesthetic score of 6.42 with no performance decline over the original multi-step counterpart, clearly outperforming the other one-step generators including SDXL-TURBO of 5.33, SDXL-LIGHTNING of 5.34 and HYPER-SDXL of 5.85. We will release this industry-ready one-step transformer-based T2I generator along with this paper.
comment: Accepted by NeurIPS 2024
☆ Sample-Efficient Curriculum Reinforcement Learning for Complex Reward Functions
Reinforcement learning (RL) shows promise in control problems, but its practical application is often hindered by the complexity arising from intricate reward functions with constraints. While the reward hypothesis suggests these competing demands can be encapsulated in a single scalar reward function, designing such functions remains challenging. Building on existing work, we start by formulating preferences over trajectories to derive a realistic reward function that balances goal achievement with constraint satisfaction in the application of mobile robotics with dynamic obstacles. To mitigate reward exploitation in such complex settings, we propose a novel two-stage reward curriculum combined with a flexible replay buffer that adaptively samples experiences. Our approach first learns on a subset of rewards before transitioning to the full reward, allowing the agent to learn trade-offs between objectives and constraints. After transitioning to a new stage, our method continues to make use of past experiences by updating their rewards for sample-efficient learning. We investigate the efficacy of our approach in robot navigation tasks and demonstrate superior performance compared to baselines in terms of true reward achievement and task completion, underlining its effectiveness.
☆ Beyond Retrieval: Generating Narratives in Conversational Recommender Systems
The recent advances in Large Language Model's generation and reasoning capabilities present an opportunity to develop truly conversational recommendation systems. However, effectively integrating recommender system knowledge into LLMs for natural language generation which is tailored towards recommendation tasks remains a challenge. This paper addresses this challenge by making two key contributions. First, we introduce a new dataset (REGEN) for natural language generation tasks in conversational recommendations. REGEN (Reviews Enhanced with GEnerative Narratives) extends the Amazon Product Reviews dataset with rich user narratives, including personalized explanations of product preferences, product endorsements for recommended items, and summaries of user purchase history. REGEN is made publicly available to facilitate further research. Furthermore, we establish benchmarks using well-known generative metrics, and perform an automated evaluation of the new dataset using a rater LLM. Second, the paper introduces a fusion architecture (CF model with an LLM) which serves as a baseline for REGEN. And to the best of our knowledge, represents the first attempt to analyze the capabilities of LLMs in understanding recommender signals and generating rich narratives. We demonstrate that LLMs can effectively learn from simple fusion architectures utilizing interaction-based CF embeddings, and this can be further enhanced using the metadata and personalization data associated with items. Our experiments show that combining CF and content embeddings leads to improvements of 4-12% in key language metrics compared to using either type of embedding individually. We also provide an analysis to interpret how CF and content embeddings contribute to this new generative task.
☆ Survival Models: Proper Scoring Rule and Stochastic Optimization with Competing Risks
When dealing with right-censored data, where some outcomes are missing due to a limited observation period, survival analysis -- known as time-to-event analysis -- focuses on predicting the time until an event of interest occurs. Multiple classes of outcomes lead to a classification variant: predicting the most likely event, a less explored area known as competing risks. Classic competing risks models couple architecture and loss, limiting scalability.To address these issues, we design a strictly proper censoring-adjusted separable scoring rule, allowing optimization on a subset of the data as each observation is evaluated independently. The loss estimates outcome probabilities and enables stochastic optimization for competing risks, which we use for efficient gradient boosting trees. SurvivalBoost not only outperforms 12 state-of-the-art models across several metrics on 4 real-life datasets, both in competing risks and survival settings, but also provides great calibration, the ability to predict across any time horizon, and computation times faster than existing methods.
comment: arXiv admin note: substantial text overlap with arXiv:2406.14085
☆ Efficient Frequency Selective Surface Analysis via End-to-End Model-Based Learning
This paper introduces an innovative end-to-end model-based deep learning approach for efficient electromagnetic analysis of high-dimensional frequency selective surfaces (FSS). Unlike traditional data-driven methods that require large datasets, this approach combines physical insights from equivalent circuit models with deep learning techniques to significantly reduce model complexity and enhance prediction accuracy. Compared to previously introduced model-based learning approaches, the proposed method is trained end-to-end from the physical structure of the FSS (geometric parameters) to its electromagnetic response (S-parameters). Additionally, an improvement in phase prediction accuracy through a modified loss function is presented. Comparisons with direct models, including deep neural networks (DNN) and radial basis function networks (RBFN), demonstrate the superiority of the model-based approach in terms of computational efficiency, model size, and generalization capability.
☆ Theoretical Convergence Guarantees for Variational Autoencoders
Variational Autoencoders (VAE) are popular generative models used to sample from complex data distributions. Despite their empirical success in various machine learning tasks, significant gaps remain in understanding their theoretical properties, particularly regarding convergence guarantees. This paper aims to bridge that gap by providing non-asymptotic convergence guarantees for VAE trained using both Stochastic Gradient Descent and Adam algorithms.We derive a convergence rate of $\mathcal{O}(\log n / \sqrt{n})$, where $n$ is the number of iterations of the optimization algorithm, with explicit dependencies on the batch size, the number of variational samples, and other key hyperparameters. Our theoretical analysis applies to both Linear VAE and Deep Gaussian VAE, as well as several VAE variants, including $\beta$-VAE and IWAE. Additionally, we empirically illustrate the impact of hyperparameters on convergence, offering new insights into the theoretical understanding of VAE training.
☆ Corrected Soft Actor Critic for Continuous Control
The Soft Actor-Critic (SAC) algorithm is known for its stability and high sample efficiency in deep reinforcement learning. However, the tanh transformation applied to sampled actions in SAC distorts the action distribution, hindering the selection of the most probable actions. This paper presents a novel action sampling method that directly identifies and selects the most probable actions within the transformed distribution, thereby addressing this issue. Extensive experiments on standard continuous control benchmarks demonstrate that the proposed method significantly enhances SAC's performance, resulting in faster convergence and higher cumulative rewards compared to the original algorithm.
☆ LLM-Assisted Red Teaming of Diffusion Models through "Failures Are Fated, But Can Be Faded"
In large deep neural networks that seem to perform surprisingly well on many tasks, we also observe a few failures related to accuracy, social biases, and alignment with human values, among others. Therefore, before deploying these models, it is crucial to characterize this failure landscape for engineers to debug or audit models. Nevertheless, it is infeasible to exhaustively test for all possible combinations of factors that could lead to a model's failure. In this paper, we improve the "Failures are fated, but can be faded" framework (arXiv:2406.07145)--a post-hoc method to explore and construct the failure landscape in pre-trained generative models--with a variety of deep reinforcement learning algorithms, screening tests, and LLM-based rewards and state generation. With the aid of limited human feedback, we then demonstrate how to restructure the failure landscape to be more desirable by moving away from the discovered failure modes. We empirically demonstrate the effectiveness of the proposed method on diffusion models. We also highlight the strengths and weaknesses of each algorithm in identifying failure modes.
comment: 13 pages, 11 figures. arXiv admin note: substantial text overlap with arXiv:2406.07145
☆ Interactive Residual Domain Adaptation Networks for Partial Transfer Industrial Fault Diagnosis
The partial domain adaptation (PDA) challenge is a prevalent issue in industrial fault diagnosis. Current PDA approaches primarily rely on adversarial learning for domain adaptation and use reweighting strategies to exclude source samples deemed outliers. However, the transferability of features diminishes from general feature extraction layers to higher task-specific layers in adversarial learning-based adaptation modules, leading to significant negative transfer in PDA settings. We term this issue the adaptation-discrimination paradox (ADP). Furthermore, reweighting strategies often suffer from unreliable pseudo-labels, compromising their effectiveness. Drawing inspiration from traditional classification settings where such partial challenge is not a concern, we propose a novel PDA framework called Interactive Residual Domain Adaptation Networks (IRDAN), which introduces domain-wise models for each domain to provide a new perspective for the PDA challenge. Each domain-wise model is equipped with a residual domain adaptation (RDA) block to mitigate the ADP problem. Additionally, we introduce a confident information flow via an interactive learning strategy, training the modules of IRDAN sequentially to avoid cross-interference. We also establish a reliable stopping criterion for selecting the best-performing model, ensuring practical usability in real-world applications. Experiments have demonstrated the superior performance of the proposed IRDAN.
☆ Progressive Compositionality In Text-to-Image Generative Models
Despite the impressive text-to-image (T2I) synthesis capabilities of diffusion models, they often struggle to understand compositional relationships between objects and attributes, especially in complex settings. Existing solutions have tackled these challenges by optimizing the cross-attention mechanism or learning from the caption pairs with minimal semantic changes. However, can we generate high-quality complex contrastive images that diffusion models can directly discriminate based on visual representations? In this work, we leverage large-language models (LLMs) to compose realistic, complex scenarios and harness Visual-Question Answering (VQA) systems alongside diffusion models to automatically curate a contrastive dataset, ConPair, consisting of 15k pairs of high-quality contrastive images. These pairs feature minimal visual discrepancies and cover a wide range of attribute categories, especially complex and natural scenarios. To learn effectively from these error cases, i.e., hard negative images, we propose EvoGen, a new multi-stage curriculum for contrastive learning of diffusion models. Through extensive experiments across a wide range of compositional scenarios, we showcase the effectiveness of our proposed framework on compositional T2I benchmarks.
☆ Optimal Partial Graph Matching
Partial graph matching addresses the limitations of traditional graph matching by allowing some nodes to remain unmatched, making it applicable to more complex scenarios. However, this flexibility introduces additional complexity, as both the subset of nodes to match and the optimal mapping must be determined. While recent studies have explored deep learning techniques for partial graph matching, a significant limitation remains: the absence of an optimization objective that fully captures the problem's intrinsic nature while enabling efficient solutions. In this paper, we propose a novel optimization framework for partial graph matching, inspired by optimal partial transport. Our approach formulates an objective that enables partial assignments while incorporating matching biases, using weighted total variation as the divergence function to guarantee optimal partial assignments. We employ the Hungarian algorithm to achieve efficient, exact solutions with cubic time complexity. Our contributions are threefold: (i) we introduce a robust optimization objective that balances matched and unmatched nodes; (ii) we establish a connection between partial graph matching and the linear sum assignment problem, enabling efficient solutions; (iii) we propose a deep graph matching architecture with a novel partial matching loss, providing an end-to-end solution. The empirical evaluations on standard graph matching benchmarks demonstrate the efficacy of the proposed approach.
☆ Collapse or Thrive? Perils and Promises of Synthetic Data in a Self-Generating World
The increasing presence of AI-generated content on the internet raises a critical question: What happens when generative machine learning models are pretrained on web-scale datasets containing data created by earlier models? Some authors prophesy $\textit{model collapse}$ under a "$\textit{replace}$" scenario: a sequence of models, the first trained with real data and each later one trained only on synthetic data from its preceding model. In this scenario, models successively degrade. Others see collapse as easily avoidable; in an "$\textit{accumulate}$' scenario, a sequence of models is trained, but each training uses all real and synthetic data generated so far. In this work, we deepen and extend the study of these contrasting scenarios. First, collapse versus avoidance of collapse is studied by comparing the replace and accumulate scenarios on each of three prominent generative modeling settings; we find the same contrast emerges in all three settings. Second, we study a compromise scenario; the available data remains the same as in the accumulate scenario -- but unlike $\textit{accumulate}$ and like $\textit{replace}$, each model is trained using a fixed compute budget; we demonstrate that model test loss on real data is larger than in the $\textit{accumulate}$ scenario, but apparently plateaus, unlike the divergence seen with $\textit{replace}$. Third, we study the relative importance of cardinality and proportion of real data for avoiding model collapse. Surprisingly, we find a non-trivial interaction between real and synthetic data, where the value of synthetic data for reducing test loss depends on the absolute quantity of real data. Our insights are particularly important when forecasting whether future frontier generative models will collapse or thrive, and our results open avenues for empirically and mathematically studying the context-dependent value of synthetic data.
☆ Influential Language Data Selection via Gradient Trajectory Pursuit
Curating a desirable dataset for training has been the core of building highly capable large language models (Touvron et al., 2023; Achiam et al., 2023; Team et al.,2024). Gradient influence scores (Pruthi et al., 2020; Xia et al., 2024) are shown to be correlated with model performance and are commonly used as the criterion for data selection. However, existing methods are built upon either individual sample rankings or inefficient matching process, leading to suboptimal performance or scaling up issues.In this paper, we propose Gradient Trajectory Pursuit (GTP), an algorithm that performs pursuit of gradient trajectories via jointly selecting data points under an L0-norm regularized objective. The proposed algorithm highlights: (1) joint selection instead of independent top-k selection, which automatically de-duplicates samples; (2) higher efficiency with compressive sampling processes, which can be further sped up using a distributed framework. In the experiments, we demonstrate the algorithm in both in-domain and target-domain selection benchmarks and show that it outperforms top-k selection and competitive algorithms consistently, for example, our algorithm chooses as low as 0.5% data to achieve full performance on the targeted instruction tuning tasks
☆ Universal approximation property of ODENet and ResNet with a single activation function
We study a universal approximation property of ODENet and ResNet. The ODENet is a map from an initial value to the final value of an ODE system in a finite interval. It is considered a mathematical model of a ResNet-type deep learning system. We consider dynamical systems with vector fields given by a single composition of the activation function and an affine mapping, which is the most common choice of the ODENet or ResNet vector field in actual machine learning systems. We show that such an ODENet and ResNet with a restricted vector field can uniformly approximate ODENet with a general vector field.
comment: 14 pages
☆ Privacy-hardened and hallucination-resistant synthetic data generation with logic-solvers
Machine-generated data is a valuable resource for training Artificial Intelligence algorithms, evaluating rare workflows, and sharing data under stricter data legislations. The challenge is to generate data that is accurate and private. Current statistical and deep learning methods struggle with large data volumes, are prone to hallucinating scenarios incompatible with reality, and seldom quantify privacy meaningfully. Here we introduce Genomator, a logic solving approach (SAT solving), which efficiently produces private and realistic representations of the original data. We demonstrate the method on genomic data, which arguably is the most complex and private information. Synthetic genomes hold great potential for balancing underrepresented populations in medical research and advancing global data exchange. We benchmark Genomator against state-of-the-art methodologies (Markov generation, Restricted Boltzmann Machine, Generative Adversarial Network and Conditional Restricted Boltzmann Machines), demonstrating an 84-93% accuracy improvement and 95-98% higher privacy. Genomator is also 1000-1600 times more efficient, making it the only tested method that scales to whole genomes. We show the universal trade-off between privacy and accuracy, and use Genomator's tuning capability to cater to all applications along the spectrum, from provable private representations of sensitive cohorts, to datasets with indistinguishable pharmacogenomic profiles. Demonstrating the production-scale generation of tuneable synthetic data can increase trust and pave the way into the clinic.
☆ ClimaQA: An Automated Evaluation Framework for Climate Foundation Models
The use of foundation models in climate science has recently gained significant attention. However, a critical issue remains: the lack of a comprehensive evaluation framework capable of assessing the quality and scientific validity of model outputs. To address this issue, we develop ClimaGen (Climate QA Generator), an automated algorithmic framework that generates question-answer pairs from graduate textbooks with climate scientists in the loop. As a result, we present ClimaQA-Gold, an expert-annotated benchmark dataset alongside ClimaQA-Silver, a large-scale, comprehensive synthetic QA dataset for climate science. Finally, we develop evaluation strategies and compare different Large Language Models (LLMs) on our benchmarks. Our results offer novel insights into various approaches used to enhance climate foundation models.
☆ Graph Transformers Dream of Electric Flow
We show theoretically and empirically that the linear Transformer, when applied to graph data, can implement algorithms that solve canonical problems such as electric flow and eigenvector decomposition. The input to the Transformer is simply the graph incidence matrix; no other explicit positional encoding information is provided. We present explicit weight configurations for implementing each such graph algorithm, and we bound the errors of the constructed Transformers by the errors of the underlying algorithms. Our theoretical findings are corroborated by experiments on synthetic data. Additionally, on a real-world molecular regression task, we observe that the linear Transformer is capable of learning a more effective positional encoding than the default one based on Laplacian eigenvectors. Our work is an initial step towards elucidating the inner-workings of the Transformer for graph data.
☆ Hyperboloid GPLVM for Discovering Continuous Hierarchies via Nonparametric Estimation
Dimensionality reduction (DR) offers a useful representation of complex high-dimensional data. Recent DR methods focus on hyperbolic geometry to derive a faithful low-dimensional representation of hierarchical data. However, existing methods are based on neighbor embedding, frequently ruining the continual relation of the hierarchies. This paper presents hyperboloid Gaussian process (GP) latent variable models (hGP-LVMs) to embed high-dimensional hierarchical data with implicit continuity via nonparametric estimation. We adopt generative modeling using the GP, which brings effective hierarchical embedding and executes ill-posed hyperparameter tuning. This paper presents three variants that employ original point, sparse point, and Bayesian estimations. We establish their learning algorithms by incorporating the Riemannian optimization and active approximation scheme of GP-LVM. For Bayesian inference, we further introduce the reparameterization trick to realize Bayesian latent variable learning. In the last part of this paper, we apply hGP-LVMs to several datasets and show their ability to represent high-dimensional hierarchies in low-dimensional spaces.
☆ Governing equation discovery of a complex system from snapshots
Complex systems in physics, chemistry, and biology that evolve over time with inherent randomness are typically described by stochastic differential equations (SDEs). A fundamental challenge in science and engineering is to determine the governing equations of a complex system from snapshot data. Traditional equation discovery methods often rely on stringent assumptions, such as the availability of the trajectory information or time-series data, and the presumption that the underlying system is deterministic. In this work, we introduce a data-driven, simulation-free framework, called Sparse Identification of Differential Equations from Snapshots (SpIDES), that discovers the governing equations of a complex system from snapshots by utilizing the advanced machine learning techniques to perform three essential steps: probability flow reconstruction, probability density estimation, and Bayesian sparse identification. We validate the effectiveness and robustness of SpIDES by successfully identifying the governing equation of an over-damped Langevin system confined within two potential wells. By extracting interpretable drift and diffusion terms from the SDEs, our framework provides deeper insights into system dynamics, enhances predictive accuracy, and facilitates more effective strategies for managing and simulating stochastic systems.
☆ Lower Bounds for Time-Varying Kernelized Bandits
The optimization of black-box functions with noisy observations is a fundamental problem with widespread applications, and has been widely studied under the assumption that the function lies in a reproducing kernel Hilbert space (RKHS). This problem has been studied extensively in the stationary setting, and near-optimal regret bounds are known via developments in both upper and lower bounds. In this paper, we consider non-stationary scenarios, which are crucial for certain applications but are currently less well-understood. Specifically, we provide the first algorithm-independent lower bounds, where the time variations are subject satisfying a total variation budget according to some function norm. Under $\ell_{\infty}$-norm variations, our bounds are found to be close to the state-of-the-art upper bound (Hong \emph{et al.}, 2023). Under RKHS norm variations, the upper and lower bounds are still reasonably close but with more of a gap, raising the interesting open question of whether non-minor improvements in the upper bound are possible.
☆ Methods of improving LLM training stability
Training stability of large language models(LLMs) is an important research topic. Reproducing training instabilities can be costly, so we use a small language model with 830M parameters and experiment with higher learning rates to force models to diverge. One of the sources of training instability is the growth of logits in attention layers. We extend the focus of the previous work and look not only at the magnitude of the logits but at all outputs of linear layers in the Transformer block. We observe that with a high learning rate the L2 norm of all linear layer outputs can grow with each training step and the model diverges. Specifically we observe that QKV, Proj and FC2 layers have the largest growth of the output magnitude. This prompts us to explore several options: 1) apply layer normalization not only after QK layers but also after Proj and FC2 layers too; 2) apply layer normalization after the QKV layer (and remove pre normalization). 3) apply QK layer normalization together with softmax capping. We show that with the last two methods we can increase learning rate by 1.5x (without model divergence) in comparison to an approach based on QK layer normalization only. Also we observe significant perplexity improvements for all three methods in comparison to the baseline model.
☆ Efficient Antibody Structure Refinement Using Energy-Guided SE(3) Flow Matching
Antibodies are proteins produced by the immune system that recognize and bind to specific antigens, and their 3D structures are crucial for understanding their binding mechanism and designing therapeutic interventions. The specificity of antibody-antigen binding predominantly depends on the complementarity-determining regions (CDR) within antibodies. Despite recent advancements in antibody structure prediction, the quality of predicted CDRs remains suboptimal. In this paper, we develop a novel antibody structure refinement method termed FlowAB based on energy-guided flow matching. FlowAB adopts the powerful deep generative method SE(3) flow matching and simultaneously incorporates important physical prior knowledge into the flow model to guide the generation process. The extensive experiments demonstrate that FlowAB can significantly improve the antibody CDR structures. It achieves new state-of-the-art performance on the antibody structure prediction task when used in conjunction with an appropriate prior model while incurring only marginal computational overhead. This advantage makes FlowAB a practical tool in antibody engineering.
comment: BIBM 2024 regular paper
☆ CoPS: Empowering LLM Agents with Provable Cross-Task Experience Sharing
Sequential reasoning in agent systems has been significantly advanced by large language models (LLMs), yet existing approaches face limitations. Reflection-driven reasoning relies solely on knowledge in pretrained models, limiting performance in novel scenarios, while experience-assisted reasoning often depends on external experiences and lacks clear principles for selecting representative experiences. We address these limitations by proposing CoPS (Cross-Task Experience Sharing), a generalizable algorithm that enhances sequential reasoning by cross-task experience sharing and selection. In detail, CoPS leverages agents' experiences on previous tasks, selecting distribution-matched experiences via a provable pessimism-based strategy to maximize utility while minimizing risks from distribution shifts. Extensive experimental results on benchmarks like Alfworld, Webshop, and HotPotQA demonstrate that CoPS consistently outperforms state-of-the-art baselines, with superior sample efficiency suitable for resource-constrained scenarios. Theoretically, we show that the performance of our algorithm depends on both the quality of the pretrained LLM and the matching between the agent's task-dependent trial distribution and that generated by the LLM. Our work bridges the gap between existing sequential reasoning paradigms and validates the effectiveness of leveraging cross-task experiences, shedding light on the potential to improve agents' generalization and adaptability across diverse tasks. Our codes are available at $\href{https://github.com/uclaml/COPS}{\text{https://github.com/uclaml/COPS}}$.
comment: 25 pages, 5 tables, 3 figures
☆ Linear Partial Gromov-Wasserstein Embedding
The Gromov Wasserstein (GW) problem, a variant of the classical optimal transport (OT) problem, has attracted growing interest in the machine learning and data science communities due to its ability to quantify similarity between measures in different metric spaces. However, like the classical OT problem, GW imposes an equal mass constraint between measures, which restricts its application in many machine learning tasks. To address this limitation, the partial Gromov-Wasserstein (PGW) problem has been introduced, which relaxes the equal mass constraint, enabling the comparison of general positive Radon measures. Despite this, both GW and PGW face significant computational challenges due to their non-convex nature. To overcome these challenges, we propose the linear partial Gromov-Wasserstein (LPGW) embedding, a linearized embedding technique for the PGW problem. For $K$ different metric measure spaces, the pairwise computation of the PGW distance requires solving the PGW problem $\mathcal{O}(K^2)$ times. In contrast, the proposed linearization technique reduces this to $\mathcal{O}(K)$ times. Similar to the linearization technique for the classical OT problem, we prove that LPGW defines a valid metric for metric measure spaces. Finally, we demonstrate the effectiveness of LPGW in practical applications such as shape retrieval and learning with transport-based embeddings, showing that LPGW preserves the advantages of PGW in partial matching while significantly enhancing computational efficiency.
☆ QuasiNav: Asymmetric Cost-Aware Navigation Planning with Constrained Quasimetric Reinforcement Learning ICRA 2025
Autonomous navigation in unstructured outdoor environments is inherently challenging due to the presence of asymmetric traversal costs, such as varying energy expenditures for uphill versus downhill movement. Traditional reinforcement learning methods often assume symmetric costs, which can lead to suboptimal navigation paths and increased safety risks in real-world scenarios. In this paper, we introduce QuasiNav, a novel reinforcement learning framework that integrates quasimetric embeddings to explicitly model asymmetric costs and guide efficient, safe navigation. QuasiNav formulates the navigation problem as a constrained Markov decision process (CMDP) and employs quasimetric embeddings to capture directionally dependent costs, allowing for a more accurate representation of the terrain. This approach is combined with adaptive constraint tightening within a constrained policy optimization framework to dynamically enforce safety constraints during learning. We validate QuasiNav across three challenging navigation scenarios-undulating terrains, asymmetric hill traversal, and directionally dependent terrain traversal-demonstrating its effectiveness in both simulated and real-world environments. Experimental results show that QuasiNav significantly outperforms conventional methods, achieving higher success rates, improved energy efficiency, and better adherence to safety constraints.
comment: Under Review for ICRA 2025
☆ FastAttention: Extend FlashAttention2 to NPUs and Low-resource GPUs
FlashAttention series has been widely applied in the inference of large language models (LLMs). However, FlashAttention series only supports the high-level GPU architectures, e.g., Ampere and Hopper. At present, FlashAttention series is not easily transferrable to NPUs and low-resource GPUs. Moreover, FlashAttention series is inefficient for multi- NPUs or GPUs inference scenarios. In this work, we propose FastAttention which pioneers the adaptation of FlashAttention series for NPUs and low-resource GPUs to boost LLM inference efficiency. Specifically, we take Ascend NPUs and Volta-based GPUs as representatives for designing our FastAttention. We migrate FlashAttention series to Ascend NPUs by proposing a novel two-level tiling strategy for runtime speedup, tiling-mask strategy for memory saving and the tiling-AllReduce strategy for reducing communication overhead, respectively. Besides, we adapt FlashAttention for Volta-based GPUs by redesigning the operands layout in shared memory and introducing a simple yet effective CPU-GPU cooperative strategy for efficient memory utilization. On Ascend NPUs, our FastAttention can achieve a 10.7$\times$ speedup compared to the standard attention implementation. Llama-7B within FastAttention reaches up to 5.16$\times$ higher throughput than within the standard attention. On Volta architecture GPUs, FastAttention yields 1.43$\times$ speedup compared to its equivalents in \texttt{xformers}. Pangu-38B within FastAttention brings 1.46$\times$ end-to-end speedup using FasterTransformer. Coupled with the propose CPU-GPU cooperative strategy, FastAttention supports a maximal input length of 256K on 8 V100 GPUs. All the codes will be made available soon.
☆ RKadiyala at SemEval-2024 Task 8: Black-Box Word-Level Text Boundary Detection in Partially Machine Generated Texts
With increasing usage of generative models for text generation and widespread use of machine generated texts in various domains, being able to distinguish between human written and machine generated texts is a significant challenge. While existing models and proprietary systems focus on identifying whether given text is entirely human written or entirely machine generated, only a few systems provide insights at sentence or paragraph level at likelihood of being machine generated at a non reliable accuracy level, working well only for a set of domains and generators. This paper introduces few reliable approaches for the novel task of identifying which part of a given text is machine generated at a word level while comparing results from different approaches and methods. We present a comparison with proprietary systems , performance of our model on unseen domains' and generators' texts. The findings reveal significant improvements in detection accuracy along with comparison on other aspects of detection capabilities. Finally we discuss potential avenues for improvement and implications of our work. The proposed model is also well suited for detecting which parts of a text are machine generated in outputs of Instruct variants of many LLMs.
comment: published at naacl 2024
☆ Dual-Model Defense: Safeguarding Diffusion Models from Membership Inference Attacks through Disjoint Data Splitting
Diffusion models have demonstrated remarkable capabilities in image synthesis, but their recently proven vulnerability to Membership Inference Attacks (MIAs) poses a critical privacy concern. This paper introduces two novel and efficient approaches (DualMD and DistillMD) to protect diffusion models against MIAs while maintaining high utility. Both methods are based on training two separate diffusion models on disjoint subsets of the original dataset. DualMD then employs a private inference pipeline that utilizes both models. This strategy significantly reduces the risk of black-box MIAs by limiting the information any single model contains about individual training samples. The dual models can also generate "soft targets" to train a private student model in DistillMD, enhancing privacy guarantees against all types of MIAs. Extensive evaluations of DualMD and DistillMD against state-of-the-art MIAs across various datasets in white-box and black-box settings demonstrate their effectiveness in substantially reducing MIA success rates while preserving competitive image generation performance. Notably, our experiments reveal that DistillMD not only defends against MIAs but also mitigates model memorization, indicating that both vulnerabilities stem from overfitting and can be addressed simultaneously with our unified approach.
☆ Parsimonious Dynamic Mode Decomposition: A Robust and Automated Approach for Optimally Sparse Mode Selection in Complex Systems
This paper introduces the Parsimonious Dynamic Mode Decomposition (parsDMD), a novel algorithm designed to automatically select an optimally sparse subset of dynamic modes for both spatiotemporal and purely temporal data. By incorporating time-delay embedding and leveraging Orthogonal Matching Pursuit (OMP), parsDMD ensures robustness against noise and effectively handles complex, nonlinear dynamics. The algorithm is validated on a diverse range of datasets, including standing wave signals, identifying hidden dynamics, fluid dynamics simulations (flow past a cylinder and transonic buffet), and atmospheric sea-surface temperature (SST) data. ParsDMD addresses a significant limitation of the traditional sparsity-promoting DMD (spDMD), which requires manual tuning of sparsity parameters through a rigorous trial-and-error process to balance between single-mode and all-mode solutions. In contrast, parsDMD autonomously determines the optimally sparse subset of modes without user intervention, while maintaining minimal computational complexity. Comparative analyses demonstrate that parsDMD consistently outperforms spDMD by providing more accurate mode identification and effective reconstruction in noisy environments. These advantages render parsDMD an effective tool for real-time diagnostics, forecasting, and reduced-order model construction across various disciplines.
comment: 42 pages, 16 Figures
☆ Enhancing Two-Player Performance Through Single-Player Knowledge Transfer: An Empirical Study on Atari 2600 Games
Playing two-player games using reinforcement learning and self-play can be challenging due to the complexity of two-player environments and the possible instability in the training process. We propose that a reinforcement learning algorithm can train more efficiently and achieve improved performance in a two-player game if it leverages the knowledge from the single-player version of the same game. This study examines the proposed idea in ten different Atari 2600 environments using the Atari 2600 RAM as the input state. We discuss the advantages of using transfer learning from a single-player training process over training in a two-player setting from scratch, and demonstrate our results in a few measures such as training time and average total reward. We also discuss a method of calculating RAM complexity and its relationship to performance.
☆ GE2E-KWS: Generalized End-to-End Training and Evaluation for Zero-shot Keyword Spotting
We propose GE2E-KWS -- a generalized end-to-end training and evaluation framework for customized keyword spotting. Specifically, enrollment utterances are separated and grouped by keywords from the training batch and their embedding centroids are compared to all other test utterance embeddings to compute the loss. This simulates runtime enrollment and verification stages, and improves convergence stability and training speed by optimizing matrix operations compared to SOTA triplet loss approaches. To benchmark different models reliably, we propose an evaluation process that mimics the production environment and compute metrics that directly measure keyword matching accuracy. Trained with GE2E loss, our 419KB quantized conformer model beats a 7.5GB ASR encoder by 23.6% relative AUC, and beats a same size triplet loss model by 60.7% AUC. Our KWS models are natively streamable with low memory footprints, and designed to continuously run on-device with no retraining needed for new keywords (zero-shot).
comment: 8 pages, 6 figures, 2 tables The paper is accepted in IEEE Spoken Language Technology (SLT) 2024
☆ LLMScan: Causal Scan for LLM Misbehavior Detection
Despite the success of Large Language Models (LLMs) across various fields, their potential to generate untruthful, biased and harmful responses poses significant risks, particularly in critical applications. This highlights the urgent need for systematic methods to detect and prevent such misbehavior. While existing approaches target specific issues such as harmful responses, this work introduces LLMScan, an innovative LLM monitoring technique based on causality analysis, offering a comprehensive solution. LLMScan systematically monitors the inner workings of an LLM through the lens of causal inference, operating on the premise that the LLM's `brain' behaves differently when misbehaving. By analyzing the causal contributions of the LLM's input tokens and transformer layers, LLMScan effectively detects misbehavior. Extensive experiments across various tasks and models reveal clear distinctions in the causal distributions between normal behavior and misbehavior, enabling the development of accurate, lightweight detectors for a variety of misbehavior detection tasks.
☆ General Frameworks for Conditional Two-Sample Testing
We study the problem of conditional two-sample testing, which aims to determine whether two populations have the same distribution after accounting for confounding factors. This problem commonly arises in various applications, such as domain adaptation and algorithmic fairness, where comparing two groups is essential while controlling for confounding variables. We begin by establishing a hardness result for conditional two-sample testing, demonstrating that no valid test can have significant power against any single alternative without proper assumptions. We then introduce two general frameworks that implicitly or explicitly target specific classes of distributions for their validity and power. Our first framework allows us to convert any conditional independence test into a conditional two-sample test in a black-box manner, while preserving the asymptotic properties of the original conditional independence test. The second framework transforms the problem into comparing marginal distributions with estimated density ratios, which allows us to leverage existing methods for marginal two-sample testing. We demonstrate this idea in a concrete manner with classification and kernel-based methods. Finally, simulation studies are conducted to illustrate the proposed frameworks in finite-sample scenarios.
comment: 39 pages, 6 figures
☆ Benchmarking Smoothness and Reducing High-Frequency Oscillations in Continuous Control Policies IROS 2024
Reinforcement learning (RL) policies are prone to high-frequency oscillations, especially undesirable when deploying to hardware in the real-world. In this paper, we identify, categorize, and compare methods from the literature that aim to mitigate high-frequency oscillations in deep RL. We define two broad classes: loss regularization and architectural methods. At their core, these methods incentivize learning a smooth mapping, such that nearby states in the input space produce nearby actions in the output space. We present benchmarks in terms of policy performance and control smoothness on traditional RL environments from the Gymnasium and a complex manipulation task, as well as three robotics locomotion tasks that include deployment and evaluation with real-world hardware. Finally, we also propose hybrid methods that combine elements from both loss regularization and architectural methods. We find that the best-performing hybrid outperforms other methods, and improves control smoothness by 26.8% over the baseline, with a worst-case performance degradation of just 2.8%.
comment: Presented in IROS 2024
☆ SoK: Dataset Copyright Auditing in Machine Learning Systems
As the implementation of machine learning (ML) systems becomes more widespread, especially with the introduction of larger ML models, we perceive a spring demand for massive data. However, it inevitably causes infringement and misuse problems with the data, such as using unauthorized online artworks or face images to train ML models. To address this problem, many efforts have been made to audit the copyright of the model training dataset. However, existing solutions vary in auditing assumptions and capabilities, making it difficult to compare their strengths and weaknesses. In addition, robustness evaluations usually consider only part of the ML pipeline and hardly reflect the performance of algorithms in real-world ML applications. Thus, it is essential to take a practical deployment perspective on the current dataset copyright auditing tools, examining their effectiveness and limitations. Concretely, we categorize dataset copyright auditing research into two prominent strands: intrusive methods and non-intrusive methods, depending on whether they require modifications to the original dataset. Then, we break down the intrusive methods into different watermark injection options and examine the non-intrusive methods using various fingerprints. To summarize our results, we offer detailed reference tables, highlight key points, and pinpoint unresolved issues in the current literature. By combining the pipeline in ML systems and analyzing previous studies, we highlight several future directions to make auditing tools more suitable for real-world copyright protection requirements.
comment: To appear in the IEEE Symposium on Security and Privacy 2025, San Francisco, CA, USA
☆ Real-time Sub-milliwatt Epilepsy Detection Implemented on a Spiking Neural Network Edge Inference Processor
Analyzing electroencephalogram (EEG) signals to detect the epileptic seizure status of a subject presents a challenge to existing technologies aimed at providing timely and efficient diagnosis. In this study, we aimed to detect interictal and ictal periods of epileptic seizures using a spiking neural network (SNN). Our proposed approach provides an online and real-time preliminary diagnosis of epileptic seizures and helps to detect possible pathological conditions.To validate our approach, we conducted experiments using multiple datasets. We utilized a trained SNN to identify the presence of epileptic seizures and compared our results with those of related studies. The SNN model was deployed on Xylo, a digital SNN neuromorphic processor designed to process temporal signals. Xylo efficiently simulates spiking leaky integrate-and-fire neurons with exponential input synapses. Xylo has much lower energy requirments than traditional approaches to signal processing, making it an ideal platform for developing low-power seizure detection systems.Our proposed method has a high test accuracy of 93.3% and 92.9% when classifying ictal and interictal periods. At the same time, the application has an average power consumption of 87.4 uW(IO power) + 287.9 uW(computational power) when deployed to Xylo. Our method demonstrates excellent low-latency performance when tested on multiple datasets. Our work provides a new solution for seizure detection, and it is expected to be widely used in portable and wearable devices in the future.
☆ Assessing and improving reliability of neighbor embedding methods: a map-continuity perspective
Visualizing high-dimensional data is an important routine for understanding biomedical data and interpreting deep learning models. Neighbor embedding methods, such as t-SNE, UMAP, and LargeVis, among others, are a family of popular visualization methods which reduce high-dimensional data to two dimensions. However, recent studies suggest that these methods often produce visual artifacts, potentially leading to incorrect scientific conclusions. Recognizing that the current limitation stems from a lack of data-independent notions of embedding maps, we introduce a novel conceptual and computational framework, LOO-map, that learns the embedding maps based on a classical statistical idea known as the leave-one-out. LOO-map extends the embedding over a discrete set of input points to the entire input space, enabling a systematic assessment of map continuity, and thus the reliability of the visualizations. We find for many neighbor embedding methods, their embedding maps can be intrinsically discontinuous. The discontinuity induces two types of observed map distortion: ``overconfidence-inducing discontinuity," which exaggerates cluster separation, and ``fracture-inducing discontinuity," which creates spurious local structures. Building upon LOO-map, we propose two diagnostic point-wise scores -- perturbation score and singularity score -- to address these limitations. These scores can help identify unreliable embedding points, detect out-of-distribution data, and guide hyperparameter selection. Our approach is flexible and works as a wrapper around many neighbor embedding algorithms. We test our methods across multiple real-world datasets from computer vision and single-cell omics to demonstrate their effectiveness in enhancing the interpretability and accuracy of visualizations.
comment: 43 pages, 15 figures
☆ GALA: Graph Diffusion-based Alignment with Jigsaw for Source-free Domain Adaptation
Source-free domain adaptation is a crucial machine learning topic, as it contains numerous applications in the real world, particularly with respect to data privacy. Existing approaches predominantly focus on Euclidean data, such as images and videos, while the exploration of non-Euclidean graph data remains scarce. Recent graph neural network (GNN) approaches can suffer from serious performance decline due to domain shift and label scarcity in source-free adaptation scenarios. In this study, we propose a novel method named Graph Diffusion-based Alignment with Jigsaw (GALA), tailored for source-free graph domain adaptation. To achieve domain alignment, GALA employs a graph diffusion model to reconstruct source-style graphs from target data. Specifically, a score-based graph diffusion model is trained using source graphs to learn the generative source styles. Then, we introduce perturbations to target graphs via a stochastic differential equation instead of sampling from a prior, followed by the reverse process to reconstruct source-style graphs. We feed the source-style graphs into an off-the-shelf GNN and introduce class-specific thresholds with curriculum learning, which can generate accurate and unbiased pseudo-labels for target graphs. Moreover, we develop a simple yet effective graph-mixing strategy named graph jigsaw to combine confident graphs and unconfident graphs, which can enhance generalization capabilities and robustness via consistency learning. Extensive experiments on benchmark datasets validate the effectiveness of GALA.
comment: IEEE TPAMI
☆ Graph Sampling for Scalable and Expressive Graph Neural Networks on Homophilic Graphs
Graph Neural Networks (GNNs) excel in many graph machine learning tasks but face challenges when scaling to large networks. GNN transferability allows training on smaller graphs and applying the model to larger ones, but existing methods often rely on random subsampling, leading to disconnected subgraphs and reduced model expressivity. We propose a novel graph sampling algorithm that leverages feature homophily to preserve graph structure. By minimizing the trace of the data correlation matrix, our method better preserves the graph Laplacian's rank than random sampling while achieving lower complexity than spectral methods. Experiments on citation networks show improved performance in preserving graph rank and GNN transferability compared to random sampling.
☆ ViMGuard: A Novel Multi-Modal System for Video Misinformation Guarding
The rise of social media and short-form video (SFV) has facilitated a breeding ground for misinformation. With the emergence of large language models, significant research has gone into curbing this misinformation problem with automatic false claim detection for text. Unfortunately, the automatic detection of misinformation in SFV is a more complex problem that remains largely unstudied. While text samples are monomodal (only containing words), SFVs comprise three different modalities: words, visuals, and non-linguistic audio. In this work, we introduce Video Masked Autoencoders for Misinformation Guarding (ViMGuard), the first deep-learning architecture capable of fact-checking an SFV through analysis of all three of its constituent modalities. ViMGuard leverages a dual-component system. First, Video and Audio Masked Autoencoders analyze the visual and non-linguistic audio elements of a video to discern its intention; specifically whether it intends to make an informative claim. If it is deemed that the SFV has informative intent, it is passed through our second component: a Retrieval Augmented Generation system that validates the factual accuracy of spoken words. In evaluation, ViMGuard outperformed three cutting-edge fact-checkers, thus setting a new standard for SFV fact-checking and marking a significant stride toward trustworthy news on social platforms. To promote further testing and iteration, VimGuard was deployed into a Chrome extension and all code was open-sourced on GitHub.
comment: 7 pages, 2 figures
☆ Do Robot Snakes Dream like Electric Sheep? Investigating the Effects of Architectural Inductive Biases on Hallucination
The growth in prominence of large language models (LLMs) in everyday life can be largely attributed to their generative abilities, yet some of this is also owed to the risks and costs associated with their use. On one front is their tendency to \textit{hallucinate} false or misleading information, limiting their reliability. On another is the increasing focus on the computational limitations associated with traditional self-attention based LLMs, which has brought about new alternatives, in particular recurrent models, meant to overcome them. Yet it remains uncommon to consider these two concerns simultaneously. Do changes in architecture exacerbate/alleviate existing concerns about hallucinations? Do they affect how and where they occur? Through an extensive evaluation, we study how these architecture-based inductive biases affect the propensity to hallucinate. While hallucination remains a general phenomenon not limited to specific architectures, the situations in which they occur and the ease with which specific types of hallucinations can be induced can significantly differ based on the model architecture. These findings highlight the need for better understanding both these problems in conjunction with each other, as well as consider how to design more universal techniques for handling hallucinations.
☆ DROP: Distributional and Regular Optimism and Pessimism for Reinforcement Learning
In reinforcement learning (RL), temporal difference (TD) error is known to be related to the firing rate of dopamine neurons. It has been observed that each dopamine neuron does not behave uniformly, but each responds to the TD error in an optimistic or pessimistic manner, interpreted as a kind of distributional RL. To explain such a biological data, a heuristic model has also been designed with learning rates asymmetric for the positive and negative TD errors. However, this heuristic model is not theoretically-grounded and unknown whether it can work as a RL algorithm. This paper therefore introduces a novel theoretically-grounded model with optimism and pessimism, which is derived from control as inference. In combination with ensemble learning, a distributional value function as a critic is estimated from regularly introduced optimism and pessimism. Based on its central value, a policy in an actor is improved. This proposed algorithm, so-called DROP (distributional and regular optimism and pessimism), is compared on dynamic tasks. Although the heuristic model showed poor learning performance, DROP showed excellent one in all tasks with high generality. In other words, it was suggested that DROP is a new model that can elicit the potential contributions of optimism and pessimism.
comment: 16 pages, 11 figures
☆ Multi-modal Machine Learning Analysis of X-ray Absorption Near-Edge Spectra and Pair Distribution Functions: Performance and Interpretability towards Experimental Design
We used off-the-shelf interpretable ML techniques to combine information from multiple heterogeneous spectra: X-ray absorption near-edge spectra (XANES) and atomic pair distribution functions (PDFs), to extract information about local structure and chemistry of transition metal oxides. This approach enabled us to analyze the relative contributions of the different spectra to different prediction tasks. Specifically, we trained random forest models on XANES, PDF, and both of them combined, to extract charge (oxidation) state, coordination number, and mean nearest-neighbor bond length of transition metal cations in oxides. We find that XANES-only models tend to outperform the PDF-only models for all the tasks, and information from XANES often dominated when the two inputs were combined. This was even true for structural tasks where we might expect PDF to dominate. However, the performance gap closes when we used species-specific differential PDFs (dPDFs) as the inputs instead of total PDFs. Our results highlight that XANES contains rich structural information and may be further developed as a structural probe. Our interpretable, multimodal approach is quick and easy to implement when suitable structural and spectroscopic databases are available. This approach provides valuable insights into the relative strengths of different modalities for a practical scientific goal, guiding researchers in their experiment design tasks such as deciding when it is useful to combine complementary techniques in a scientific investigation.
☆ Evolution with Opponent-Learning Awareness
The universe involves many independent co-learning agents as an ever-evolving part of our observed environment. Yet, in practice, Multi-Agent Reinforcement Learning (MARL) applications are usually constrained to small, homogeneous populations and remain computationally intensive. In this paper, we study how large heterogeneous populations of learning agents evolve in normal-form games. We show how, under assumptions commonly made in the multi-armed bandit literature, Multi-Agent Policy Gradient closely resembles the Replicator Dynamic, and we further derive a fast, parallelizable implementation of Opponent-Learning Awareness tailored for evolutionary simulations. This enables us to simulate the evolution of very large populations made of heterogeneous co-learning agents, under both naive and advanced learning strategies. We demonstrate our approach in simulations of 200,000 agents, evolving in the classic games of Hawk-Dove, Stag-Hunt, and Rock-Paper-Scissors. Each game highlights distinct ways in which Opponent-Learning Awareness affects evolution.
comment: 12 pages, 10 figures
☆ Bauplan: zero-copy, scale-up FaaS for data pipelines
Chaining functions for longer workloads is a key use case for FaaS platforms in data applications. However, modern data pipelines differ significantly from typical serverless use cases (e.g., webhooks and microservices); this makes it difficult to retrofit existing pipeline frameworks due to structural constraints. In this paper, we describe these limitations in detail and introduce bauplan, a novel FaaS programming model and serverless runtime designed for data practitioners. bauplan enables users to declaratively define functional Directed Acyclic Graphs (DAGs) along with their runtime environments, which are then efficiently executed on cloud-based workers. We show that bauplan achieves both better performance and a superior developer experience for data workloads by making the trade-off of reducing generality in favor of data-awareness
comment: Accepted for the 10th International Workshop on Serverless Computing (pre-print)
♻ ☆ The Persian Rug: solving toy models of superposition using large-scale symmetries
We present a complete mechanistic description of the algorithm learned by a minimal non-linear sparse data autoencoder in the limit of large input dimension. The model, originally presented in arXiv:2209.10652, compresses sparse data vectors through a linear layer and decompresses using another linear layer followed by a ReLU activation. We notice that when the data is permutation symmetric (no input feature is privileged) large models reliably learn an algorithm that is sensitive to individual weights only through their large-scale statistics. For these models, the loss function becomes analytically tractable. Using this understanding, we give the explicit scalings of the loss at high sparsity, and show that the model is near-optimal among recently proposed architectures. In particular, changing or adding to the activation function any elementwise or filtering operation can at best improve the model's performance by a constant factor. Finally, we forward-engineer a model with the requisite symmetries and show that its loss precisely matches that of the trained models. Unlike the trained model weights, the low randomness in the artificial weights results in miraculous fractal structures resembling a Persian rug, to which the algorithm is oblivious. Our work contributes to neural network interpretability by introducing techniques for understanding the structure of autoencoders. Code to reproduce our results can be found at https://github.com/KfirD/PersianRug .
comment: Improved arguments, presentation. No changes to results
♻ ☆ On high-dimensional modifications of the nearest neighbor classifier
Nearest neighbor classifier is arguably the most simple and popular nonparametric classifier available in the literature. However, due to the concentration of pairwise distances and the violation of the neighborhood structure, this classifier often suffers in high-dimension, low-sample size (HDLSS) situations, especially when the scale difference between the competing classes dominates their location difference. Several attempts have been made in the literature to take care of this problem. In this article, we discuss some of these existing methods and propose some new ones. We carry out some theoretical investigations in this regard and analyze several simulated and benchmark datasets to compare the empirical performances of proposed methods with some of the existing ones.
♻ ☆ Context-Parametric Inversion: Why Instruction Finetuning May Not Actually Improve Context Reliance
A standard practice when using large language models is for users to supplement their instruction with an input context containing new information for the model to process. However, models struggle to reliably follow the input context, especially when it conflicts with their parametric knowledge from pretraining. In-principle, one would expect models to adapt to the user context better after instruction finetuning, particularly when handling knowledge conflicts. However, we observe a surprising failure mode: during instruction tuning, the context reliance under knowledge conflicts initially increases as expected, but then gradually decreases as instruction finetuning progresses. This happens while the performance on standard benchmarks keeps on increasing far after this drop. We call this phenomenon context-parametric inversion and observe it across multiple general purpose instruction tuning datasets such as TULU, Alpaca and Ultrachat, across different model families like Llama, Mistral, and Pythia. We perform various controlled studies and theoretical analysis to show that context-parametric inversion occurs due to examples in the instruction finetuning data where the input context provides information that aligns with model's parametric knowledge. Our analysis suggests some natural mitigation strategies with limited but insightful gains, and serves as a useful starting point in addressing this deficiency in instruction finetuning.
comment: Under Review
♻ ☆ Pessimistic asynchronous sampling in high-cost Bayesian optimization
Asynchronous Bayesian optimization is a recently implemented technique that allows for parallel operation of experimental systems and disjointed workflows. Contrasting with serial Bayesian optimization which individually selects experiments one at a time after conducting a measurement for each experiment, asynchronous policies sequentially assign multiple experiments before measurements can be taken and evaluate new measurements continuously as they are made available. This technique allows for faster data generation and therefore faster optimization of an experimental space. This work extends the capabilities of asynchronous optimization methods beyond prior studies by evaluating four additional policies that incorporate pessimistic predictions in the training data set. Combined with a conventional policy that uses model predictions, the five total policies were evaluated in a simulated environment and benchmarked with serial sampling. Under some conditions and parameter space dimensionalities, the pessimistic prediction asynchronous policy reached optimum experimental conditions in significantly fewer experiments than equivalent serial policies and proved to be less susceptible to convergence onto local optima at higher dimensions. Without accounting for the faster sampling rate, the pessimistic asynchronous algorithm presented in this work could result in more efficient algorithm driven optimization of high-cost experimental spaces. Accounting for sampling rate, the presented asynchronous algorithm could allow for faster optimization in experimental spaces where multiple experiments can be run before results are collected.
♻ ☆ SMARLA: A Safety Monitoring Approach for Deep Reinforcement Learning Agents
Deep Reinforcement Learning (DRL) has made significant advancements in various fields, such as autonomous driving, healthcare, and robotics, by enabling agents to learn optimal policies through interactions with their environments. However, the application of DRL in safety-critical domains presents challenges, particularly concerning the safety of the learned policies. DRL agents, which are focused on maximizing rewards, may select unsafe actions, leading to safety violations. Runtime safety monitoring is thus essential to ensure the safe operation of these agents, especially in unpredictable and dynamic environments. This paper introduces SMARLA, a black-box safety monitoring approach specifically designed for DRL agents. SMARLA utilizes machine learning to predict safety violations by observing the agent's behavior during execution. The approach is based on Q-values, which reflect the expected reward for taking actions in specific states. SMARLA employs state abstraction to reduce the complexity of the state space, enhancing the predictive capabilities of the monitoring model. Such abstraction enables the early detection of unsafe states, allowing for the implementation of corrective and preventive measures before incidents occur. We quantitatively and qualitatively validated SMARLA on three well-known case studies widely used in DRL research. Empirical results reveal that SMARLA is accurate at predicting safety violations, with a low false positive rate, and can predict violations at an early stage, approximately halfway through the execution of the agent, before violations occur. We also discuss different decision criteria, based on confidence intervals of the predicted violation probabilities, to trigger safety mechanisms aiming at a trade-off between early detection and low false positive rates.
♻ ☆ Universal approximation property of Banach space-valued random feature models including random neural networks
We introduce a Banach space-valued extension of random feature learning, a data-driven supervised machine learning technique for large-scale kernel approximation. By randomly initializing the feature maps, only the linear readout needs to be trained, which reduces the computational complexity substantially. Viewing random feature models as Banach space-valued random variables, we prove a universal approximation result in the corresponding Bochner space. Moreover, we derive approximation rates and an explicit algorithm to learn an element of the given Banach space by such models. The framework of this paper includes random trigonometric/Fourier regression and in particular random neural networks which are single-hidden-layer feedforward neural networks whose weights and biases are randomly initialized, whence only the linear readout needs to be trained. For the latter, we can then lift the universal approximation property of deterministic neural networks to random neural networks, even within function spaces over non-compact domains, e.g., weighted spaces, $L^p$-spaces, and (weighted) Sobolev spaces, where the latter includes the approximation of the (weak) derivatives. In addition, we analyze when the training costs for approximating a given function grow polynomially in both the input/output dimension and the reciprocal of a pre-specified tolerated approximation error. Furthermore, we demonstrate in a numerical example the empirical advantages of random feature models over their deterministic counterparts.
comment: 64 pages, 3 figures
♻ ☆ Data-driven rainfall prediction at a regional scale: a case study with Ghana
With a warming planet, tropical regions are expected to experience the brunt of climate change, with more intense and more volatile rainfall events. Currently, state-of-the-art numerical weather prediction (NWP) models are known to struggle to produce skillful rainfall forecasts in tropical regions of Africa. There is thus a pressing need for improved rainfall forecasting in these regions. Over the last decade or so, the increased availability of large-scale meteorological datasets and the development of powerful machine learning models have opened up new opportunities for data-driven weather forecasting. Focusing on Ghana in this study, we use these tools to develop two U-Net convolutional neural network (CNN) models, to predict 24h rainfall at 12h and 30h lead-time. The models were trained using data from the ERA5 reanalysis dataset, and the GPM-IMERG dataset. A special attention was paid to interpretability. We developed a novel statistical methodology that allowed us to probe the relative importance of the meteorological variables input in our model, offering useful insights into the factors that drive precipitation in the Ghana region. Empirically, we found that our 12h lead-time model has performances that match, and in some accounts are better than the 18h lead-time forecasts produced by the ECMWF (as available in the TIGGE dataset). We also found that combining our data-driven model with classical NWP further improves forecast accuracy.
♻ ☆ Sample Compression Unleashed: New Generalization Bounds for Real Valued Losses
The sample compression theory provides generalization guarantees for predictors that can be fully defined using a subset of the training dataset and a (short) message string, generally defined as a binary sequence. Previous works provided generalization bounds for the zero-one loss, which is restrictive notably when applied to deep learning approaches. In this paper, we present a general framework for deriving new sample compression bounds that hold for real-valued unbounded losses. Using the Pick-To-Learn (P2L) meta-algorithm, which transforms the training method of any machine-learning predictor to yield sample-compressed predictors, we empirically demonstrate the tightness of the bounds and their versatility by evaluating them on random forests and multiple types of neural networks.
♻ ☆ The Impact of Large Language Models in Academia: from Writing to Speaking
Large language models (LLMs) are increasingly impacting human society, particularly in textual information. Based on more than 30,000 papers and 1,000 presentations from machine learning conferences, we examined and compared the words used in writing and speaking, representing the first large-scale study of how LLMs influence the two main modes of verbal communication and expression within the same group of people. Our empirical results show that LLM-style words such as "significant" have been used more frequently in abstracts and oral presentations. The impact on speaking is beginning to emerge and is likely to grow in the future, calling attention to the implicit influence and ripple effect of LLMs on human society.
comment: 23 pages
♻ ☆ A Bayesian Framework for Clustered Federated Learning
One of the main challenges of federated learning (FL) is handling non-independent and identically distributed (non-IID) client data, which may occur in practice due to unbalanced datasets and use of different data sources across clients. Knowledge sharing and model personalization are key strategies for addressing this issue. Clustered federated learning is a class of FL methods that groups clients that observe similarly distributed data into clusters, such that every client is typically associated with one data distribution and participates in training a model for that distribution along their cluster peers. In this paper, we present a unified Bayesian framework for clustered FL which associates clients to clusters. Then we propose several practical algorithms to handle the, otherwise growing, data associations in a way that trades off performance and computational complexity. This work provides insights on client-cluster associations and enables client knowledge sharing in new ways. The proposed framework circumvents the need for unique client-cluster associations, which is seen to increase the performance of the resulting models in a variety of experiments.
♻ ☆ FDINet: Protecting against DNN Model Extraction via Feature Distortion Index
Machine Learning as a Service (MLaaS) platforms have gained popularity due to their accessibility, cost-efficiency, scalability, and rapid development capabilities. However, recent research has highlighted the vulnerability of cloud-based models in MLaaS to model extraction attacks. In this paper, we introduce FDINET, a novel defense mechanism that leverages the feature distribution of deep neural network (DNN) models. Concretely, by analyzing the feature distribution from the adversary's queries, we reveal that the feature distribution of these queries deviates from that of the model's training set. Based on this key observation, we propose Feature Distortion Index (FDI), a metric designed to quantitatively measure the feature distribution deviation of received queries. The proposed FDINET utilizes FDI to train a binary detector and exploits FDI similarity to identify colluding adversaries from distributed extraction attacks. We conduct extensive experiments to evaluate FDINET against six state-of-the-art extraction attacks on four benchmark datasets and four popular model architectures. Empirical results demonstrate the following findings FDINET proves to be highly effective in detecting model extraction, achieving a 100% detection accuracy on DFME and DaST. FDINET is highly efficient, using just 50 queries to raise an extraction alarm with an average confidence of 96.08% for GTSRB. FDINET exhibits the capability to identify colluding adversaries with an accuracy exceeding 91%. Additionally, it demonstrates the ability to detect two types of adaptive attacks.
comment: Accepted to IEEE Transactions on Dependable and Secure Computing
♻ ☆ ControlSpeech: Towards Simultaneous Zero-shot Speaker Cloning and Zero-shot Language Style Control With Decoupled Codec
In this paper, we present ControlSpeech, a text-to-speech (TTS) system capable of fully cloning the speaker's voice and enabling arbitrary control and adjustment of speaking style, merely based on a few seconds of audio prompt and a simple textual style description prompt. Prior zero-shot TTS models and controllable TTS models either could only mimic the speaker's voice without further control and adjustment capabilities or were unrelated to speaker-specific voice generation. Therefore, ControlSpeech focuses on a more challenging new task-a TTS system with controllable timbre, content, and style at the same time. ControlSpeech takes speech prompts, content prompts, and style prompts as inputs and utilizes bidirectional attention and mask-based parallel decoding to capture corresponding codec representations in a discrete decoupling codec space. Moreover, we discovered the issue of text style controllability in a many-to-many mapping fashion and proposed the Style Mixture Semantic Density (SMSD) model to resolve this problem. SMSD module which is based on Gaussian mixture density networks, is designed to enhance the fine-grained partitioning and sampling capabilities of style semantic information and generate speech with more diverse styles. In terms of experiments, we make available a controllable model toolkit called ControlToolkit with a new style controllable dataset, some replicated baseline models and propose new metrics to evaluate both the control capability and the quality of generated audio in ControlSpeech. The relevant ablation studies validate the necessity of each component in ControlSpeech is necessary. We hope that ControlSpeech can establish the next foundation paradigm of controllable speech synthesis. The relevant code and demo are available at https://github.com/jishengpeng/ControlSpeech .
♻ ☆ Boosting Sample Efficiency and Generalization in Multi-agent Reinforcement Learning via Equivariance NeurIPS 2024
Multi-Agent Reinforcement Learning (MARL) struggles with sample inefficiency and poor generalization [1]. These challenges are partially due to a lack of structure or inductive bias in the neural networks typically used in learning the policy. One such form of structure that is commonly observed in multi-agent scenarios is symmetry. The field of Geometric Deep Learning has developed Equivariant Graph Neural Networks (EGNN) that are equivariant (or symmetric) to rotations, translations, and reflections of nodes. Incorporating equivariance has been shown to improve learning efficiency and decrease error [ 2 ]. In this paper, we demonstrate that EGNNs improve the sample efficiency and generalization in MARL. However, we also show that a naive application of EGNNs to MARL results in poor early exploration due to a bias in the EGNN structure. To mitigate this bias, we present Exploration-enhanced Equivariant Graph Neural Networks or E2GN2. We compare E2GN2 to other common function approximators using common MARL benchmarks MPE and SMACv2. E2GN2 demonstrates a significant improvement in sample efficiency, greater final reward convergence, and a 2x-5x gain in over standard GNNs in our generalization tests. These results pave the way for more reliable and effective solutions in complex multi-agent systems.
comment: accepted as a poster at NeurIPS 2024
♻ ☆ Credal Bayesian Deep Learning
Uncertainty quantification and robustness to distribution shifts are important goals in machine learning and artificial intelligence. Although Bayesian Neural Networks (BNNs) allow for uncertainty in the predictions to be assessed, different sources of predictive uncertainty cannot be distinguished properly. We present Credal Bayesian Deep Learning (CBDL). Heuristically, CBDL allows to train an (uncountably) infinite ensemble of BNNs, using only finitely many elements. This is possible thanks to prior and likelihood finitely generated credal sets (FGCSs), a concept from the imprecise probability literature. Intuitively, convex combinations of a finite collection of prior-likelihood pairs are able to represent infinitely many such pairs. After training, CBDL outputs a set of posteriors on the parameters of the neural network. At inference time, such posterior set is used to derive a set of predictive distributions that is in turn utilized to distinguish between (predictive) aleatoric and epistemic uncertainties, and to quantify them. The predictive set also produces either (i) a collection of outputs enjoying desirable probabilistic guarantees, or (ii) the single output that is deemed the best, that is, the one having the highest predictive lower probability -- another imprecise-probabilistic concept. CBDL is more robust than single BNNs to prior and likelihood misspecification, and to distribution shift. We show that CBDL is better at quantifying and disentangling different types of (predictive) uncertainties than single BNNs and ensemble of BNNs. In addition, we apply CBDL to two case studies to demonstrate its downstream tasks capabilities: one, for motion prediction in autonomous driving scenarios, and two, to model blood glucose and insulin dynamics for artificial pancreas control. We show that CBDL performs better when compared to an ensemble of BNNs baseline.
♻ ☆ Stochastic Gradient Descent for Nonparametric Regression
This paper introduces an iterative algorithm for training nonparametric additive models that enjoys favorable memory storage and computational requirements. The algorithm can be viewed as the functional counterpart of stochastic gradient descent, applied to the coefficients of a truncated basis expansion of the component functions. We show that the resulting estimator satisfies an oracle inequality that allows for model mis-specification. In the well-specified setting, by choosing the learning rate carefully across three distinct stages of training, we demonstrate that its risk is minimax optimal in terms of the dependence on the dimensionality of the data and the size of the training sample. We also provide polynomial convergence rates even when the covariates do not have full support on their domain.
♻ ☆ ReCAP: Recursive Cross Attention Network for Pseudo-Label Generation in Robotic Surgical Skill Assessment
In surgical skill assessment, Objective Structured Assessments of Technical Skills (OSATS scores) and the Global Rating Scale (GRS) are established tools for evaluating the performance of surgeons during training. These metrics, coupled with feedback on their performance, enable surgeons to improve and achieve standards of practice. Recent studies on the open-source dataset JIGSAW, which contains both GRS and OSATS labels, have focused on regressing GRS scores from kinematic signals, video data, or a combination of both. In this paper, we argue that regressing the GRS score, a unitless value, by itself is too restrictive, and variations throughout the surgical trial do not hold significant clinical meaning. To address this gap, we developed a recurrent transformer model that outputs the surgeon's performance throughout their training session by relating the model's hidden states to five OSATS scores derived from kinematic signals. These scores are averaged and aggregated to produce a GRS prediction, enabling assessment of the model's performance against the state-of-the-art (SOTA). We report Spearman's Correlation Coefficient (SCC), demonstrating that our model outperforms SOTA models for all tasks, except for Suturing under the leave-one-subject-out (LOSO) scheme (SCC 0.68-0.89), while achieving comparable performance for suturing and across tasks under the leave-one-user-out (LOUO) scheme (SCC 0.45-0.68) and beating SOTA for Needle Passing (0.69). We argue that relating final OSATS scores to short instances throughout a surgeon's procedure is more clinically meaningful than a single GRS score. This approach also allows us to translate quantitative predictions into qualitative feedback, which is crucial for any automated surgical skill assessment pipeline. A senior surgeon validated our model's behaviour and agreed with the semi-supervised predictions 77 \% (p = 0.006) of the time.
♻ ☆ Towards Enhancing the Reproducibility of Deep Learning Bugs: An Empirical Study
Context: Deep learning has achieved remarkable progress in various domains. However, like any software system, deep learning systems contain bugs, some of which can have severe impacts, as evidenced by crashes involving autonomous vehicles. Despite substantial advancements in deep learning techniques, little research has focused on reproducing deep learning bugs, which is an essential step for their resolution. Existing literature suggests that only 3% of deep learning bugs are reproducible, underscoring the need for further research. Objective: This paper examines the reproducibility of deep learning bugs. We identify edit actions and useful information that could improve the reproducibility of deep learning bugs. Method: First, we construct a dataset of 668 deep-learning bugs from Stack Overflow and GitHub across three frameworks and 22 architectures. Second, out of the 668 bugs, we select 165 bugs using stratified sampling and attempt to determine their reproducibility. While reproducing these bugs, we identify edit actions and useful information for their reproduction. Third, we used the Apriori algorithm to identify useful information and edit actions required to reproduce specific types of bugs. Finally, we conducted a user study involving 22 developers to assess the effectiveness of our findings in real-life settings. Results: We successfully reproduced 148 out of 165 bugs attempted. We identified ten edit actions and five useful types of component information that can help us reproduce the deep learning bugs. With the help of our findings, the developers were able to reproduce 22.92% more bugs and reduce their reproduction time by 24.35%. Conclusions: Our research addresses the critical issue of deep learning bug reproducibility. Practitioners and researchers can leverage our findings to improve deep learning bug reproducibility.
comment: Accepted at the Journal of Empirical Software Engineering (EMSE)
♻ ☆ WavTokenizer: an Efficient Acoustic Discrete Codec Tokenizer for Audio Language Modeling
Language models have been effectively applied to modeling natural signals, such as images, video, speech, and audio. A crucial component of these models is the codec tokenizer, which compresses high-dimensional natural signals into lower-dimensional discrete tokens. In this paper, we introduce WavTokenizer, which offers several advantages over previous SOTA acoustic codec models in the audio domain: 1)extreme compression. By compressing the layers of quantizers and the temporal dimension of the discrete codec, one-second audio of 24kHz sampling rate requires only a single quantizer with 40 or 75 tokens. 2)improved subjective quality. Despite the reduced number of tokens, WavTokenizer achieves state-of-the-art reconstruction quality with outstanding UTMOS scores and inherently contains richer semantic information. Specifically, we achieve these results by designing a broader VQ space, extended contextual windows, and improved attention networks, as well as introducing a powerful multi-scale discriminator and an inverse Fourier transform structure. We conducted extensive reconstruction experiments in the domains of speech, audio, and music. WavTokenizer exhibited strong performance across various objective and subjective metrics compared to state-of-the-art models. We also tested semantic information, VQ utilization, and adaptability to generative models. Comprehensive ablation studies confirm the necessity of each module in WavTokenizer. The related code, demos, and pre-trained models are available at https://github.com/jishengpeng/WavTokenizer.
comment: Working in progress
♻ ☆ Online Tensor Learning: Computational and Statistical Trade-offs, Adaptivity and Optimal Regret
Large tensor learning algorithms are typically computationally expensive and require storing a vast amount of data. In this paper, we propose a unified online Riemannian gradient descent (oRGrad) algorithm for tensor learning, which is computationally efficient, consumes much less memory, and can handle sequentially arriving data while making timely predictions. The algorithm is applicable to both linear and generalized linear models. If the time horizon T is known, oRGrad achieves statistical optimality by choosing an appropriate fixed step size. We find that noisy tensor completion particularly benefits from online algorithms by avoiding the trimming procedure and ensuring sharp entry-wise statistical error, which is often technically challenging for offline methods. The regret of oRGrad is analyzed, revealing a fascinating trilemma concerning the computational convergence rate, statistical error, and regret bound. By selecting an appropriate constant step size, oRGrad achieves an $O(T^{1/2})$ regret. We then introduce the adaptive-oRGrad algorithm, which can achieve the optimal $O(\log T)$ regret by adaptively selecting step sizes, regardless of whether the time horizon is known. The adaptive-oRGrad algorithm can attain a statistically optimal error rate without knowing the horizon. Comprehensive numerical simulations corroborate our theoretical findings. We show that oRGrad significantly outperforms its offline counterpart in predicting the solar F10.7 index with tensor predictors that monitor space weather impacts.
comment: Add initialization algorithms and new application
♻ ☆ RectifID: Personalizing Rectified Flow with Anchored Classifier Guidance NeurIPS 2024
Customizing diffusion models to generate identity-preserving images from user-provided reference images is an intriguing new problem. The prevalent approaches typically require training on extensive domain-specific images to achieve identity preservation, which lacks flexibility across different use cases. To address this issue, we exploit classifier guidance, a training-free technique that steers diffusion models using an existing classifier, for personalized image generation. Our study shows that based on a recent rectified flow framework, the major limitation of vanilla classifier guidance in requiring a special classifier can be resolved with a simple fixed-point solution, allowing flexible personalization with off-the-shelf image discriminators. Moreover, its solving procedure proves to be stable when anchored to a reference flow trajectory, with a convergence guarantee. The derived method is implemented on rectified flow with different off-the-shelf image discriminators, delivering advantageous personalization results for human faces, live subjects, and certain objects. Code is available at https://github.com/feifeiobama/RectifID.
comment: NeurIPS 2024
♻ ☆ Agent-driven Generative Semantic Communication with Cross-Modality and Prediction
In the era of 6G, with compelling visions of intelligent transportation systems and digital twins, remote surveillance is poised to become a ubiquitous practice. Substantial data volume and frequent updates present challenges in wireless networks. To address these challenges, we propose a novel agent-driven generative semantic communication (A-GSC) framework based on reinforcement learning. In contrast to the existing research on semantic communication (SemCom), which mainly focuses on either semantic extraction or semantic sampling, we seamlessly integrate both by jointly considering the intrinsic attributes of source information and the contextual information regarding the task. Notably, the introduction of generative artificial intelligence (GAI) enables the independent design of semantic encoders and decoders. In this work, we develop an agent-assisted semantic encoder with cross-modality capability, which can track the semantic changes, channel condition, to perform adaptive semantic extraction and sampling. Accordingly, we design a semantic decoder with both predictive and generative capabilities, consisting of two tailored modules. Moreover, the effectiveness of the designed models has been verified using the UA-DETRAC dataset, demonstrating the performance gains of the overall A-GSC framework in both energy saving and reconstruction accuracy.
♻ ☆ Developing a Thailand solar irradiance map using Himawari-8 satellite imageries and deep learning models
This paper presents an online platform that shows Thailand's solar irradiance map every 30 minutes. It is available at https://www.cusolarforecast.com. The methodology for estimating global horizontal irradiance (GHI) across Thailand relies on cloud index extracted from Himawari-8 satellite imagery, Ineichen clear-sky model with locally-tuned Linke turbidity, and machine learning models. The methods take clear-sky irradiance, cloud index, re-analyzed GHI and temperature data from the MERRA-2 database, and date-time as inputs for GHI estimation models, including LightGBM, LSTM, Informer, and Transformer. These are benchmarked with the estimate from a commercial service X by evaluating 15-minute ground GHI data from 53 ground stations over 1.5 years from 2022-2023. The results show that the four models have competitive performances and outperform the service X. The best model is LightGBM, with an MAE of 78.58 W/sqm and RMSE of 118.97 W/sqm. Obtaining re-analyzed MERRA-2 data for Thailand is not economically feasible for deployment. When removing these features, the Informer model has a winning performance of 78.67 W/sqm in MAE. The obtained performance aligns with existing literature by taking the climate zone and time granularity of data into consideration. As the map shows an estimate of GHI over 93,000 grids with a frequent update, the paper also describes a computational framework for displaying the entire map. It tests the runtime performance of deep learning models in the GHI estimation process.
comment: 23 pages, 14 figures
♻ ☆ Moonshine: Speech Recognition for Live Transcription and Voice Commands
This paper introduces Moonshine, a family of speech recognition models optimized for live transcription and voice command processing. Moonshine is based on an encoder-decoder transformer architecture and employs Rotary Position Embedding (RoPE) instead of traditional absolute position embeddings. The model is trained on speech segments of various lengths, but without using zero-padding, leading to greater efficiency for the encoder during inference time. When benchmarked against OpenAI's Whisper tiny-en, Moonshine Tiny demonstrates a 5x reduction in compute requirements for transcribing a 10-second speech segment while incurring no increase in word error rates across standard evaluation datasets. These results highlight Moonshine's potential for real-time and resource-constrained applications.
comment: 7 pages, 6 figures, 3 tables
♻ ☆ Temporal Graph Rewiring with Expander Graphs
Evolving relations in real-world networks are often modelled by temporal graphs. Temporal Graph Neural Networks (TGNNs) emerged to model evolutionary behaviour of such graphs by leveraging the message passing primitive at the core of Graph Neural Networks (GNNs). It is well-known that GNNs are vulnerable to several issues directly related to the input graph topology, such as under-reaching and over-squashing - we argue that these issues can often get exacerbated in temporal graphs, particularly as the result of stale nodes and edges. While graph rewiring techniques have seen frequent usage in GNNs to make the graph topology more favourable for message passing, they have not seen any mainstream usage on TGNNs. In this work, we propose Temporal Graph Rewiring (TGR), the first approach for graph rewiring on temporal graphs, to the best of our knowledge. TGR constructs message passing highways between temporally distant nodes in a continuous-time dynamic graph by utilizing expander graph propagation, a prominent framework used for graph rewiring on static graphs which makes minimal assumptions on the underlying graph structure. On the challenging TGB benchmark, TGR achieves state-of-the-art results on tgbl-review, tgbl-coin, tgbl-comment and tgbl-flight datasets at the time of writing. For tgbl-review, TGR has 50.5% improvement in MRR over the base TGN model and 22.2% improvement over the base TNCN model. The significant improvement over base models demonstrates clear benefits of temporal graph rewiring.
comment: 14 pages, 2 figures
♻ ☆ Causal Fairness under Unobserved Confounding: A Neural Sensitivity Framework
Fairness for machine learning predictions is widely required in practice for legal, ethical, and societal reasons. Existing work typically focuses on settings without unobserved confounding, even though unobserved confounding can lead to severe violations of causal fairness and, thus, unfair predictions. In this work, we analyze the sensitivity of causal fairness to unobserved confounding. Our contributions are three-fold. First, we derive bounds for causal fairness metrics under different sources of unobserved confounding. This enables practitioners to examine the sensitivity of their machine learning models to unobserved confounding in fairness-critical applications. Second, we propose a novel neural framework for learning fair predictions, which allows us to offer worst-case guarantees of the extent to which causal fairness can be violated due to unobserved confounding. Third, we demonstrate the effectiveness of our framework in a series of experiments, including a real-world case study about predicting prison sentences. To the best of our knowledge, ours is the first work to study causal fairness under unobserved confounding. To this end, our work is of direct practical value as a refutation strategy to ensure the fairness of predictions in high-stakes applications.
♻ ☆ PRIMER: Perception-Aware Robust Learning-based Multiagent Trajectory Planner
In decentralized multiagent trajectory planners, agents need to communicate and exchange their positions to generate collision-free trajectories. However, due to localization errors/uncertainties, trajectory deconfliction can fail even if trajectories are perfectly shared between agents. To address this issue, we first present PARM and PARM*, perception-aware, decentralized, asynchronous multiagent trajectory planners that enable a team of agents to navigate uncertain environments while deconflicting trajectories and avoiding obstacles using perception information. PARM* differs from PARM as it is less conservative, using more computation to find closer-to-optimal solutions. While these methods achieve state-of-the-art performance, they suffer from high computational costs as they need to solve large optimization problems onboard, making it difficult for agents to replan at high rates. To overcome this challenge, we present our second key contribution, PRIMER, a learning-based planner trained with imitation learning (IL) using PARM* as the expert demonstrator. PRIMER leverages the low computational requirements at deployment of neural networks and achieves a computation speed up to 5500 times faster than optimization-based approaches.
comment: 7 pages, 3 figures
♻ ☆ Thinking Forward: Memory-Efficient Federated Finetuning of Language Models NeurIPS 2024
Finetuning large language models (LLMs) in federated learning (FL) settings has become increasingly important as it allows resource-constrained devices to finetune a model using private data. However, finetuning LLMs using backpropagation requires excessive memory (especially from intermediate activations) for resource-constrained devices. While Forward-mode Auto-Differentiation (AD) can significantly reduce memory footprint from activations, we observe that directly applying it to LLM finetuning results in slow convergence and poor accuracy. In this paper, we introduce Spry, an FL algorithm that splits trainable weights of an LLM among participating clients, such that each client computes gradients using forward-mode AD that are closer estimations of the true gradients. Spry achieves a low memory footprint, high accuracy, and fast convergence. We formally prove that the global gradients in Spry are unbiased estimators of true global gradients for homogeneous data distributions across clients, while heterogeneity increases bias of the estimates. We also derive Spry's convergence rate, showing that the gradients decrease inversely proportional to the number of FL rounds, indicating the convergence up to the limits of heterogeneity. Empirically, Spry reduces the memory footprint during training by 1.4-7.1x in contrast to backpropagation, while reaching comparable accuracy, across a wide range of language tasks, models, and FL settings. Spry reduces the convergence time by 1.2-20.3x and achieves 5.2-13.5% higher accuracy against zero-order methods. When finetuning Llama2-7B with LoRA, compared to the peak memory consumption of 33.9GB of backpropagation, Spry only consumes 6.2GB of peak memory. For OPT13B, the reduction is from 76.5GB to 10.8GB. Spry makes feasible previously impossible FL deployments on commodity edge devices. Our source code is available at https://github.com/Astuary/Spry.
comment: Accepted to NeurIPS 2024
♻ ☆ Slicing Through Bias: Explaining Performance Gaps in Medical Image Analysis using Slice Discovery Methods MICCAI 2024
Machine learning models have achieved high overall accuracy in medical image analysis. However, performance disparities on specific patient groups pose challenges to their clinical utility, safety, and fairness. This can affect known patient groups - such as those based on sex, age, or disease subtype - as well as previously unknown and unlabeled groups. Furthermore, the root cause of such observed performance disparities is often challenging to uncover, hindering mitigation efforts. In this paper, to address these issues, we leverage Slice Discovery Methods (SDMs) to identify interpretable underperforming subsets of data and formulate hypotheses regarding the cause of observed performance disparities. We introduce a novel SDM and apply it in a case study on the classification of pneumothorax and atelectasis from chest x-rays. Our study demonstrates the effectiveness of SDMs in hypothesis formulation and yields an explanation of previously observed but unexplained performance disparities between male and female patients in widely used chest X-ray datasets and models. Our findings indicate shortcut learning in both classification tasks, through the presence of chest drains and ECG wires, respectively. Sex-based differences in the prevalence of these shortcut features appear to cause the observed classification performance gap, representing a previously underappreciated interaction between shortcut learning and model fairness analyses.
comment: MICCAI 2024 Workshop on Fairness of AI in Medical Imaging
♻ ☆ Adversarial Online Collaborative Filtering
We investigate the problem of online collaborative filtering under no-repetition constraints, whereby users need to be served content in an online fashion and a given user cannot be recommended the same content item more than once. We start by designing and analyzing an algorithm that works under biclustering assumptions on the user-item preference matrix, and show that this algorithm exhibits an optimal regret guarantee, while being fully adaptive, in that it is oblivious to any prior knowledge about the sequence of users, the universe of items, as well as the biclustering parameters of the preference matrix. We then propose a more robust version of this algorithm which operates with general matrices. Also this algorithm is parameter free, and we prove regret guarantees that scale with the amount by which the preference matrix deviates from a biclustered structure. To our knowledge, these are the first results on online collaborative filtering that hold at this level of generality and adaptivity under no-repetition constraints. Finally, we complement our theoretical findings with simple experiments on real-world datasets aimed at both validating the theory and empirically comparing to standard baselines. This comparison shows the competitive advantage of our approach over these baselines.
♻ ☆ Transfer Learning Bayesian Optimization to Design Competitor DNA Molecules for Use in Diagnostic Assays
With the rise in engineered biomolecular devices, there is an increased need for tailor-made biological sequences. Often, many similar biological sequences need to be made for a specific application meaning numerous, sometimes prohibitively expensive, lab experiments are necessary for their optimization. This paper presents a transfer learning design of experiments workflow to make this development feasible. By combining a transfer learning surrogate model with Bayesian optimization, we show how the total number of experiments can be reduced by sharing information between optimization tasks. We demonstrate the reduction in the number of experiments using data from the development of DNA competitors for use in an amplification-based diagnostic assay. We use cross-validation to compare the predictive accuracy of different transfer learning models, and then compare the performance of the models for both single objective and penalized optimization tasks.
♻ ☆ LLM Gesticulator: Leveraging Large Language Models for Scalable and Controllable Co-Speech Gesture Synthesis
In this work, we present LLM Gesticulator, an LLM-based audio-driven co-speech gesture generation framework that synthesizes full-body animations that are rhythmically aligned with the input audio while exhibiting natural movements and editability. Compared to previous work, our model demonstrates substantial scalability. As the size of the backbone LLM model increases, our framework shows proportional improvements in evaluation metrics (a.k.a. scaling law). Our method also exhibits strong controllability where the content, style of the generated gestures can be controlled by text prompt. To the best of our knowledge, LLM gesticulator is the first work that use LLM on the co-speech generation task. Evaluation with existing objective metrics and user studies indicate that our framework outperforms prior works.
♻ ☆ Beyond Trading Data: The Hidden Influence of Public Awareness and Interest on Cryptocurrency Volatility CIKM 2023
Since Bitcoin first appeared on the scene in 2009, cryptocurrencies have become a worldwide phenomenon as important decentralized financial assets. Their decentralized nature, however, leads to notable volatility against traditional fiat currencies, making the task of accurately forecasting the crypto-fiat exchange rate complex. This study examines the various independent factors that affect the volatility of the Bitcoin-Dollar exchange rate. To this end, we propose CoMForE, a multimodal AdaBoost-LSTM ensemble model, which not only utilizes historical trading data but also incorporates public sentiments from related tweets, public interest demonstrated by search volumes, and blockchain hash-rate data. Our developed model goes a step further by predicting fluctuations in the overall cryptocurrency value distribution, thus increasing its value for investment decision-making. We have subjected this method to extensive testing via comprehensive experiments, thereby validating the importance of multimodal combination over exclusive reliance on trading data. Further experiments show that our method significantly surpasses existing forecasting tools and methodologies, demonstrating a 19.29% improvement. This result underscores the influence of external independent factors on cryptocurrency volatility.
comment: Published at the 32nd ACM International Conference on Information and Knowledge Management (CIKM 2023)
♻ ☆ Adaptive $Q$-Aid for Conditional Supervised Learning in Offline Reinforcement Learning NeurIPS2024
Offline reinforcement learning (RL) has progressed with return-conditioned supervised learning (RCSL), but its lack of stitching ability remains a limitation. We introduce $Q$-Aided Conditional Supervised Learning (QCS), which effectively combines the stability of RCSL with the stitching capability of $Q$-functions. By analyzing $Q$-function over-generalization, which impairs stable stitching, QCS adaptively integrates $Q$-aid into RCSL's loss function based on trajectory return. Empirical results show that QCS significantly outperforms RCSL and value-based methods, consistently achieving or exceeding the maximum trajectory returns across diverse offline RL benchmarks.
comment: Accepted to NeurIPS2024. The project page is available at https://beanie00.com/publications/qcs
♻ ☆ Targeted Separation and Convergence with Kernel Discrepancies
Maximum mean discrepancies (MMDs) like the kernel Stein discrepancy (KSD) have grown central to a wide range of applications, including hypothesis testing, sampler selection, distribution approximation, and variational inference. In each setting, these kernel-based discrepancy measures are required to (i) separate a target P from other probability measures or even (ii) control weak convergence to P. In this article we derive new sufficient and necessary conditions to ensure (i) and (ii). For MMDs on separable metric spaces, we characterize those kernels that separate Bochner embeddable measures and introduce simple conditions for separating all measures with unbounded kernels and for controlling convergence with bounded kernels. We use these results on $\mathbb{R}^d$ to substantially broaden the known conditions for KSD separation and convergence control and to develop the first KSDs known to exactly metrize weak convergence to P. Along the way, we highlight the implications of our results for hypothesis testing, measuring and improving sample quality, and sampling with Stein variational gradient descent.
♻ ☆ Rethinking Complex Queries on Knowledge Graphs with Neural Link Predictors ICLR 2024
Reasoning on knowledge graphs is a challenging task because it utilizes observed information to predict the missing one. Particularly, answering complex queries based on first-order logic is one of the crucial tasks to verify learning to reason abilities for generalization and composition. Recently, the prevailing method is query embedding which learns the embedding of a set of entities and treats logic operations as set operations and has shown great empirical success. Though there has been much research following the same formulation, many of its claims lack a formal and systematic inspection. In this paper, we rethink this formulation and justify many of the previous claims by characterizing the scope of queries investigated previously and precisely identifying the gap between its formulation and its goal, as well as providing complexity analysis for the currently investigated queries. Moreover, we develop a new dataset containing ten new types of queries with features that have never been considered and therefore can provide a thorough investigation of complex queries. Finally, we propose a new neural-symbolic method, Fuzzy Inference with Truth value (FIT), where we equip the neural link predictors with fuzzy logic theory to support end-to-end learning using complex queries with provable reasoning capability. Empirical results show that our method outperforms previous methods significantly in the new dataset and also surpasses previous methods in the existing dataset at the same time.
comment: Received in ICLR 2024
♻ ☆ ExDBN: Exact learning of Dynamic Bayesian Networks
Causal learning from data has received much attention in recent years. One way of capturing causal relationships is by utilizing Bayesian networks. There, one recovers a weighted directed acyclic graph, in which random variables are represented by vertices, and the weights associated with each edge represent the strengths of the causal relationships between them. This concept is extended to capture dynamic effects by introducing a dependency on past data, which may be captured by the structural equation model, which is utilized in the present contribution to formulate a score-based learning approach. A mixed-integer quadratic program is formulated and an algorithmic solution proposed, in which the pre-generation of exponentially many acyclicity constraints is avoided by utilizing the so-called branch-and-cut ("lazy constraint") method. Comparing the novel approach to the state of the art, we show that the proposed approach turns out to produce excellent results when applied to small and medium-sized synthetic instances of up to 25 time-series. Lastly, two interesting applications in bio-science and finance, to which the method is directly applied, further stress the opportunities in developing highly accurate, globally convergent solvers that can handle modest instances.
comment: 12 pages
♻ ☆ Finite Operator Learning: Bridging Neural Operators and Numerical Methods for Efficient Parametric Solution and Optimization of PDEs
We introduce a method that combines neural operators, physics-informed machine learning, and standard numerical methods for solving PDEs. The proposed approach extends each of the aforementioned methods and unifies them within a single framework. We can parametrically solve partial differential equations in a data-free manner and provide accurate sensitivities, meaning the derivatives of the solution space with respect to the design space. These capabilities enable gradient-based optimization without the typical sensitivity analysis costs, unlike adjoint methods that scale directly with the number of response functions. Our Finite Operator Learning (FOL) approach uses an uncomplicated feed-forward neural network model to directly map the discrete design space (i.e. parametric input space) to the discrete solution space (i.e. finite number of sensor points in the arbitrary shape domain) ensuring compliance with physical laws by designing them into loss functions. The discretized governing equations, as well as the design and solution spaces, can be derived from any well-established numerical techniques. In this work, we employ the Finite Element Method (FEM) to approximate fields and their spatial derivatives. Subsequently, we conduct Sobolev training to minimize a multi-objective loss function, which includes the discretized weak form of the energy functional, boundary conditions violations, and the stationarity of the residuals with respect to the design variables. Our study focuses on the steady-state heat equation within heterogeneous materials that exhibits significant phase contrast and possibly temperature-dependent conductivity. The network's tangent matrix is directly used for gradient-based optimization to improve the microstructure's heat transfer characteristics. ...
comment: arXiv admin note: text overlap with arXiv:2401.02363
♻ ☆ Towards Understanding the Working Mechanism of Text-to-Image Diffusion Model
Recently, the strong latent Diffusion Probabilistic Model (DPM) has been applied to high-quality Text-to-Image (T2I) generation (e.g., Stable Diffusion), by injecting the encoded target text prompt into the gradually denoised diffusion image generator. Despite the success of DPM in practice, the mechanism behind it remains to be explored. To fill this blank, we begin by examining the intermediate statuses during the gradual denoising generation process in DPM. The empirical observations indicate, the shape of image is reconstructed after the first few denoising steps, and then the image is filled with details (e.g., texture). The phenomenon is because the low-frequency signal (shape relevant) of the noisy image is not corrupted until the final stage in the forward process (initial stage of generation) of adding noise in DPM. Inspired by the observations, we proceed to explore the influence of each token in the text prompt during the two stages. After a series of experiments of T2I generations conditioned on a set of text prompts. We conclude that in the earlier generation stage, the image is mostly decided by the special token [\texttt{EOS}] in the text prompt, and the information in the text prompt is already conveyed in this stage. After that, the diffusion model completes the details of generated images by information from themselves. Finally, we propose to apply this observation to accelerate the process of T2I generation by properly removing text guidance, which finally accelerates the sampling up to 25\%+.
♻ ☆ Understanding Linear Probing then Fine-tuning Language Models from NTK Perspective NeurIPS 2024
The two-stage fine-tuning (FT) method, linear probing (LP) then fine-tuning (LP-FT), outperforms linear probing and FT alone. This holds true for both in-distribution (ID) and out-of-distribution (OOD) data. One key reason for its success is the preservation of pre-trained features, achieved by obtaining a near-optimal linear head during LP. However, despite the widespread use of large language models, there has been limited exploration of more complex architectures such as Transformers. In this paper, we analyze the training dynamics of LP-FT for classification tasks on the basis of the neural tangent kernel (NTK) theory. Our analysis decomposes the NTK matrix into two components. This decomposition highlights the importance of the linear head norm alongside the prediction accuracy at the start of the FT stage. We also observe a significant increase in the linear head norm during LP, which stems from training with the cross-entropy (CE) loss. This increase in the linear head norm effectively reduces changes in learned features. Furthermore, we find that this increased norm can adversely affect model calibration, which can be corrected using temperature scaling. Additionally, we extend our analysis with the NTK to the low-rank adaptation (LoRA) method and validate its effectiveness. Our experiments using a Transformer-based model on multiple natural language processing datasets confirm our theoretical analysis. Our study demonstrates the effectiveness of LP-FT for fine-tuning language models. Code is available at https://github.com/tom4649/lp-ft_ntk.
comment: Accepted at NeurIPS 2024
♻ ☆ Mimicking Better by Matching the Approximate Action Distribution
In this paper, we introduce MAAD, a novel, sample-efficient on-policy algorithm for Imitation Learning from Observations. MAAD utilizes a surrogate reward signal, which can be derived from various sources such as adversarial games, trajectory matching objectives, or optimal transport criteria. To compensate for the non-availability of expert actions, we rely on an inverse dynamics model that infers plausible actions distribution given the expert's state-state transitions; we regularize the imitator's policy by aligning it to the inferred action distribution. MAAD leads to significantly improved sample efficiency and stability. We demonstrate its effectiveness in a number of MuJoCo environments, both int the OpenAI Gym and the DeepMind Control Suite. We show that it requires considerable fewer interactions to achieve expert performance, outperforming current state-of-the-art on-policy methods. Remarkably, MAAD often stands out as the sole method capable of attaining expert performance levels, underscoring its simplicity and efficacy.
♻ ☆ Higher-Order Message Passing for Glycan Representation Learning NeurIPS 2024
Glycans are the most complex biological sequence, with monosaccharides forming extended, non-linear sequences. As post-translational modifications, they modulate protein structure, function, and interactions. Due to their diversity and complexity, predictive models of glycan properties and functions are still insufficient. Graph Neural Networks (GNNs) are deep learning models designed to process and analyze graph-structured data. These architectures leverage the connectivity and relational information in graphs to learn effective representations of nodes, edges, and entire graphs. Iteratively aggregating information from neighboring nodes, GNNs capture complex patterns within graph data, making them particularly well-suited for tasks such as link prediction or graph classification across domains. This work presents a new model architecture based on combinatorial complexes and higher-order message passing to extract features from glycan structures into a latent space representation. The architecture is evaluated on an improved GlycanML benchmark suite, establishing a new state-of-the-art performance. We envision that these improvements will spur further advances in computational glycosciences and reveal the roles of glycans in biology.
comment: Accepted to MLSB Workshop at NeurIPS 2024
♻ ☆ Advancing Training Efficiency of Deep Spiking Neural Networks through Rate-based Backpropagation NeurIPS 2024
Recent insights have revealed that rate-coding is a primary form of information representation captured by surrogate-gradient-based Backpropagation Through Time (BPTT) in training deep Spiking Neural Networks (SNNs). Motivated by these findings, we propose rate-based backpropagation, a training strategy specifically designed to exploit rate-based representations to reduce the complexity of BPTT. Our method minimizes reliance on detailed temporal derivatives by focusing on averaged dynamics, streamlining the computational graph to reduce memory and computational demands of SNNs training. We substantiate the rationality of the gradient approximation between BPTT and the proposed method through both theoretical analysis and empirical observations. Comprehensive experiments on CIFAR-10, CIFAR-100, ImageNet, and CIFAR10-DVS validate that our method achieves comparable performance to BPTT counterparts, and surpasses state-of-the-art efficient training techniques. By leveraging the inherent benefits of rate-coding, this work sets the stage for more scalable and efficient SNNs training within resource-constrained environments. Our code is available at https://github.com/Tab-ct/rate-based-backpropagation.
comment: Accepted by NeurIPS 2024
♻ ☆ GLBench: A Comprehensive Benchmark for Graph with Large Language Models
The emergence of large language models (LLMs) has revolutionized the way we interact with graphs, leading to a new paradigm called GraphLLM. Despite the rapid development of GraphLLM methods in recent years, the progress and understanding of this field remain unclear due to the lack of a benchmark with consistent experimental protocols. To bridge this gap, we introduce GLBench, the first comprehensive benchmark for evaluating GraphLLM methods in both supervised and zero-shot scenarios. GLBench provides a fair and thorough evaluation of different categories of GraphLLM methods, along with traditional baselines such as graph neural networks. Through extensive experiments on a collection of real-world datasets with consistent data processing and splitting strategies, we have uncovered several key findings. Firstly, GraphLLM methods outperform traditional baselines in supervised settings, with LLM-as-enhancers showing the most robust performance. However, using LLMs as predictors is less effective and often leads to uncontrollable output issues. We also notice that no clear scaling laws exist for current GraphLLM methods. In addition, both structures and semantics are crucial for effective zero-shot transfer, and our proposed simple baseline can even outperform several models tailored for zero-shot scenarios. The data and code of the benchmark can be found at https://github.com/NineAbyss/GLBench.
♻ ☆ A Historical Trajectory Assisted Optimization Method for Zeroth-Order Federated Learning
Federated learning heavily relies on distributed gradient descent techniques. In the situation where gradient information is not available, the gradients need to be estimated from zeroth-order information, which typically involves computing finite-differences along isotropic random directions. This method suffers from high estimation errors, as the geometric features of the objective landscape may be overlooked during the isotropic sampling. In this work, we propose a non-isotropic sampling method to improve the gradient estimation procedure. Gradients in our method are estimated in a subspace spanned by historical trajectories of solutions, aiming to encourage the exploration of promising regions and hence improve the convergence. The proposed method uses a covariance matrix for sampling which is a convex combination of two parts. The first part is a thin projection matrix containing the basis of the subspace which is designed to improve the exploitation ability. The second part is the historical trajectories. We implement this method in zeroth-order federated settings, and show that the convergence rate aligns with existing ones while introducing no significant overheads in communication or local computation. The effectiveness of our proposal is verified on several numerical experiments in comparison to several commonly-used zeroth-order federated optimization algorithms.
♻ ☆ Point Cloud Matters: Rethinking the Impact of Different Observation Spaces on Robot Learning NeurIPS 2024
In robot learning, the observation space is crucial due to the distinct characteristics of different modalities, which can potentially become a bottleneck alongside policy design. In this study, we explore the influence of various observation spaces on robot learning, focusing on three predominant modalities: RGB, RGB-D, and point cloud. We introduce OBSBench, a benchmark comprising two simulators and 125 tasks, along with standardized pipelines for various encoders and policy baselines. Extensive experiments on diverse contact-rich manipulation tasks reveal a notable trend: point cloud-based methods, even those with the simplest designs, frequently outperform their RGB and RGB-D counterparts. This trend persists in both scenarios: training from scratch and utilizing pre-training. Furthermore, our findings demonstrate that point cloud observations often yield better policy performance and significantly stronger generalization capabilities across various geometric and visual conditions. These outcomes suggest that the 3D point cloud is a valuable observation modality for intricate robotic tasks. We also suggest that incorporating both appearance and coordinate information can enhance the performance of point cloud methods. We hope our work provides valuable insights and guidance for designing more generalizable and robust robotic models. Codes are available at https://github.com/HaoyiZhu/PointCloudMatters.
comment: 38th Conference on Neural Information Processing Systems (NeurIPS 2024) Track on Datasets and Benchmarks
♻ ☆ CartesianMoE: Boosting Knowledge Sharing among Experts via Cartesian Product Routing in Mixture-of-Experts
Large language models (LLM) have been attracting much attention from the community recently, due to their remarkable performance in all kinds of downstream tasks. According to the well-known scaling law, scaling up a dense LLM enhances its capabilities, but also significantly increases the computational complexity. Mixture-of-Experts (MoE) models address that by allowing the model size to grow without substantially raising training or inference costs. Yet MoE models face challenges regarding knowledge sharing among experts, making their performance somehow sensitive to routing accuracy. To tackle that, previous works introduced shared experts and combined their outputs with those of the top $K$ routed experts in an ``addition'' manner. In this paper, inspired by collective matrix factorization to learn shared knowledge among data, we propose CartesianMoE, which implements more effective knowledge sharing among experts in more like a ``multiplication'' manner. Extensive experimental results indicate that CartesianMoE outperforms previous MoE models for building LLMs, in terms of both perplexity and downstream task performance. And we also find that CartesianMoE achieves better expert routing robustness.
♻ ☆ Critical Phase Transition in Large Language Models
Large Language Models (LLMs) have demonstrated impressive performance. To understand their behaviors, we need to consider the fact that LLMs sometimes show qualitative changes. The natural world also presents such changes called phase transitions, which are defined by singular, divergent statistical quantities. Therefore, an intriguing question is whether qualitative changes in LLMs are phase transitions. In this work, we have conducted extensive analysis on texts generated by LLMs and suggested that a phase transition occurs in LLMs when varying the temperature parameter. Specifically, statistical quantities have divergent properties just at the point between the low-temperature regime, where LLMs generate sentences with clear repetitive structures, and the high-temperature regime, where generated sentences are often incomprehensible. In addition, critical behaviors near the phase transition point, such as a power-law decay of correlation and slow convergence toward the stationary state, are similar to those in natural languages. Our results suggest a meaningful analogy between LLMs and natural phenomena.
comment: 10 pages, 6 figures
♻ ☆ Knowledge Distillation-Based Model Extraction Attack using GAN-based Private Counterfactual Explanations
In recent years, there has been a notable increase in the deployment of machine learning (ML) models as services (MLaaS) across diverse production software applications. In parallel, explainable AI (XAI) continues to evolve, addressing the necessity for transparency and trustworthiness in ML models. XAI techniques aim to enhance the transparency of ML models by providing insights, in terms of model's explanations, into their decision-making process. Simultaneously, some MLaaS platforms now offer explanations alongside the ML prediction outputs. This setup has elevated concerns regarding vulnerabilities in MLaaS, particularly in relation to privacy leakage attacks such as model extraction attacks (MEA). This is due to the fact that explanations can unveil insights about the inner workings of the model which could be exploited by malicious users. In this work, we focus on investigating how model explanations, particularly counterfactual explanations (CFs), can be exploited for performing MEA within the MLaaS platform. We also delve into assessing the effectiveness of incorporating differential privacy (DP) as a mitigation strategy. To this end, we first propose a novel approach for MEA based on Knowledge Distillation (KD) to enhance the efficiency of extracting a substitute model of a target model exploiting CFs, without any knowledge about the training data distribution by the attacker. Then, we advise an approach for training CF generators incorporating DP to generate private CFs. We conduct thorough experimental evaluations on real-world datasets and demonstrate that our proposed KD-based MEA can yield a high-fidelity substitute model with a reduced number of queries with respect to baseline approaches. Furthermore, our findings reveal that including a privacy layer can allow mitigating the MEA. However, on the account of the quality of CFs, impacts the performance of the explanations.
comment: 19 pages
♻ ☆ A Self-Organizing Clustering System for Unsupervised Distribution Shift Detection IJCNN'2024
Modeling non-stationary data is a challenging problem in the field of continual learning, and data distribution shifts may result in negative consequences on the performance of a machine learning model. Classic learning tools are often vulnerable to perturbations of the input covariates, and are sensitive to outliers and noise, and some tools are based on rigid algebraic assumptions. Distribution shifts are frequently occurring due to changes in raw materials for production, seasonality, a different user base, or even adversarial attacks. Therefore, there is a need for more effective distribution shift detection techniques. In this work, we propose a continual learning framework for monitoring and detecting distribution changes. We explore the problem in a latent space generated by a bio-inspired self-organizing clustering and statistical aspects of the latent space. In particular, we investigate the projections made by two topology-preserving maps: the Self-Organizing Map and the Scale Invariant Map. Our method can be applied in both a supervised and an unsupervised context. We construct the assessment of changes in the data distribution as a comparison of Gaussian signals, making the proposed method fast and robust. We compare it to other unsupervised techniques, specifically Principal Component Analysis (PCA) and Kernel-PCA. Our comparison involves conducting experiments using sequences of images (based on MNIST and injected shifts with adversarial samples), chemical sensor measurements, and the environmental variable related to ozone levels. The empirical study reveals the potential of the proposed approach.
comment: Revised version of the accepted manuscript to IJCNN'2024. Main corrections were in Section 2.2 and Section 3.3. In Section 2.2 was corrected expression (3), and in Section 3.3 in the definition of the elements of the matrix $D$ it was a typo where $\phi(x)$ was written instead of $x$
♻ ☆ Online Structured Prediction with Fenchel--Young Losses and Improved Surrogate Regret for Online Multiclass Classification with Logistic Loss
This paper studies online structured prediction with full-information feedback. For online multiclass classification, Van der Hoeven (2020) established \emph{finite} surrogate regret bounds, which are independent of the time horizon, by introducing an elegant \emph{exploit-the-surrogate-gap} framework. However, this framework has been limited to multiclass classification primarily because it relies on a classification-specific procedure for converting estimated scores to outputs. We extend the exploit-the-surrogate-gap framework to online structured prediction with \emph{Fenchel--Young losses}, a large family of surrogate losses that includes the logistic loss for multiclass classification as a special case, obtaining finite surrogate regret bounds in various structured prediction problems. To this end, we propose and analyze \emph{randomized decoding}, which converts estimated scores to general structured outputs. Moreover, by applying our decoding to online multiclass classification with the logistic loss, we obtain a surrogate regret bound of $O(\| \mathbf{U} \|_\mathrm{F}^2)$, where $\mathbf{U}$ is the best offline linear estimator and $\| \cdot \|_\mathrm{F}$ denotes the Frobenius norm. This bound is tight up to logarithmic factors and improves the previous bound of $O(d\| \mathbf{U} \|_\mathrm{F}^2)$ due to Van der Hoeven (2020) by a factor of $d$, the number of classes.
♻ ☆ Enhancing Algorithm Performance Understanding through tsMorph: Generating Semi-Synthetic Time Series for Robust Forecasting Evaluation
Time series forecasting is a subject of significant scientific and industrial importance. Despite the widespread utilization of forecasting methods, there is a dearth of research aimed at comprehending the conditions under which these methods yield favorable or unfavorable performances. Empirical studies, although common, are challenged by the limited availability of time series datasets, restricting the extraction of reliable insights. To address this limitation, we present tsMorph, a tool for generating semi-synthetic time series through dataset morphing. tsMorph works by creating a sequence of datasets from two original datasets. The characteristics of the generated datasets progressively depart from those of one of the datasets and converge toward the attributes of the other dataset. This method provides a valuable alternative for obtaining substantial datasets. In this paper, we show the benefits of tsMorph by assessing the predictive performance of the Long Short-Term Memory Network and DeepAR forecasting algorithms. The time series used for the experiments comes from the NN5 Competition. The experimental results provide important insights. Notably, the performances of the two algorithms improve proportionally with the frequency of the time series. These experiments confirm that tsMorph can be an effective tool for better understanding the behavior of forecasting algorithms, delivering a pathway to overcoming the limitations posed by empirical studies and enabling more extensive and reliable experiments.
♻ ☆ PLaMo-100B: A Ground-Up Language Model Designed for Japanese Proficiency
We introduce PLaMo-100B, a large-scale language model designed for Japanese proficiency. The model was trained from scratch using 2 trillion tokens, with architecture such as QK Normalization and Z-Loss to ensure training stability during the training process. Post-training techniques, including Supervised Fine-Tuning and Direct Preference Optimization, were applied to refine the model's performance. Benchmark evaluations suggest that PLaMo-100B performs well, particularly in Japanese-specific tasks, achieving results that are competitive with frontier models like GPT-4. The base model is available at https://huggingface.co/pfnet/plamo-100b.
♻ ☆ Multimodal hierarchical Variational AutoEncoders with Factor Analysis latent space
Purpose: Handling heterogeneous and mixed data types has become increasingly critical with the exponential growth in real-world databases. While deep generative models attempt to merge diverse data views into a common latent space, they often sacrifice interpretability, flexibility, and modularity. This study proposes a novel method to address these limitations by combining Variational AutoEncoders (VAEs) with a Factor Analysis latent space (FA-VAE). Methods: The proposed FA-VAE method employs multiple VAEs to learn a private representation for each heterogeneous data view in a continuous latent space. Information is shared between views using a low-dimensional latent space, generated via a linear projection matrix. This modular design creates a hierarchical dependency between private and shared latent spaces, allowing for the flexible addition of new views and conditioning of pre-trained models. Results: The FA-VAE approach facilitates cross-generation of data from different domains and enables transfer learning between generative models. This allows for effective integration of information across diverse data views while preserving their distinct characteristics. Conclusions: By overcoming the limitations of existing methods, the FA-VAE provides a more interpretable, flexible, and modular solution for managing heterogeneous data types. It offers a pathway to more efficient and scalable data-handling strategies, enhancing the potential for cross-domain data synthesis and model transferability.
comment: 21 pages main work, 2 pages supplementary, 14 figures
♻ ☆ Mesa-Extrapolation: A Weave Position Encoding Method for Enhanced Extrapolation in LLMs NeurIPS 2024
Large language models (LLMs), although having revolutionized many fields, still suffer from the challenging extrapolation problem, where the inference ability of LLMs sharply declines beyond their max training lengths. In this work, we conduct a theoretical analysis to better understand why No Position Encoding (NoPE) fails outside its effective range, as well as examining the power of Position Encoding (PE) in this context. Our findings reveal that with meticulous weave position, PE can indeed be extended beyond effective range. Our theorems establish that LLMs equipped with weave PE can achieve improved extrapolation performance without additional cost. Furthermore, we introduce a novel weave PE method, Mesa-Extrapolation, which utilizes a chunk-based triangular attention matrix and applies Stair PE to manage the final chunk. This method not only retains competitive performance but also offers substantial benefits such as significantly reduced memory demand and faster inference speed. Extensive experiments validate the effectiveness of Mesa-Extrapolation, demonstrating its potential as a scalable solution to enhancing LLMs applicative reach.
comment: Accepted by NeurIPS 2024; 13 pages and 30 pages appendix
♻ ☆ A Lightweight Modular Framework for Low-Cost Open-Vocabulary Object Detection Training
Object detection is a fundamental challenge in computer vision, centered on recognizing objects within images, with diverse applications in areas like image analysis, robotics, and autonomous vehicles. Although existing methods have achieved great success, they are often constrained by a fixed vocabulary of objects. To overcome this limitation, approaches like MDETR have redefined object detection by incorporating region-level vision-language pre-training, enabling open-vocabulary object detectors. However, these methods are computationally heavy due to the simultaneous training of large models for both vision and language representations. To address this, we introduce a lightweight framework that significantly reduces the number of parameters while preserving, or even improving, performance. Our solution is applied to MDETR, resulting in the development of Lightweight MDETR (LightMDETR), an optimized version of MDETR designed to enhance computational efficiency without sacrificing accuracy. The core of our approach involves freezing the MDETR backbone and training only the Universal Projection module (UP), which bridges vision and language representations. A learnable modality token parameter allows the UP to seamlessly switch between modalities. Evaluations on tasks like phrase grounding, referring expression comprehension, and segmentation show that LightMDETR not only reduces computational costs but also outperforms several state-of-the-art methods in terms of accuracy.
♻ ☆ TensorProjection Layer: A Tensor-Based Dimension Reduction Method in Deep Neural Networks
In this paper, we propose a dimension reduction method specifically designed for tensor-structured feature data in deep neural networks. The method is implemented as a hidden layer, called the TensorProjection layer, which transforms input tensors into output tensors with reduced dimensions through mode-wise projections. The projection directions are treated as model parameters of the layer and are optimized during model training. Our method can serve as an alternative to pooling layers for summarizing image data, or to convolutional layers as a technique for reducing the number of channels. We conduct experiments on tasks such as medical image classification and segmentation, integrating the TensorProjection layer into commonly used baseline architectures to evaluate its effectiveness. Numerical experiments indicate that the proposed method can outperform traditional downsampling methods, such as pooling layers, in our tasks, suggesting it as a promising alternative for feature summarization.
♻ ☆ Unsupervised Assessment of Landscape Shifts Based on Persistent Entropy and Topological Preservation KDD'2024
In Continual Learning (CL) contexts, concept drift typically refers to the analysis of changes in data distribution. A drift in the input data can have negative consequences on a learning predictor and the system's stability. The majority of concept drift methods emphasize the analysis of statistical changes in non-stationary data over time. In this context, we consider another perspective, where the concept drift also integrates substantial changes in the topological characteristics of the data stream. In this article, we introduce a novel framework for monitoring changes in multi-dimensional data streams. We explore variations in the topological structures of the data, presenting another angle on the standard concept drift. Our developed approach is based on persistent entropy and topology-preserving projections in a continual learning scenario. The framework operates in both unsupervised and supervised environments. To show the utility of the proposed framework, we analyze the model across three scenarios using data streams generated with MNIST samples. The obtained results reveal the potential of applying topological data analysis for shift detection and encourage further research in this area.
comment: KDD'2024. Workshop on Drift Detection and Landscape Shifts
♻ ☆ Exploring Edge Probability Graph Models Beyond Edge Independency: Concepts, Analyses, and Algorithms
Desirable random graph models (RGMs) should (i) generate realistic structures such as high clustering (i.e., high subgraph densities), (ii) generate variable (i.e., not overly similar) graphs, and (iii) remain tractable to compute and control graph statistics. A common class of RGMs (e.g., Erd\H{o}s-R'{e}nyi and stochastic Kronecker) outputs edge probabilities, and we need to realize (i.e., sample from) the edge probabilities to generate graphs. Typically, each edge's existence is assumed to be determined independently for simplicity and tractability. However, with edge independency, RGMs theoretically cannot produce high subgraph densities and high output variability simultaneously. In this work, we explore realization beyond edge independence that can produce more realistic structures while maintaining high traceability and variability. Theoretically, we propose an edge-dependent realization framework called binding that provably preserves output variability, and derive closed-form tractability results on subgraph (e.g., triangle) densities in generated graphs. Practically, we propose algorithms for graph generation with binding and parameter fitting of binding. Our empirical results demonstrate that binding exhibits high tractability and generates realistic graphs with high clustering, significantly improving upon existing RGMs assuming edge independency.
♻ ☆ Scaling transformer neural networks for skillful and reliable medium-range weather forecasting NeurIPS 2024
Weather forecasting is a fundamental problem for anticipating and mitigating the impacts of climate change. Recently, data-driven approaches for weather forecasting based on deep learning have shown great promise, achieving accuracies that are competitive with operational systems. However, those methods often employ complex, customized architectures without sufficient ablation analysis, making it difficult to understand what truly contributes to their success. Here we introduce Stormer, a simple transformer model that achieves state-of-the-art performance on weather forecasting with minimal changes to the standard transformer backbone. We identify the key components of Stormer through careful empirical analyses, including weather-specific embedding, randomized dynamics forecast, and pressure-weighted loss. At the core of Stormer is a randomized forecasting objective that trains the model to forecast the weather dynamics over varying time intervals. During inference, this allows us to produce multiple forecasts for a target lead time and combine them to obtain better forecast accuracy. On WeatherBench 2, Stormer performs competitively at short to medium-range forecasts and outperforms current methods beyond 7 days, while requiring orders-of-magnitude less training data and compute. Additionally, we demonstrate Stormer's favorable scaling properties, showing consistent improvements in forecast accuracy with increases in model size and training tokens. Code and checkpoints are available at https://github.com/tung-nd/stormer.
comment: Neural Information Processing Systems (NeurIPS 2024)
♻ ☆ Improving Reinforcement Learning from Human Feedback with Efficient Reward Model Ensemble
Reinforcement Learning from Human Feedback (RLHF) is a widely adopted approach for aligning large language models with human values. However, RLHF relies on a reward model that is trained with a limited amount of human preference data, which could lead to inaccurate predictions. As a result, RLHF may produce outputs that are misaligned with human values. To mitigate this issue, we contribute a reward ensemble method that allows the reward model to make more accurate predictions. As using an ensemble of large language model-based reward models can be computationally and resource-expensive, we explore efficient ensemble methods including linear-layer ensemble and LoRA-based ensemble. Empirically, we run Best-of-$n$ and Proximal Policy Optimization with our ensembled reward models, and verify that our ensemble methods help improve the alignment performance of RLHF outputs.
♻ ☆ Position Engineering: Boosting Large Language Models through Positional Information Manipulation
The performance of large language models (LLMs) is significantly influenced by the quality of the prompts provided. In response, researchers have developed enormous prompt engineering strategies aimed at modifying the prompt text to enhance task performance. In this paper, we introduce a novel technique termed position engineering, which offers a more efficient way to guide large language models. Unlike prompt engineering, which requires substantial effort to modify the text provided to LLMs, position engineering merely involves altering the positional information in the prompt without modifying the text itself. We have evaluated position engineering in two widely-used LLM scenarios: retrieval-augmented generation (RAG) and in-context learning (ICL). Our findings show that position engineering substantially improves upon the baseline in both cases. Position engineering thus represents a promising new strategy for exploiting the capabilities of large language models.
♻ ☆ Learning to Denoise Biomedical Knowledge Graph for Robust Molecular Interaction Prediction
Molecular interaction prediction plays a crucial role in forecasting unknown interactions between molecules, such as drug-target interaction (DTI) and drug-drug interaction (DDI), which are essential in the field of drug discovery and therapeutics. Although previous prediction methods have yielded promising results by leveraging the rich semantics and topological structure of biomedical knowledge graphs (KGs), they have primarily focused on enhancing predictive performance without addressing the presence of inevitable noise and inconsistent semantics. This limitation has hindered the advancement of KG-based prediction methods. To address this limitation, we propose BioKDN (Biomedical Knowledge Graph Denoising Network) for robust molecular interaction prediction. BioKDN refines the reliable structure of local subgraphs by denoising noisy links in a learnable manner, providing a general module for extracting task-relevant interactions. To enhance the reliability of the refined structure, BioKDN maintains consistent and robust semantics by smoothing relations around the target interaction. By maximizing the mutual information between reliable structure and smoothed relations, BioKDN emphasizes informative semantics to enable precise predictions. Experimental results on real-world datasets show that BioKDN surpasses state-of-the-art models in DTI and DDI prediction tasks, confirming the effectiveness and robustness of BioKDN in denoising unreliable interactions within contaminated KGs
comment: 13 pages, Accepted at TKDE
♻ ☆ Diverse Policies Recovering via Pointwise Mutual Information Weighted Imitation Learning
Recovering a spectrum of diverse policies from a set of expert trajectories is an important research topic in imitation learning. After determining a latent style for a trajectory, previous diverse policies recovering methods usually employ a vanilla behavioral cloning learning objective conditioned on the latent style, treating each state-action pair in the trajectory with equal importance. Based on an observation that in many scenarios, behavioral styles are often highly relevant with only a subset of state-action pairs, this paper presents a new principled method in diverse polices recovery. In particular, after inferring or assigning a latent style for a trajectory, we enhance the vanilla behavioral cloning by incorporating a weighting mechanism based on pointwise mutual information. This additional weighting reflects the significance of each state-action pair's contribution to learning the style, thus allowing our method to focus on state-action pairs most representative of that style. We provide theoretical justifications for our new objective, and extensive empirical evaluations confirm the effectiveness of our method in recovering diverse policies from expert data.
comment: 18 pages, 6 figures
♻ ☆ Reducing Hallucinations in Vision-Language Models via Latent Space Steering
Hallucination poses a challenge to the deployment of large vision-language models (LVLMs) in applications. Unlike in large language models (LLMs), hallucination in LVLMs often arises from misalignments between visual inputs and textual outputs. This paper investigates the underlying mechanisms of hallucination, focusing on the unique structure of LVLMs that distinguishes them from large language models (LLMs). We identify that hallucinations often arise from the sensitivity of text decoders to vision inputs, a natural phenomenon when image encoders and text decoders are pre-trained separately. Inspired by this, we introduce Visual and Textual Intervention (VTI), a novel technique designed to reduce hallucinations by steering latent space representations during inference to enhance the stability of vision features. As a task-agnostic test-time intervention, VTI can be easily applied to any problem without additional cost. Extensive experiments demonstrate that it can effectively reduce hallucinations and outperform baseline methods across multiple metrics, highlighting the critical role of vision feature stability in LVLMs.
comment: 21 pages
♻ ☆ FDF: Flexible Decoupled Framework for Time Series Forecasting with Conditional Denoising and Polynomial Modeling
Time series forecasting is vital in numerous web applications, influencing critical decision-making across industries. While diffusion models have recently gained increasing popularity for this task, we argue they suffer from a significant drawback: indiscriminate noise addition to the original time series followed by denoising, which can obscure underlying dynamic evolving trend and complicate forecasting. To address this limitation, we propose a novel flexible decoupled framework (FDF) that learns high-quality time series representations for enhanced forecasting performance. A key characteristic of our approach leverages the inherent inductive bias of time series data by decomposing it into trend and seasonal components, each modeled separately to enable decoupled analysis and modeling. Specifically, we propose an innovative Conditional Denoising Seasonal Module (CDSM) within the diffusion model, which leverages statistical information from the historical window to conditionally model the complex seasonal component. Notably, we incorporate a Polynomial Trend Module (PTM) to effectively capture the smooth trend component, thereby enhancing the model's ability to represent temporal dependencies. Extensive experiments validate the effectiveness of our framework, demonstrating superior performance over existing methods and higlighting its flexibility in time series forecasting. We hope our work can bring a new perspective for time series forecasting. We intend to make our code publicly available as open-source in the future.
♻ ☆ BeGin: Extensive Benchmark Scenarios and An Easy-to-use Framework for Graph Continual Learning
Continual Learning (CL) is the process of learning ceaselessly a sequence of tasks. Most existing CL methods deal with independent data (e.g., images and text) for which many benchmark frameworks and results under standard experimental settings are available. Compared to them, however, CL methods for graph data (graph CL) are relatively underexplored because of (a) the lack of standard experimental settings, especially regarding how to deal with the dependency between instances, (b) the lack of benchmark datasets and scenarios, and (c) high complexity in implementation and evaluation due to the dependency. In this paper, regarding (a) we define four standard incremental settings (task-, class-, domain-, and time-incremental) for node-, link-, and graph-level problems, extending the previously explored scope. Regarding (b), we provide 35 benchmark scenarios based on 24 real-world graphs. Regarding (c), we develop BeGin, an easy and fool-proof framework for graph CL. BeGin is easily extended since it is modularized with reusable modules for data processing, algorithm design, and evaluation. Especially, the evaluation module is completely separated from user code to eliminate potential mistakes. Regarding benchmark results, we cover 3x more combinations of incremental settings and levels of problems than the latest benchmark. All assets for the benchmark framework are publicly available at https://github.com/ShinhwanKang/BeGin.
comment: Full version of the ACM TIST paper with the same title
♻ ☆ Treeffuser: Probabilistic Predictions via Conditional Diffusions with Gradient-Boosted Trees NeurIPS 2024
Probabilistic prediction aims to compute predictive distributions rather than single point predictions. These distributions enable practitioners to quantify uncertainty, compute risk, and detect outliers. However, most probabilistic methods assume parametric responses, such as Gaussian or Poisson distributions. When these assumptions fail, such models lead to bad predictions and poorly calibrated uncertainty. In this paper, we propose Treeffuser, an easy-to-use method for probabilistic prediction on tabular data. The idea is to learn a conditional diffusion model where the score function is estimated using gradient-boosted trees. The conditional diffusion model makes Treeffuser flexible and non-parametric, while the gradient-boosted trees make it robust and easy to train on CPUs. Treeffuser learns well-calibrated predictive distributions and can handle a wide range of regression tasks -- including those with multivariate, multimodal, and skewed responses. We study Treeffuser on synthetic and real data and show that it outperforms existing methods, providing better calibrated probabilistic predictions. We further demonstrate its versatility with an application to inventory allocation under uncertainty using sales data from Walmart. We implement Treeffuser in https://github.com/blei-lab/treeffuser.
comment: NeurIPS 2024
♻ ☆ Generative Kaleidoscopic Networks
We discovered that the neural networks, especially the deep ReLU networks, demonstrate an `over-generalization' phenomenon. That is, the output values for the inputs that were not seen during training are mapped close to the output range that were observed during the learning process. In other words, the neural networks learn a many-to-one mapping and this effect is more prominent as we increase the number of layers or the depth of the neural network. We utilize this property of neural networks to design a dataset kaleidoscope, termed as `Generative Kaleidoscopic Networks'. Succinctly, if we learn a model to map from input $x\in\mathbb{R}^D$ to itself $f_\mathcal{N}(x)\rightarrow x$, the proposed `Kaleidoscopic sampling' procedure starts with a random input noise $z\in\mathbb{R}^D$ and recursively applies $f_\mathcal{N}(\cdots f_\mathcal{N}(z)\cdots )$. After a burn-in period duration, we start observing samples from the input distribution and the quality of samples recovered improves as we increase the depth of the model. Scope: We observed this phenomenon to various degrees for the other deep learning architectures like CNNs, Transformers & U-Nets and we are currently investigating them further.
♻ ☆ Designing Network Algorithms via Large Language Models
We introduce NADA, the first framework to autonomously design network algorithms by leveraging the generative capabilities of large language models (LLMs). Starting with an existing algorithm implementation, NADA enables LLMs to create a wide variety of alternative designs in the form of code blocks. It then efficiently identifies the top-performing designs through a series of filtering techniques, minimizing the need for full-scale evaluations and significantly reducing computational costs. Using adaptive bitrate (ABR) streaming as a case study, we demonstrate that NADA produces novel ABR algorithms -- previously unknown to human developers -- that consistently outperform the original algorithm in diverse network environments, including broadband, satellite, 4G, and 5G.
♻ ☆ Can AI weather models predict out-of-distribution gray swan tropical cyclones?
Predicting gray swan weather extremes, which are possible but so rare that they are absent from the training dataset, is a major concern for AI weather/climate models. An important open question is whether AI models can extrapolate from weaker weather events present in the training set to stronger, unseen weather extremes. To test this, we train independent versions of the AI model FourCastNet on the 1979-2015 ERA5 dataset with all data, or with Category 3-5 tropical cyclones (TCs) removed, either globally or only over the North Atlantic or Western Pacific basin. We then test these versions of FourCastNet on 2018-2023 Category 5 TCs (gray swans). All versions yield similar accuracy for global weather, but the one trained without Category 3-5 TCs cannot accurately forecast Category 5 TCs, indicating that these models cannot extrapolate from weaker storms. The versions trained without Category 3-5 TCs in one basin show some skill forecasting Category 5 TCs in that basin, suggesting that FourCastNet can generalize across tropical basins. This is encouraging and surprising because regional information is implicitly encoded in inputs. No version satisfies gradient-wind balance, implying that enforcing such physical constraints may not improve generalizability to gray swans. Given that current state-of-the-art AI weather/climate models have similar learning strategies, we expect our findings to apply to other models and extreme events. Our work demonstrates that novel learning strategies are needed for AI weather/climate models to provide early warning or estimated statistics for the rarest, most impactful weather extremes.
♻ ☆ Federated Learning with Neural Graphical Models
Federated Learning (FL) addresses the need to create models based on proprietary data in such a way that multiple clients retain exclusive control over their data, while all benefit from improved model accuracy due to pooled resources. Recently proposed Neural Graphical Models (NGMs) are Probabilistic Graphical models that utilize the expressive power of neural networks to learn complex non-linear dependencies between the input features. They learn to capture the underlying data distribution and have efficient algorithms for inference and sampling. We develop a FL framework which maintains a global NGM model that learns the averaged information from the local NGM models while keeping the training data within the client's environment. Our design, FedNGMs, avoids the pitfalls and shortcomings of neuron matching frameworks like Federated Matched Averaging that suffers from model parameter explosion. Our global model size remains constant throughout the process. In the cases where clients have local variables that are not part of the combined global distribution, we propose a `Stitching' algorithm, which personalizes the global NGM models by merging the additional variables using the client's data. FedNGM is robust to data heterogeneity, large number of participants, and limited communication bandwidth. We experimentally demonstrated the use of FedNGMs for extracting insights from CDC's Infant Mortality dataset and discuss interesting future applications.
♻ ☆ Warped Diffusion: Solving Video Inverse Problems with Image Diffusion Models NeurIPS 2024
Using image models naively for solving inverse video problems often suffers from flickering, texture-sticking, and temporal inconsistency in generated videos. To tackle these problems, in this paper, we view frames as continuous functions in the 2D space, and videos as a sequence of continuous warping transformations between different frames. This perspective allows us to train function space diffusion models only on images and utilize them to solve temporally correlated inverse problems. The function space diffusion models need to be equivariant with respect to the underlying spatial transformations. To ensure temporal consistency, we introduce a simple post-hoc test-time guidance towards (self)-equivariant solutions. Our method allows us to deploy state-of-the-art latent diffusion models such as Stable Diffusion XL to solve video inverse problems. We demonstrate the effectiveness of our method for video inpainting and $8\times$ video super-resolution, outperforming existing techniques based on noise transformations. We provide generated video results: https://giannisdaras.github.io/warped_diffusion.github.io/.
comment: Accepted in NeurIPS 2024
♻ ☆ Q-WSL: Optimizing Goal-Conditioned RL with Weighted Supervised Learning via Dynamic Programming
A novel class of advanced algorithms, termed Goal-Conditioned Weighted Supervised Learning (GCWSL), has recently emerged to tackle the challenges posed by sparse rewards in goal-conditioned reinforcement learning (RL). GCWSL consistently delivers strong performance across a diverse set of goal-reaching tasks due to its simplicity, effectiveness, and stability. However, GCWSL methods lack a crucial capability known as trajectory stitching, which is essential for learning optimal policies when faced with unseen skills during testing. This limitation becomes particularly pronounced when the replay buffer is predominantly filled with sub-optimal trajectories. In contrast, traditional TD-based RL methods, such as Q-learning, which utilize Dynamic Programming, do not face this issue but often experience instability due to the inherent difficulties in value function approximation. In this paper, we propose Q-learning Weighted Supervised Learning (Q-WSL), a novel framework designed to overcome the limitations of GCWSL by incorporating the strengths of Dynamic Programming found in Q-learning. Q-WSL leverages Dynamic Programming results to output the optimal action of (state, goal) pairs across different trajectories within the replay buffer. This approach synergizes the strengths of both Q-learning and GCWSL, effectively mitigating their respective weaknesses and enhancing overall performance. Empirical evaluations on challenging goal-reaching tasks demonstrate that Q-WSL surpasses other goal-conditioned approaches in terms of both performance and sample efficiency. Additionally, Q-WSL exhibits notable robustness in environments characterized by binary reward structures and environmental stochasticity.
♻ ☆ Is the MMI Criterion Necessary for Interpretability? Degenerating Non-causal Features to Plain Noise for Self-Rationalization NeurIPS 2024
An important line of research in the field of explainability is to extract a small subset of crucial rationales from the full input. The most widely used criterion for rationale extraction is the maximum mutual information (MMI) criterion. However, in certain datasets, there are spurious features non-causally correlated with the label and also get high mutual information, complicating the loss landscape of MMI. Although some penalty-based methods have been developed to penalize the spurious features (e.g., invariance penalty, intervention penalty, etc) to help MMI work better, these are merely remedial measures. In the optimization objectives of these methods, spurious features are still distinguished from plain noise, which hinders the discovery of causal rationales. This paper aims to develop a new criterion that treats spurious features as plain noise, allowing the model to work on datasets rich in spurious features as if it were working on clean datasets, thereby making rationale extraction easier. We theoretically observe that removing either plain noise or spurious features from the input does not alter the conditional distribution of the remaining components relative to the task label. However, significant changes in the conditional distribution occur only when causal features are eliminated. Based on this discovery, the paper proposes a criterion for \textbf{M}aximizing the \textbf{R}emaining \textbf{D}iscrepancy (MRD). Experiments on six widely used datasets show that our MRD criterion improves rationale quality (measured by the overlap with human-annotated rationales) by up to $10.4\%$ as compared to several recent competitive MMI variants. Code: \url{https://github.com/jugechengzi/Rationalization-MRD}.
comment: Accepted at NeurIPS 2024. arXiv admin note: text overlap with arXiv:2309.13391
♻ ☆ Boardwalk Empire: How Generative AI is Revolutionizing Economic Paradigms
The relentless pursuit of technological advancements has ushered in a new era where artificial intelligence (AI) is not only a powerful tool but also a critical economic driver. At the forefront of this transformation is Generative AI, which is catalyzing a paradigm shift across industries. Deep generative models, an integration of generative and deep learning techniques, excel in creating new data beyond analyzing existing ones, revolutionizing sectors from production and manufacturing to finance. By automating design, optimization, and innovation cycles, Generative AI is reshaping core industrial processes. In the financial sector, it is transforming risk assessment, trading strategies, and forecasting, demonstrating its profound impact. This paper explores the sweeping changes driven by deep learning models like Large Language Models (LLMs), highlighting their potential to foster innovative business models, disruptive technologies, and novel economic landscapes. As we stand at the threshold of an AI-driven economic era, Generative AI is emerging as a pivotal force, driving innovation, disruption, and economic evolution on a global scale.
comment: 25 pages, 8 figures, Accepted at National Conference on Advances in Marketing Paradigms for Research, Innovation and Technology (AMRIT 2023)
♻ ☆ Neural Graph Revealers
Sparse graph recovery methods work well where the data follows their assumptions but often they are not designed for doing downstream probabilistic queries. This limits their adoption to only identifying connections among the input variables. On the other hand, the Probabilistic Graphical Models (PGMs) assume an underlying base graph between variables and learns a distribution over them. PGM design choices are carefully made such that the inference \& sampling algorithms are efficient. This brings in certain restrictions and often simplifying assumptions. In this work, we propose Neural Graph Revealers (NGRs), that are an attempt to efficiently merge the sparse graph recovery methods with PGMs into a single flow. The problem setting consists of an input data X with D features and M samples and the task is to recover a sparse graph showing connection between the features and jointly learn a probability distribution over them. NGRs view the neural networks as a `glass box' or more specifically as a multitask learning framework. We introduce `Graph-constrained path norm' that NGRs leverage to learn a graphical model that captures complex non-linear functional dependencies between the features in the form of an undirected sparse graph. Furthermore, NGRs can handle multimodal inputs like images, text, categorical data, embeddings etc. which is not straightforward to incorporate in the existing methods. We show experimental results of doing sparse graph recovery and probabilistic inference on data from Gaussian graphical models and a multimodal infant mortality dataset by Centers for Disease Control and Prevention.
♻ ☆ Stochastic Sparse Sampling: A Framework for Variable-Length Medical Time Series Classification
While the majority of time series classification research has focused on modeling fixed-length sequences, variable-length time series classification (VTSC) remains critical in healthcare, where sequence length may vary among patients and events. To address this challenge, we propose $\textbf{S}$tochastic $\textbf{S}$parse $\textbf{S}$ampling (SSS), a novel VTSC framework developed for medical time series. SSS manages variable-length sequences by sparsely sampling fixed windows to compute local predictions, which are then aggregated and calibrated to form a global prediction. We apply SSS to the task of seizure onset zone (SOZ) localization, a critical VTSC problem requiring identification of seizure-inducing brain regions from variable-length electrophysiological time series. We evaluate our method on the Epilepsy iEEG Multicenter Dataset, a heterogeneous collection of intracranial electroencephalography (iEEG) recordings obtained from four independent medical centers. SSS demonstrates superior performance compared to state-of-the-art (SOTA) baselines across most medical centers, and superior performance on all out-of-distribution (OOD) unseen medical centers. Additionally, SSS naturally provides post-hoc insights into local signal characteristics related to the SOZ, by visualizing temporally averaged local predictions throughout the signal.
comment: 20 pages, 8 figures, 2 tables
♻ ☆ Multi-Layer Feature Fusion with Cross-Channel Attention-Based U-Net for Kidney Tumor Segmentation
Renal tumors, especially renal cell carcinoma (RCC), show significant heterogeneity, posing challenges for diagnosis using radiology images such as MRI, echocardiograms, and CT scans. U-Net based deep learning techniques are emerging as a promising approach for automated medical image segmentation for minimally invasive diagnosis of renal tumors. However, current techniques need further improvements in accuracy to become clinically useful to radiologists. In this study, we present an improved U-Net based model for end-to-end automated semantic segmentation of CT scan images to identify renal tumors. The model uses residual connections across convolution layers, integrates a multi-layer feature fusion (MFF) and cross-channel attention (CCA) within encoder blocks, and incorporates skip connections augmented with additional information derived using MFF and CCA. We evaluated our model on the KiTS19 dataset, which contains data from 210 patients. For kidney segmentation, our model achieves a Dice Similarity Coefficient (DSC) of 0.97 and a Jaccard index (JI) of 0.95. For renal tumor segmentation, our model achieves a DSC of 0.96 and a JI of 0.91. Based on a comparison of available DSC scores, our model outperforms the current leading models.
comment: 8 pages
♻ ☆ From PINNs to PIKANs: Recent Advances in Physics-Informed Machine Learning
Physics-Informed Neural Networks (PINNs) have emerged as a key tool in Scientific Machine Learning since their introduction in 2017, enabling the efficient solution of ordinary and partial differential equations using sparse measurements. Over the past few years, significant advancements have been made in the training and optimization of PINNs, covering aspects such as network architectures, adaptive refinement, domain decomposition, and the use of adaptive weights and activation functions. A notable recent development is the Physics-Informed Kolmogorov-Arnold Networks (PIKANS), which leverage a representation model originally proposed by Kolmogorov in 1957, offering a promising alternative to traditional PINNs. In this review, we provide a comprehensive overview of the latest advancements in PINNs, focusing on improvements in network design, feature expansion, optimization techniques, uncertainty quantification, and theoretical insights. We also survey key applications across a range of fields, including biomedicine, fluid and solid mechanics, geophysics, dynamical systems, heat transfer, chemical engineering, and beyond. Finally, we review computational frameworks and software tools developed by both academia and industry to support PINN research and applications.
comment: physics-informed neural networks, Kolmogorov-Arnold networks, optimization algorithms, separable PINNs, self-adaptive weights, uncertainty quantification
♻ ☆ Are uGLAD? Time will tell!
We frequently encounter multiple series that are temporally correlated in our surroundings, such as EEG data to examine alterations in brain activity or sensors to monitor body movements. Segmentation of multivariate time series data is a technique for identifying meaningful patterns or changes in the time series that can signal a shift in the system's behavior. However, most segmentation algorithms have been designed primarily for univariate time series, and their performance on multivariate data remains largely unsatisfactory, making this a challenging problem. In this work, we introduce a novel approach for multivariate time series segmentation using conditional independence (CI) graphs. CI graphs are probabilistic graphical models that represents the partial correlations between the nodes. We propose a domain agnostic multivariate segmentation framework $\texttt{tGLAD}$ which draws a parallel between the CI graph nodes and the variables of the time series. Consider applying a graph recovery model $\texttt{uGLAD}$ to a short interval of the time series, it will result in a CI graph that shows partial correlations among the variables. We extend this idea to the entire time series by utilizing a sliding window to create a batch of time intervals and then run a single $\texttt{uGLAD}$ model in multitask learning mode to recover all the CI graphs simultaneously. As a result, we obtain a corresponding temporal CI graphs representation. We then designed a first-order and second-order based trajectory tracking algorithms to study the evolution of these graphs across distinct intervals. Finally, an `Allocation' algorithm is used to determine a suitable segmentation of the temporal graph sequence. $\texttt{tGLAD}$ provides a competitive time complexity of $O(N)$ for settings where number of variables $D<
♻ ☆ Knowledge Propagation over Conditional Independence Graphs
Conditional Independence (CI) graph is a special type of a Probabilistic Graphical Model (PGM) where the feature connections are modeled using an undirected graph and the edge weights show the partial correlation strength between the features. Since the CI graphs capture direct dependence between features, they have been garnering increasing interest within the research community for gaining insights into the systems from various domains, in particular discovering the domain topology. In this work, we propose algorithms for performing knowledge propagation over the CI graphs. Our experiments demonstrate that our techniques improve upon the state-of-the-art on the publicly available Cora and PubMed datasets.
♻ ☆ Forgettable Federated Linear Learning with Certified Data Unlearning
The advent of Federated Learning (FL) has revolutionized the way distributed systems handle collaborative model training while preserving user privacy. Recently, Federated Unlearning (FU) has emerged to address demands for the "right to be forgotten"" and unlearning of the impact of poisoned clients without requiring retraining in FL. Most FU algorithms require the cooperation of retained or target clients (clients to be unlearned), introducing additional communication overhead and potential security risks. In addition, some FU methods need to store historical models to execute the unlearning process. These challenges hinder the efficiency and memory constraints of the current FU methods. Moreover, due to the complexity of nonlinear models and their training strategies, most existing FU methods for deep neural networks (DNN) lack theoretical certification. In this work, we introduce a novel FL training and unlearning strategy in DNN, termed Forgettable Federated Linear Learning (F^2L^2). F^2L^2 considers a common practice of using pre-trained models to approximate DNN linearly, allowing them to achieve similar performance as the original networks via Federated Linear Training (FLT). We then present FedRemoval, a certified, efficient, and secure unlearning strategy that enables the server to unlearn a target client without requiring client communication or adding additional storage. We have conducted extensive empirical validation on small- to large-scale datasets, using both convolutional neural networks and modern foundation models. These experiments demonstrate the effectiveness of F^2L^2 in balancing model accuracy with the successful unlearning of target clients. F^2L^2 represents a promising pipeline for efficient and trustworthy FU. The code is available here.
comment: 17 pages, 6 figures
♻ ☆ Towards Fair Graph Representation Learning in Social Networks
With the widespread use of Graph Neural Networks (GNNs) for representation learning from network data, the fairness of GNN models has raised great attention lately. Fair GNNs aim to ensure that node representations can be accurately classified, but not easily associated with a specific group. Existing advanced approaches essentially enhance the generalisation of node representation in combination with data augmentation strategy, and do not directly impose constraints on the fairness of GNNs. In this work, we identify that a fundamental reason for the unfairness of GNNs in social network learning is the phenomenon of social homophily, i.e., users in the same group are more inclined to congregate. The message-passing mechanism of GNNs can cause users in the same group to have similar representations due to social homophily, leading model predictions to establish spurious correlations with sensitive attributes. Inspired by this reason, we propose a method called Equity-Aware GNN (EAGNN) towards fair graph representation learning. Specifically, to ensure that model predictions are independent of sensitive attributes while maintaining prediction performance, we introduce constraints for fair representation learning based on three principles: sufficiency, independence, and separation. We theoretically demonstrate that our EAGNN method can effectively achieve group fairness. Extensive experiments on three datasets with varying levels of social homophily illustrate that our EAGNN method achieves the state-of-the-art performance across two fairness metrics and offers competitive effectiveness.
♻ ☆ TopoNav: Topological Navigation for Efficient Exploration in Sparse Reward Environments IROS
Autonomous robots exploring unknown environments face a significant challenge: navigating effectively without prior maps and with limited external feedback. This challenge intensifies in sparse reward environments, where traditional exploration techniques often fail. In this paper, we present TopoNav, a novel topological navigation framework that integrates active mapping, hierarchical reinforcement learning, and intrinsic motivation to enable efficient goal-oriented exploration and navigation in sparse-reward settings. TopoNav dynamically constructs a topological map of the environment, capturing key locations and pathways. A two-level hierarchical policy architecture, comprising a high-level graph traversal policy and low-level motion control policies, enables effective navigation and obstacle avoidance while maintaining focus on the overall goal. Additionally, TopoNav incorporates intrinsic motivation to guide exploration toward relevant regions and frontier nodes in the topological map, addressing the challenges of sparse extrinsic rewards. We evaluate TopoNav both in the simulated and real-world off-road environments using a Clearpath Jackal robot, across three challenging navigation scenarios: goal-reaching, feature-based navigation, and navigation in complex terrains. We observe an increase in exploration coverage by 7- 20%, in success rates by 9-19%, and reductions in navigation times by 15-36% across various scenarios, compared to state-of-the-art methods
comment: Accepted at the 37th IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2024
♻ ☆ Merging LoRAs like Playing LEGO: Pushing the Modularity of LoRA to Extremes Through Rank-Wise Clustering
Low-Rank Adaptation (LoRA) has emerged as a popular technique for fine-tuning large language models (LLMs) to various domains due to its modular design and widespread availability on platforms like Huggingface. This modularity has sparked interest in combining multiple LoRAs to enhance LLM capabilities. However, existing methods for LoRA composition primarily focus on task-specific adaptations that require additional training, and current model merging techniques often fail to fully leverage LoRA's modular nature, leading to parameter interference and performance degradation. In this paper, we investigate the feasibility of disassembling and reassembling multiple LoRAs at a finer granularity, analogous to assembling LEGO blocks. We introduce the concept of Minimal Semantic Units (MSUs), where the parameters corresponding to each rank in LoRA function as independent units. These MSUs demonstrate permutation invariance and concatenation-summation equivalence properties, enabling flexible combinations to create new LoRAs. Building on these insights, we propose the LoRA-LEGO framework. This framework conducts rank-wise parameter clustering by grouping MSUs from different LoRAs into $k$ clusters. The centroid of each cluster serves as a representative MSU, enabling the assembly of a merged LoRA with an adjusted rank of $k$. Additionally, we apply a dual reweighting strategy to optimize the scale of the merged LoRA. Experiments across various benchmarks demonstrate that our method outperforms existing approaches in LoRA merging.
♻ ☆ Towards Domain Adaptive Neural Contextual Bandits
Contextual bandit algorithms are essential for solving real-world decision making problems. In practice, collecting a contextual bandit's feedback from different domains may involve different costs. For example, measuring drug reaction from mice (as a source domain) and humans (as a target domain). Unfortunately, adapting a contextual bandit algorithm from a source domain to a target domain with distribution shift still remains a major challenge and largely unexplored. In this paper, we introduce the first general domain adaptation method for contextual bandits. Our approach learns a bandit model for the target domain by collecting feedback from the source domain. Our theoretical analysis shows that our algorithm maintains a sub-linear regret bound even adapting across domains. Empirical results show that our approach outperforms the state-of-the-art contextual bandit algorithms on real-world datasets.
♻ ☆ Wireless Link Quality Estimation Using LSTM Model
In recent years, various services have been provided through high-speed and high-capacity wireless networks on mobile communication devices, necessitating stable communication regardless of indoor or outdoor environments. To achieve stable communication, it is essential to implement proactive measures, such as switching to an alternative path and ensuring data buffering before the communication quality becomes unstable. The technology of Wireless Link Quality Estimation (WLQE), which predicts the communication quality of wireless networks in advance, plays a crucial role in this context. In this paper, we propose a novel WLQE model for estimating the communication quality of wireless networks by leveraging sequential information. Our proposed method is based on Long Short-Term Memory (LSTM), enabling highly accurate estimation by considering the sequential information of link quality. We conducted a comparative evaluation with the conventional model, stacked autoencoder-based link quality estimator (LQE-SAE), using a dataset recorded in real-world environmental conditions. Our LSTM-based LQE model demonstrates its superiority, achieving a 4.0% higher accuracy and a 4.6% higher macro-F1 score than the LQE-SAE model in the evaluation.
comment: This paper was submitted to IEEE Network Operations and Management Symposium
♻ ☆ Diffusion-Reward Adversarial Imitation Learning
Imitation learning aims to learn a policy from observing expert demonstrations without access to reward signals from environments. Generative adversarial imitation learning (GAIL) formulates imitation learning as adversarial learning, employing a generator policy learning to imitate expert behaviors and discriminator learning to distinguish the expert demonstrations from agent trajectories. Despite its encouraging results, GAIL training is often brittle and unstable. Inspired by the recent dominance of diffusion models in generative modeling, we propose Diffusion-Reward Adversarial Imitation Learning (DRAIL), which integrates a diffusion model into GAIL, aiming to yield more robust and smoother rewards for policy learning. Specifically, we propose a diffusion discriminative classifier to construct an enhanced discriminator, and design diffusion rewards based on the classifier's output for policy learning. Extensive experiments are conducted in navigation, manipulation, and locomotion, verifying DRAIL's effectiveness compared to prior imitation learning methods. Moreover, additional experimental results demonstrate the generalizability and data efficiency of DRAIL. Visualized learned reward functions of GAIL and DRAIL suggest that DRAIL can produce more robust and smoother rewards. Project page: https://nturobotlearninglab.github.io/DRAIL/
♻ ☆ Soft ascent-descent as a stable and flexible alternative to flooding NeurIPS 2024
As a heuristic for improving test accuracy in classification, the "flooding" method proposed by Ishida et al. (2020) sets a threshold for the average surrogate loss at training time; above the threshold, gradient descent is run as usual, but below the threshold, a switch to gradient ascent is made. While setting the threshold is non-trivial and is usually done with validation data, this simple technique has proved remarkably effective in terms of accuracy. On the other hand, what if we are also interested in other metrics such as model complexity or average surrogate loss at test time? As an attempt to achieve better overall performance with less fine-tuning, we propose a softened, pointwise mechanism called SoftAD (soft ascent-descent) that downweights points on the borderline, limits the effects of outliers, and retains the ascent-descent effect of flooding, with no additional computational overhead. We contrast formal stationarity guarantees with those for flooding, and empirically demonstrate how SoftAD can realize classification accuracy competitive with flooding (and the more expensive alternative SAM) while enjoying a much smaller loss generalization gap and model norm.
comment: Revised version accepted to NeurIPS 2024
♻ ☆ Hierarchical Simplicity Bias of Neural Networks
Neural networks often exhibit simplicity bias, favoring simpler features over more complex ones, even when both are equally predictive. We introduce a novel method called imbalanced label coupling to explore and extend this simplicity bias across multiple hierarchical levels. Our approach demonstrates that trained networks sequentially consider features of increasing complexity based on their correlation with labels in the training set, regardless of their actual predictive power. For example, in CIFAR-10, simple spurious features can cause misclassifications where most cats are predicted as dogs and most trucks as automobiles. We empirically show that last-layer retraining with target data distribution \citep{kirichenko2022last} is insufficient to fully recover core features when spurious features perfectly correlate with target labels in our synthetic datasets. Our findings deepen the understanding of the implicit biases inherent in neural networks.
comment: 20 pages, 21 figures, revised version, accepted at OPT2024: 16th Annual Workshop on Optimization for Machine Learning
♻ ☆ On the Power of Foundation Models ICML'23
With infinitely many high-quality data points, infinite computational power, an infinitely large foundation model with a perfect training algorithm and guaranteed zero generalization error on the pretext task, can the model be used for everything? This question cannot be answered by the existing theory of representation, optimization or generalization, because the issues they mainly investigate are assumed to be nonexistent here. In this paper, we show that category theory provides powerful machinery to answer this question. We have proved three results. The first one limits the power of prompt-based learning, saying that the model can solve a downstream task with prompts if and only if the task is representable. The second one says fine tuning does not have this limit, as a foundation model with the minimum required power (up to symmetry) can theoretically solve downstream tasks for the category defined by pretext task, with fine tuning and enough resources. Our final result can be seen as a new type of generalization theorem, showing that the foundation model can generate unseen objects from the target category (e.g., images) using the structural information from the source category (e.g., texts). Along the way, we provide a categorical framework for supervised and self-supervised learning, which might be of independent interest.
comment: ICML'23. This version fixed a bug when applying prompt tuning theorem to LLM
♻ ☆ Community Detection Guarantees Using Embeddings Learned by Node2Vec
Embedding the nodes of a large network into an Euclidean space is a common objective in modern machine learning, with a variety of tools available. These embeddings can then be used as features for tasks such as community detection/node clustering or link prediction, where they achieve state of the art performance. With the exception of spectral clustering methods, there is little theoretical understanding for commonly used approaches to learning embeddings. In this work we examine the theoretical properties of the embeddings learned by node2vec. Our main result shows that the use of $k$-means clustering on the embedding vectors produced by node2vec gives weakly consistent community recovery for the nodes in (degree corrected) stochastic block models. We also discuss the use of these embeddings for node and link prediction tasks. We demonstrate this result empirically, and examine how this relates to other embedding tools for network data.
comment: Camera ready version for Neurips 2024
♻ ☆ Debiasing Text Safety Classifiers through a Fairness-Aware Ensemble
Increasing use of large language models (LLMs) demand performant guardrails to ensure the safety of inputs and outputs of LLMs. When these safeguards are trained on imbalanced data, they can learn the societal biases. We present a light-weight, post-processing method for mitigating counterfactual fairness in closed-source text safety classifiers. Our approach involves building an ensemble that not only outperforms the input classifiers and policy-aligns them, but also acts as a debiasing regularizer. We introduce two threshold-agnostic metrics to assess the counterfactual fairness of a model, and demonstrate how combining these metrics with Fair Data Reweighting (FDW) helps mitigate biases. We create an expanded Open AI dataset, and a new templated LLM-generated dataset based on user-prompts, both of which are counterfactually balanced across identity groups and cover four key areas of safety; we will work towards publicly releasing these datasets. Our results show that our approach improves counterfactual fairness with minimal impact on model performance.
♻ ☆ An Empirical Comparison of Vocabulary Expansion and Initialization Approaches for Language Models EMNLP 2024
Language Models (LMs) excel in natural language processing tasks for English but show reduced performance in most other languages. This problem is commonly tackled by continually pre-training and fine-tuning these models for said languages. A significant issue in this process is the limited vocabulary coverage in the original model's tokenizer, leading to inadequate representation of new languages and necessitating an expansion of the tokenizer. The initialization of the embeddings corresponding to new vocabulary items presents a further challenge. Current strategies require cross-lingual embeddings and lack a solid theoretical foundation as well as comparisons with strong baselines. In this paper, we first establish theoretically that initializing within the convex hull of existing embeddings is a good initialization, followed by a novel but simple approach, Constrained Word2Vec (CW2V), which does not require cross-lingual embeddings. Our study evaluates different initialization methods for expanding RoBERTa and LLaMA 2 across four languages and five tasks. The results show that CW2V performs equally well or even better than more advanced techniques. Additionally, simpler approaches like multivariate initialization perform on par with these advanced methods indicating that efficient large-scale multilingual continued pretraining can be achieved even with simpler initialization methods. We release our code publicly (https://github.com/AI4Bharat/VocabAdaptation_LLM/tree/CW2V).
comment: CONLL 2024 (EMNLP 2024)
♻ ☆ Reinforcement Learning for Infinite-Horizon Average-Reward Linear MDPs via Approximation by Discounted-Reward MDPs
We study the infinite-horizon average-reward reinforcement learning with linear MDPs. Previous approaches either suffer from computational inefficiency or require strong assumptions on dynamics, such as ergodicity, for achieving a regret bound of $\widetilde{O}(\sqrt{T})$. In this paper, we propose an algorithm that achieves the regret bound of $\widetilde{O}(\sqrt{T})$ and is computationally efficient in the sense that the time complexity is polynomial in problem parameters. Our algorithm runs an optimistic value iteration on a discounted-reward MDP that approximates the average-reward setting. With an appropriately tuned discounting factor $\gamma$, the algorithm attains the desired $\widetilde{O}(\sqrt{T})$ regret. The challenge in our approximation approach is to get a regret bound with a sharp dependency on the effective horizon $1 / (1 - \gamma)$. We address this challenge by clipping the value function obtained at each value iteration step to limit the span of the value function.
comment: Fixes an error in the analysis in the previous version by modifying the algorithm and analysis
♻ ☆ NVLM: Open Frontier-Class Multimodal LLMs
We introduce NVLM 1.0, a family of frontier-class multimodal large language models (LLMs) that achieve state-of-the-art results on vision-language tasks, rivaling the leading proprietary models (e.g., GPT-4o) and open-access models (e.g., Llama 3-V 405B and InternVL 2). Remarkably, NVLM 1.0 shows improved text-only performance over its LLM backbone after multimodal training. In terms of model design, we perform a comprehensive comparison between decoder-only multimodal LLMs (e.g., LLaVA) and cross-attention-based models (e.g., Flamingo). Based on the strengths and weaknesses of both approaches, we propose a novel architecture that enhances both training efficiency and multimodal reasoning capabilities. Furthermore, we introduce a 1-D tile-tagging design for tile-based dynamic high-resolution images, which significantly boosts performance on multimodal reasoning and OCR-related tasks. Regarding training data, we meticulously curate and provide detailed information on our multimodal pretraining and supervised fine-tuning datasets. Our findings indicate that dataset quality and task diversity are more important than scale, even during the pretraining phase, across all architectures. Notably, we develop production-grade multimodality for the NVLM-1.0 models, enabling them to excel in vision-language tasks while maintaining and even improving text-only performance compared to their LLM backbones. To achieve this, we craft and integrate a high-quality text-only dataset into multimodal training, alongside a substantial amount of multimodal math and reasoning data, leading to enhanced math and coding capabilities across modalities. To advance research in the field, we release the model weights at https://huggingface.co/nvidia/NVLM-D-72B and will open-source the training code for the community soon.
comment: Fixed the typos. For more information, please visit our project page at: https://research.nvidia.com/labs/adlr/NVLM-1
♻ ☆ Encoder Embedding for General Graph and Node Classification
Graph encoder embedding, a recent technique for graph data, offers speed and scalability in producing vertex-level representations from binary graphs. In this paper, we extend the applicability of this method to a general graph model, which includes weighted graphs, distance matrices, and kernel matrices. We prove that the encoder embedding satisfies the law of large numbers and the central limit theorem on a per-observation basis. Under certain condition, it achieves asymptotic normality on a per-class basis, enabling optimal classification through discriminant analysis. These theoretical findings are validated through a series of experiments involving weighted graphs, as well as text and image data transformed into general graph representations using appropriate distance metrics.
comment: 16 pages
Multimedia 6
☆ Personalized Playback Technology: How Short Video Services Create Excellent User Experience
Short-form video content has become increasingly popular and influential in recent years. Its concise yet engaging format aligns well with todays' fast-paced and on-the-go lifestyles, making it a dominating trend in the digital world. As one of the front runners in the short video platform space, ByteDance has been highly successful in delivering a one-of-a-kind short video experience and attracting billions of users worldwide. One key contributing factor is its advanced end-to-end personalized short video playback technology, where we pioneered and developed the new technical field over the past five years to optimize user experience. This paper introduces the major concepts and methodologies of this personalized video playback technology that distinguish it from traditional multimedia technologies. More details, including goal setting, iterative process, modeling, experimental methods and required supporting systems, are also provided to encourage deeper research in this area.
♻ ☆ AIM 2024 Challenge on Compressed Video Quality Assessment: Methods and Results
Video quality assessment (VQA) is a crucial task in the development of video compression standards, as it directly impacts the viewer experience. This paper presents the results of the Compressed Video Quality Assessment challenge, held in conjunction with the Advances in Image Manipulation (AIM) workshop at ECCV 2024. The challenge aimed to evaluate the performance of VQA methods on a diverse dataset of 459 videos, encoded with 14 codecs of various compression standards (AVC/H.264, HEVC/H.265, AV1, and VVC/H.266) and containing a comprehensive collection of compression artifacts. To measure the methods performance, we employed traditional correlation coefficients between their predictions and subjective scores, which were collected via large-scale crowdsourced pairwise human comparisons. For training purposes, participants were provided with the Compressed Video Quality Assessment Dataset (CVQAD), a previously developed dataset of 1022 videos. Up to 30 participating teams registered for the challenge, while we report the results of 6 teams, which submitted valid final solutions and code for reproducing the results. Moreover, we calculated and present the performance of state-of-the-art VQA methods on the developed dataset, providing a comprehensive benchmark for future research. The dataset, results, and online leaderboard are publicly available at https://challenges.videoprocessing.ai/challenges/compressedvideo-quality-assessment.html.
♻ ☆ WavTokenizer: an Efficient Acoustic Discrete Codec Tokenizer for Audio Language Modeling
Language models have been effectively applied to modeling natural signals, such as images, video, speech, and audio. A crucial component of these models is the codec tokenizer, which compresses high-dimensional natural signals into lower-dimensional discrete tokens. In this paper, we introduce WavTokenizer, which offers several advantages over previous SOTA acoustic codec models in the audio domain: 1)extreme compression. By compressing the layers of quantizers and the temporal dimension of the discrete codec, one-second audio of 24kHz sampling rate requires only a single quantizer with 40 or 75 tokens. 2)improved subjective quality. Despite the reduced number of tokens, WavTokenizer achieves state-of-the-art reconstruction quality with outstanding UTMOS scores and inherently contains richer semantic information. Specifically, we achieve these results by designing a broader VQ space, extended contextual windows, and improved attention networks, as well as introducing a powerful multi-scale discriminator and an inverse Fourier transform structure. We conducted extensive reconstruction experiments in the domains of speech, audio, and music. WavTokenizer exhibited strong performance across various objective and subjective metrics compared to state-of-the-art models. We also tested semantic information, VQ utilization, and adaptability to generative models. Comprehensive ablation studies confirm the necessity of each module in WavTokenizer. The related code, demos, and pre-trained models are available at https://github.com/jishengpeng/WavTokenizer.
comment: Working in progress
♻ ☆ Reducing Hallucinations in Vision-Language Models via Latent Space Steering
Hallucination poses a challenge to the deployment of large vision-language models (LVLMs) in applications. Unlike in large language models (LLMs), hallucination in LVLMs often arises from misalignments between visual inputs and textual outputs. This paper investigates the underlying mechanisms of hallucination, focusing on the unique structure of LVLMs that distinguishes them from large language models (LLMs). We identify that hallucinations often arise from the sensitivity of text decoders to vision inputs, a natural phenomenon when image encoders and text decoders are pre-trained separately. Inspired by this, we introduce Visual and Textual Intervention (VTI), a novel technique designed to reduce hallucinations by steering latent space representations during inference to enhance the stability of vision features. As a task-agnostic test-time intervention, VTI can be easily applied to any problem without additional cost. Extensive experiments demonstrate that it can effectively reduce hallucinations and outperform baseline methods across multiple metrics, highlighting the critical role of vision feature stability in LVLMs.
comment: 21 pages
♻ ☆ Designing Network Algorithms via Large Language Models
We introduce NADA, the first framework to autonomously design network algorithms by leveraging the generative capabilities of large language models (LLMs). Starting with an existing algorithm implementation, NADA enables LLMs to create a wide variety of alternative designs in the form of code blocks. It then efficiently identifies the top-performing designs through a series of filtering techniques, minimizing the need for full-scale evaluations and significantly reducing computational costs. Using adaptive bitrate (ABR) streaming as a case study, we demonstrate that NADA produces novel ABR algorithms -- previously unknown to human developers -- that consistently outperform the original algorithm in diverse network environments, including broadband, satellite, 4G, and 5G.
♻ ☆ NVLM: Open Frontier-Class Multimodal LLMs
We introduce NVLM 1.0, a family of frontier-class multimodal large language models (LLMs) that achieve state-of-the-art results on vision-language tasks, rivaling the leading proprietary models (e.g., GPT-4o) and open-access models (e.g., Llama 3-V 405B and InternVL 2). Remarkably, NVLM 1.0 shows improved text-only performance over its LLM backbone after multimodal training. In terms of model design, we perform a comprehensive comparison between decoder-only multimodal LLMs (e.g., LLaVA) and cross-attention-based models (e.g., Flamingo). Based on the strengths and weaknesses of both approaches, we propose a novel architecture that enhances both training efficiency and multimodal reasoning capabilities. Furthermore, we introduce a 1-D tile-tagging design for tile-based dynamic high-resolution images, which significantly boosts performance on multimodal reasoning and OCR-related tasks. Regarding training data, we meticulously curate and provide detailed information on our multimodal pretraining and supervised fine-tuning datasets. Our findings indicate that dataset quality and task diversity are more important than scale, even during the pretraining phase, across all architectures. Notably, we develop production-grade multimodality for the NVLM-1.0 models, enabling them to excel in vision-language tasks while maintaining and even improving text-only performance compared to their LLM backbones. To achieve this, we craft and integrate a high-quality text-only dataset into multimodal training, alongside a substantial amount of multimodal math and reasoning data, leading to enhanced math and coding capabilities across modalities. To advance research in the field, we release the model weights at https://huggingface.co/nvidia/NVLM-D-72B and will open-source the training code for the community soon.
comment: Fixed the typos. For more information, please visit our project page at: https://research.nvidia.com/labs/adlr/NVLM-1
Computation and Language 82
☆ Raising the Stakes: Performance Pressure Improves AI-Assisted Decision Making
AI systems are used in many domains to assist with decision making, and although the potential for AI systems to assist with decision making is much discussed, human-AI collaboration often underperforms. Investigation into why the performance potential is not realized has revealed many factors, including (mis)trust in the AI system and mental models of AI capabilities on subjective tasks. Performance pressure is known to influence human decision making behavior, yet how it interacts with human-AI decision making is understudied. In this work, we show the effects of performance pressure on AI advice reliance when laypeople (Amazon Mechanical Turk crowdworkers) complete a common AI-assisted task (fake review detection) and thus have inherently low performance pressure. We manipulate performance pressure by leveraging people's loss aversion towards potential monetary gains when completing a task. We find that when the stakes are high, people use AI advice more appropriately than when stakes are lower, regardless of the presence of an AI explanation. Furthermore, when the AI system gives incorrect advice, people correctly discount the poor advice more often when the stakes are higher than when they are lower. We conclude by discussing the implications of how performance pressure influences AI-assisted decision making and encourage future research to incorporate performance pressure analysis.
☆ A Theoretical Understanding of Chain-of-Thought: Coherent Reasoning and Error-Aware Demonstration
Few-shot Chain-of-Thought (CoT) prompting has demonstrated strong performance in improving the reasoning capabilities of large language models (LLMs). While theoretical investigations have been conducted to understand CoT, the underlying transformer used in these studies isolates the CoT reasoning process into separated in-context learning steps (Stepwise ICL). In this work, we theoretically show that, compared to Stepwise ICL, the transformer gains better error correction ability and more accurate predictions if the reasoning from earlier steps (Coherent CoT) is integrated. Given that this coherent reasoning changes the behavior of the transformer, we further investigate the sensitivity of the transformer with Coherent CoT when the demonstration examples are corrupted at the inference stage. Our theoretical results indicate that the transformer is more sensitive to errors in intermediate reasoning steps than the final outcome. Building upon this observation, we propose an improvement on CoT by incorporating both correct and incorrect reasoning paths in the demonstration. Our experiments validate the effectiveness of the proposed approach.
☆ Large Body Language Models
As virtual agents become increasingly prevalent in human-computer interaction, generating realistic and contextually appropriate gestures in real-time remains a significant challenge. While neural rendering techniques have made substantial progress with static scripts, their applicability to human-computer interactions remains limited. To address this, we introduce Large Body Language Models (LBLMs) and present LBLM-AVA, a novel LBLM architecture that combines a Transformer-XL large language model with a parallelized diffusion model to generate human-like gestures from multimodal inputs (text, audio, and video). LBLM-AVA incorporates several key components enhancing its gesture generation capabilities, such as multimodal-to-pose embeddings, enhanced sequence-to-sequence mapping with redefined attention mechanisms, a temporal smoothing module for gesture sequence coherence, and an attention-based refinement module for enhanced realism. The model is trained on our large-scale proprietary open-source dataset Allo-AVA. LBLM-AVA achieves state-of-the-art performance in generating lifelike and contextually appropriate gestures with a 30% reduction in Fr\'echet Gesture Distance (FGD), and a 25% improvement in Fr\'echet Inception Distance compared to existing approaches.
☆ Bayesian scaling laws for in-context learning
In-context learning (ICL) is a powerful technique for getting language models to perform complex tasks with no training updates. Prior work has established strong correlations between the number of in-context examples provided and the accuracy of the model's predictions. In this paper, we seek to explain this correlation by showing that ICL approximates a Bayesian learner. This perspective gives rise to a family of novel Bayesian scaling laws for ICL. In experiments with \mbox{GPT-2} models of different sizes, our scaling laws exceed or match existing scaling laws in accuracy while also offering interpretable terms for task priors, learning efficiency, and per-example probabilities. To illustrate the analytic power that such interpretable scaling laws provide, we report on controlled synthetic dataset experiments designed to inform real-world studies of safety alignment. In our experimental protocol, we use SFT to suppress an unwanted existing model capability and then use ICL to try to bring that capability back (many-shot jailbreaking). We then experiment on real-world instruction-tuned LLMs using capabilities benchmarks as well as a new many-shot jailbreaking dataset. In all cases, Bayesian scaling laws accurately predict the conditions under which ICL will cause the suppressed behavior to reemerge, which sheds light on the ineffectiveness of post-training at increasing LLM safety.
comment: 10 pages main text, 26 pages total
☆ AUTALIC: A Dataset for Anti-AUTistic Ableist Language In Context
As our understanding of autism and ableism continues to increase, so does our understanding of ableist language towards autistic people. Such language poses a significant challenge in NLP research due to its subtle and context-dependent nature. Yet, detecting anti-autistic ableist language remains underexplored, with existing NLP tools often failing to capture its nuanced expressions. We present AUTALIC, the first benchmark dataset dedicated to the detection of anti-autistic ableist language in context, addressing a significant gap in the field. The dataset comprises 2,400 autism-related sentences collected from Reddit, accompanied by surrounding context, and is annotated by trained experts with backgrounds in neurodiversity. Our comprehensive evaluation reveals that current language models, including state-of-the-art LLMs, struggle to reliably identify anti-autistic ableism and align with human judgments, underscoring their limitations in this domain. We publicly release AUTALIC along with the individual annotations which serve as a valuable resource to researchers working on ableism, neurodiversity, and also studying disagreements in annotation tasks. This dataset serves as a crucial step towards developing more inclusive and context-aware NLP systems that better reflect diverse perspectives.
comment: 9 pages, 5 figures, 7 tables
☆ Learning from others' mistakes: Finetuning machine translation models with span-level error annotations
Despite growing interest in incorporating feedback to improve language models, most efforts focus only on sequence-level annotations. In this work, we explore the potential of utilizing fine-grained span-level annotations from offline datasets to improve model quality. We develop a simple finetuning algorithm, called Training with Annotations (TWA), to directly train machine translation models on such annotated data. TWA utilizes targeted span-level error information while also flexibly learning what to penalize within a span. Moreover, TWA considers the overall trajectory of a sequence when deciding which non-error spans to utilize as positive signals. Experiments on English-German and Chinese-English machine translation show that TWA outperforms baselines such as Supervised FineTuning on sequences filtered for quality and Direct Preference Optimization on pairs constructed from the same data.
☆ Allo-AVA: A Large-Scale Multimodal Conversational AI Dataset for Allocentric Avatar Gesture Animation
The scarcity of high-quality, multimodal training data severely hinders the creation of lifelike avatar animations for conversational AI in virtual environments. Existing datasets often lack the intricate synchronization between speech, facial expressions, and body movements that characterize natural human communication. To address this critical gap, we introduce Allo-AVA, a large-scale dataset specifically designed for text and audio-driven avatar gesture animation in an allocentric (third person point-of-view) context. Allo-AVA consists of $\sim$1,250 hours of diverse video content, complete with audio, transcripts, and extracted keypoints. Allo-AVA uniquely maps these keypoints to precise timestamps, enabling accurate replication of human movements (body and facial gestures) in synchronization with speech. This comprehensive resource enables the development and evaluation of more natural, context-aware avatar animation models, potentially transforming applications ranging from virtual reality to digital assistants.
☆ Rulebreakers Challenge: Revealing a Blind Spot in Large Language Models' Reasoning with Formal Logic
Formal logic has long been applied to natural language reasoning, but this approach can sometimes lead to conclusions that, while logically entailed, are factually inconsistent with the premises or are not typically inferred by humans. This study introduces the concept of "rulebreakers", which refers to instances where logical entailment diverges from factually acceptable inference. We present RULEBREAKERS, a novel dataset for evaluating Large Language Models' (LLMs) ability to distinguish between rulebreakers and non-rulebreakers. Focusing on modus tollens and disjunctive syllogism, we assess six state-of-the-art LLMs using RULEBREAKERS, measuring their performance in terms of token-level exact accuracy and model confidence. Our findings reveal that while most models perform poorly to moderately in recognizing rulebreakers, they demonstrate a latent ability to distinguish rulebreakers when assessed by their confidence levels. Further analysis suggests that the failure to recognize rulebreakers is potentially associated with the models' world knowledge and their attention distribution patterns. This research highlights the limitation of LLMs' reasoning capabilities, and contributes to the ongoing discussion on reasoning in LLMs.
comment: Preprint
☆ Natural Language Processing for Human Resources: A Survey
The domain of human resources (HR) includes a broad spectrum of tasks related to natural language processing (NLP) techniques. Recent breakthroughs in NLP have generated significant interest in its industrial applications in this domain and potentially alleviate challenges such as the difficulty of resource acquisition and the complexity of problems. At the same time, the HR domain can also present unique challenges that drive state-of-the-art in NLP research. To support this, we provide NLP researchers and practitioners with an overview of key HR tasks from an NLP perspective, illustrating how specific sub-tasks (e.g., skill extraction) contribute to broader objectives (e.g., job matching). Through this survey, we identify opportunities in NLP for HR and suggest directions for future exploration.
☆ BIG5-CHAT: Shaping LLM Personalities Through Training on Human-Grounded Data
In this work, we tackle the challenge of embedding realistic human personality traits into LLMs. Previous approaches have primarily focused on prompt-based methods that describe the behavior associated with the desired personality traits, suffering from realism and validity issues. To address these limitations, we introduce BIG5-CHAT, a large-scale dataset containing 100,000 dialogues designed to ground models in how humans express their personality in text. Leveraging this dataset, we explore Supervised Fine-Tuning and Direct Preference Optimization as training-based methods to align LLMs more naturally with human personality patterns. Our methods outperform prompting on personality assessments such as BFI and IPIP-NEO, with trait correlations more closely matching human data. Furthermore, our experiments reveal that models trained to exhibit higher conscientiousness, higher agreeableness, lower extraversion, and lower neuroticism display better performance on reasoning tasks, aligning with psychological findings on how these traits impact human cognitive performance. To our knowledge, this work is the first comprehensive study to demonstrate how training-based methods can shape LLM personalities through learning from real human behaviors.
☆ Multi-head Sequence Tagging Model for Grammatical Error Correction
To solve the Grammatical Error Correction (GEC) problem , a mapping between a source sequence and a target one is needed, where the two differ only on few spans. For this reason, the attention has been shifted to the non-autoregressive or sequence tagging models. In which, the GEC has been simplified from Seq2Seq to labeling the input tokens with edit commands chosen from a large edit space. Due to this large number of classes and the limitation of the available datasets, the current sequence tagging approaches still have some issues handling a broad range of grammatical errors just by being laser-focused on one single task. To this end, we simplified the GEC further by dividing it into seven related subtasks: Insertion, Deletion, Merge, Substitution, Transformation, Detection, and Correction, with Correction being our primary focus. A distinct classification head is dedicated to each of these subtasks. the novel multi-head and multi-task learning model is proposed to effectively utilize training data and harness the information from related task training signals. To mitigate the limited number of available training samples, a new denoising autoencoder is used to generate a new synthetic dataset to be used for pretraining. Additionally, a new character-level transformation is proposed to enhance the sequence-to-edit function and improve the model's vocabulary coverage. Our single/ensemble model achieves an F0.5 of 74.4/77.0, and 68.6/69.1 on BEA-19 (test) and CoNLL-14 (test) respectively. Moreover, evaluated on JFLEG test set, the GLEU scores are 61.6 and 61.7 for the single and ensemble models, respectively. It mostly outperforms recently published state-of-the-art results by a considerable margin.
☆ DocEdit-v2: Document Structure Editing Via Multimodal LLM Grounding EMNLP 2024
Document structure editing involves manipulating localized textual, visual, and layout components in document images based on the user's requests. Past works have shown that multimodal grounding of user requests in the document image and identifying the accurate structural components and their associated attributes remain key challenges for this task. To address these, we introduce the DocEdit-v2, a novel framework that performs end-to-end document editing by leveraging Large Multimodal Models (LMMs). It consists of three novel components: (1) Doc2Command, which simultaneously localizes edit regions of interest (RoI) and disambiguates user edit requests into edit commands; (2) LLM-based Command Reformulation prompting to tailor edit commands originally intended for specialized software into edit instructions suitable for generalist LMMs. (3) Moreover, DocEdit-v2 processes these outputs via Large Multimodal Models like GPT-4V and Gemini, to parse the document layout, execute edits on grounded Region of Interest (RoI), and generate the edited document image. Extensive experiments on the DocEdit dataset show that DocEdit-v2 significantly outperforms strong baselines on edit command generation (2-33%), RoI bounding box detection (12-31%), and overall document editing (1-12\%) tasks.
comment: EMNLP 2024 (Main)
☆ Beyond Browsing: API-Based Web Agents
Web browsers are a portal to the internet, where much of human activity is undertaken. Thus, there has been significant research work in AI agents that interact with the internet through web browsing. However, there is also another interface designed specifically for machine interaction with online content: application programming interfaces (APIs). In this paper we ask -- what if we were to take tasks traditionally tackled by browsing agents, and give AI agents access to APIs? To do so, we propose two varieties of agents: (1) an API-calling agent that attempts to perform online tasks through APIs only, similar to traditional coding agents, and (2) a Hybrid Agent that can interact with online data through both web browsing and APIs. In experiments on WebArena, a widely-used and realistic benchmark for web navigation tasks, we find that API-based agents outperform web browsing agents. Hybrid Agents out-perform both others nearly uniformly across tasks, resulting in a more than 20.0% absolute improvement over web browsing alone, achieving a success rate of 35.8%, achiving the SOTA performance among task-agnostic agents. These results strongly suggest that when APIs are available, they present an attractive alternative to relying on web browsing alone.
comment: 24 pages, 6 figures
☆ Comparative Study of Multilingual Idioms and Similes in Large Language Models
This study addresses the gap in the literature concerning the comparative performance of LLMs in interpreting different types of figurative language across multiple languages. By evaluating LLMs using two multilingual datasets on simile and idiom interpretation, we explore the effectiveness of various prompt engineering strategies, including chain-of-thought, few-shot, and English translation prompts. We extend the language of these datasets to Persian as well by building two new evaluation sets. Our comprehensive assessment involves both closed-source (GPT-3.5, GPT-4o mini, Gemini 1.5), and open-source models (Llama 3.1, Qwen2), highlighting significant differences in performance across languages and figurative types. Our findings reveal that while prompt engineering methods are generally effective, their success varies by figurative type, language, and model. We also observe that open-source models struggle particularly with low-resource languages in similes. Additionally, idiom interpretation is nearing saturation for many languages, necessitating more challenging evaluations.
comment: 22 pages, 4 figures
☆ To the Globe (TTG): Towards Language-Driven Guaranteed Travel Planning
Travel planning is a challenging and time-consuming task that aims to find an itinerary which satisfies multiple, interdependent constraints regarding flights, accommodations, attractions, and other travel arrangements. In this paper, we propose To the Globe (TTG), a real-time demo system that takes natural language requests from users, translates it to symbolic form via a fine-tuned Large Language Model, and produces optimal travel itineraries with Mixed Integer Linear Programming solvers. The overall system takes ~5 seconds to reply to the user request with guaranteed itineraries. To train TTG, we develop a synthetic data pipeline that generates user requests, flight and hotel information in symbolic form without human annotations, based on the statistics of real-world datasets, and fine-tune an LLM to translate NL user requests to their symbolic form, which is sent to the symbolic solver to compute optimal itineraries. Our NL-symbolic translation achieves ~91% exact match in a backtranslation metric (i.e., whether the estimated symbolic form of generated natural language matches the groundtruth), and its returned itineraries have a ratio of 0.979 compared to the optimal cost of the ground truth user request. When evaluated by users, TTG achieves consistently high Net Promoter Scores (NPS) of 35-40% on generated itinerary.
☆ Does your LLM truly unlearn? An embarrassingly simple approach to recover unlearned knowledge
Large language models (LLMs) have shown remarkable proficiency in generating text, benefiting from extensive training on vast textual corpora. However, LLMs may also acquire unwanted behaviors from the diverse and sensitive nature of their training data, which can include copyrighted and private content. Machine unlearning has been introduced as a viable solution to remove the influence of such problematic content without the need for costly and time-consuming retraining. This process aims to erase specific knowledge from LLMs while preserving as much model utility as possible. Despite the effectiveness of current unlearning methods, little attention has been given to whether existing unlearning methods for LLMs truly achieve forgetting or merely hide the knowledge, which current unlearning benchmarks fail to detect. This paper reveals that applying quantization to models that have undergone unlearning can restore the "forgotten" information. To thoroughly evaluate this phenomenon, we conduct comprehensive experiments using various quantization techniques across multiple precision levels. We find that for unlearning methods with utility constraints, the unlearned model retains an average of 21\% of the intended forgotten knowledge in full precision, which significantly increases to 83\% after 4-bit quantization. Based on our empirical findings, we provide a theoretical explanation for the observed phenomenon and propose a quantization-robust unlearning strategy to mitigate this intricate issue...
comment: 21 pages, 2 figures
☆ Susu Box or Piggy Bank: Assessing Cultural Commonsense Knowledge between Ghana and the U.S EMNLP 2024
Recent work has highlighted the culturally-contingent nature of commonsense knowledge. We introduce AMAMMER${\epsilon}$, a test set of 525 multiple-choice questions designed to evaluate the commonsense knowledge of English LLMs, relative to the cultural contexts of Ghana and the United States. To create AMAMMER${\epsilon}$, we select a set of multiple-choice questions (MCQs) from existing commonsense datasets and rewrite them in a multi-stage process involving surveys of Ghanaian and U.S. participants. In three rounds of surveys, participants from both pools are solicited to (1) write correct and incorrect answer choices, (2) rate individual answer choices on a 5-point Likert scale, and (3) select the best answer choice from the newly-constructed MCQ items, in a final validation step. By engaging participants at multiple stages, our procedure ensures that participant perspectives are incorporated both in the creation and validation of test items, resulting in high levels of agreement within each pool. We evaluate several off-the-shelf English LLMs on AMAMMER${\epsilon}$. Uniformly, models prefer answers choices that align with the preferences of U.S. annotators over Ghanaian annotators. Additionally, when test items specify a cultural context (Ghana or the U.S.), models exhibit some ability to adapt, but performance is consistently better in U.S. contexts than Ghanaian. As large resources are devoted to the advancement of English LLMs, our findings underscore the need for culturally adaptable models and evaluations to meet the needs of diverse English-speaking populations around the world.
comment: Accepted to EMNLP 2024
☆ Improving Neuron-level Interpretability with White-box Language Models
Neurons in auto-regressive language models like GPT-2 can be interpreted by analyzing their activation patterns. Recent studies have shown that techniques such as dictionary learning, a form of post-hoc sparse coding, enhance this neuron-level interpretability. In our research, we are driven by the goal to fundamentally improve neural network interpretability by embedding sparse coding directly within the model architecture, rather than applying it as an afterthought. In our study, we introduce a white-box transformer-like architecture named Coding RAte TransformEr (CRATE), explicitly engineered to capture sparse, low-dimensional structures within data distributions. Our comprehensive experiments showcase significant improvements (up to 103% relative improvement) in neuron-level interpretability across a variety of evaluation metrics. Detailed investigations confirm that this enhanced interpretability is steady across different layers irrespective of the model size, underlining CRATE's robust performance in enhancing neural network interpretability. Further analysis shows that CRATE's increased interpretability comes from its enhanced ability to consistently and distinctively activate on relevant tokens. These findings point towards a promising direction for creating white-box foundation models that excel in neuron-level interpretation.
☆ Enhancing Multimodal Affective Analysis with Learned Live Comment Features
Live comments, also known as Danmaku, are user-generated messages that are synchronized with video content. These comments overlay directly onto streaming videos, capturing viewer emotions and reactions in real-time. While prior work has leveraged live comments in affective analysis, its use has been limited due to the relative rarity of live comments across different video platforms. To address this, we first construct the Live Comment for Affective Analysis (LCAffect) dataset which contains live comments for English and Chinese videos spanning diverse genres that elicit a wide spectrum of emotions. Then, using this dataset, we use contrastive learning to train a video encoder to produce synthetic live comment features for enhanced multimodal affective content analysis. Through comprehensive experimentation on a wide range of affective analysis tasks (sentiment, emotion recognition, and sarcasm detection) in both English and Chinese, we demonstrate that these synthetic live comment features significantly improve performance over state-of-the-art methods.
☆ VipAct: Visual-Perception Enhancement via Specialized VLM Agent Collaboration and Tool-use
While vision-language models (VLMs) have demonstrated remarkable performance across various tasks combining textual and visual information, they continue to struggle with fine-grained visual perception tasks that require detailed pixel-level analysis. Effectively eliciting comprehensive reasoning from VLMs on such intricate visual elements remains an open challenge. In this paper, we present VipAct, an agent framework that enhances VLMs by integrating multi-agent collaboration and vision expert models, enabling more precise visual understanding and comprehensive reasoning. VipAct consists of an orchestrator agent, which manages task requirement analysis, planning, and coordination, along with specialized agents that handle specific tasks such as image captioning and vision expert models that provide high-precision perceptual information. This multi-agent approach allows VLMs to better perform fine-grained visual perception tasks by synergizing planning, reasoning, and tool use. We evaluate VipAct on benchmarks featuring a diverse set of visual perception tasks, with experimental results demonstrating significant performance improvements over state-of-the-art baselines across all tasks. Furthermore, comprehensive ablation studies reveal the critical role of multi-agent collaboration in eliciting more detailed System-2 reasoning and highlight the importance of image input for task planning. Additionally, our error analysis identifies patterns of VLMs' inherent limitations in visual perception, providing insights into potential future improvements. VipAct offers a flexible and extensible framework, paving the way for more advanced visual perception systems across various real-world applications.
☆ LLM-based Optimization of Compound AI Systems: A Survey
In a compound AI system, components such as an LLM call, a retriever, a code interpreter, or tools are interconnected. The system's behavior is primarily driven by parameters such as instructions or tool definitions. Recent advancements enable end-to-end optimization of these parameters using an LLM. Notably, leveraging an LLM as an optimizer is particularly efficient because it avoids gradient computation and can generate complex code and instructions. This paper presents a survey of the principles and emerging trends in LLM-based optimization of compound AI systems. It covers archetypes of compound AI systems, approaches to LLM-based end-to-end optimization, and insights into future directions and broader impacts. Importantly, this survey uses concepts from program analysis to provide a unified view of how an LLM optimizer is prompted to optimize a compound AI system. The exhaustive list of paper is provided at https://github.com/linyuhongg/LLM-based-Optimization-of-Compound-AI-Systems.
☆ KatzBot: Revolutionizing Academic Chatbot for Enhanced Communication
Effective communication within universities is crucial for addressing the diverse information needs of students, alumni, and external stakeholders. However, existing chatbot systems often fail to deliver accurate, context-specific responses, resulting in poor user experiences. In this paper, we present KatzBot, an innovative chatbot powered by KatzGPT, a custom Large Language Model (LLM) fine-tuned on domain-specific academic data. KatzGPT is trained on two university-specific datasets: 6,280 sentence-completion pairs and 7,330 question-answer pairs. KatzBot outperforms established existing open source LLMs, achieving higher accuracy and domain relevance. KatzBot offers a user-friendly interface, significantly enhancing user satisfaction in real-world applications. The source code is publicly available at \url{https://github.com/AiAI-99/katzbot}.
☆ xGen-MM-Vid (BLIP-3-Video): You Only Need 32 Tokens to Represent a Video Even in VLMs
We present xGen-MM-Vid (BLIP-3-Video): a multimodal language model for videos, particularly designed to efficiently capture temporal information over multiple frames. BLIP-3-Video takes advantage of the 'temporal encoder' in addition to the conventional visual tokenizer, which maps a sequence of tokens over multiple frames into a compact set of visual tokens. This enables BLIP3-Video to use much fewer visual tokens than its competing models (e.g., 32 vs. 4608 tokens). We explore different types of temporal encoders, including learnable spatio-temporal pooling as well as sequential models like Token Turing Machines. We experimentally confirm that BLIP-3-Video obtains video question-answering accuracies comparable to much larger state-of-the-art models (e.g., 34B), while being much smaller (i.e., 4B) and more efficient by using fewer visual tokens. The project website is at https://www.salesforceairesearch.com/opensource/xGen-MM-Vid/index.html
☆ CompassJudger-1: All-in-one Judge Model Helps Model Evaluation and Evolution
Efficient and accurate evaluation is crucial for the continuous improvement of large language models (LLMs). Among various assessment methods, subjective evaluation has garnered significant attention due to its superior alignment with real-world usage scenarios and human preferences. However, human-based evaluations are costly and lack reproducibility, making precise automated evaluators (judgers) vital in this process. In this report, we introduce \textbf{CompassJudger-1}, the first open-source \textbf{all-in-one} judge LLM. CompassJudger-1 is a general-purpose LLM that demonstrates remarkable versatility. It is capable of: 1. Performing unitary scoring and two-model comparisons as a reward model; 2. Conducting evaluations according to specified formats; 3. Generating critiques; 4. Executing diverse tasks like a general LLM. To assess the evaluation capabilities of different judge models under a unified setting, we have also established \textbf{JudgerBench}, a new benchmark that encompasses various subjective evaluation tasks and covers a wide range of topics. CompassJudger-1 offers a comprehensive solution for various evaluation tasks while maintaining the flexibility to adapt to diverse requirements. Both CompassJudger and JudgerBench are released and available to the research community athttps://github.com/open-compass/CompassJudger. We believe that by open-sourcing these tools, we can foster collaboration and accelerate progress in LLM evaluation methodologies.
comment: Technical Report, Code and Models: https://github.com/open-compass/CompassJudger
☆ Can Knowledge Editing Really Correct Hallucinations?
Large Language Models (LLMs) suffer from hallucinations, referring to the non-factual information in generated content, despite their superior capacities across tasks. Meanwhile, knowledge editing has been developed as a new popular paradigm to correct the erroneous factual knowledge encoded in LLMs with the advantage of avoiding retraining from scratch. However, one common issue of existing evaluation datasets for knowledge editing is that they do not ensure LLMs actually generate hallucinated answers to the evaluation questions before editing. When LLMs are evaluated on such datasets after being edited by different techniques, it is hard to directly adopt the performance to assess the effectiveness of different knowledge editing methods in correcting hallucinations. Thus, the fundamental question remains insufficiently validated: Can knowledge editing really correct hallucinations in LLMs? We proposed HalluEditBench to holistically benchmark knowledge editing methods in correcting real-world hallucinations. First, we rigorously construct a massive hallucination dataset with 9 domains, 26 topics and more than 6,000 hallucinations. Then, we assess the performance of knowledge editing methods in a holistic way on five dimensions including Efficacy, Generalization, Portability, Locality, and Robustness. Through HalluEditBench, we have provided new insights into the potentials and limitations of different knowledge editing methods in correcting hallucinations, which could inspire future improvements and facilitate the progress in the field of knowledge editing.
comment: The first two authors contributed equally to this work. The main paper is 10 pages long, with 35 pages total. The code, results, dataset, and additional resources are available on the project website: https://llm-editing.github.io/
☆ Analyzing Context Contributions in LLM-based Machine Translation
Large language models (LLMs) have achieved state-of-the-art performance in machine translation (MT) and demonstrated the ability to leverage in-context learning through few-shot examples. However, the mechanisms by which LLMs use different parts of the input context remain largely unexplored. In this work, we provide a comprehensive analysis of context utilization in MT, studying how LLMs use various context parts, such as few-shot examples and the source text, when generating translations. We highlight several key findings: (1) the source part of few-shot examples appears to contribute more than its corresponding targets, irrespective of translation direction; (2) finetuning LLMs with parallel data alters the contribution patterns of different context parts; and (3) there is a positional bias where earlier few-shot examples have higher contributions to the translated sequence. Finally, we demonstrate that inspecting anomalous context contributions can potentially uncover pathological translations, such as hallucinations. Our findings shed light on the internal workings of LLM-based MT which go beyond those known for standard encoder-decoder MT models.
☆ ToW: Thoughts of Words Improve Reasoning in Large Language Models
We introduce thoughts of words (ToW), a novel training-time data-augmentation method for next-word prediction. ToW views next-word prediction as a core reasoning task and injects fine-grained thoughts explaining what the next word should be and how it is related to the previous contexts in pre-training texts. Our formulation addresses two fundamental drawbacks of existing next-word prediction learning schemes: they induce factual hallucination and are inefficient for models to learn the implicit reasoning processes in raw texts. While there are many ways to acquire such thoughts of words, we explore the first step of acquiring ToW annotations through distilling from larger models. After continual pre-training with only 70K ToW annotations, we effectively improve models' reasoning performances by 7% to 9% on average and reduce model hallucination by up to 10%. At the same time, ToW is entirely agnostic to tasks and applications, introducing no additional biases on labels or semantics.
☆ Sketch2Code: Evaluating Vision-Language Models for Interactive Web Design Prototyping
Sketches are a natural and accessible medium for UI designers to conceptualize early-stage ideas. However, existing research on UI/UX automation often requires high-fidelity inputs like Figma designs or detailed screenshots, limiting accessibility and impeding efficient design iteration. To bridge this gap, we introduce Sketch2Code, a benchmark that evaluates state-of-the-art Vision Language Models (VLMs) on automating the conversion of rudimentary sketches into webpage prototypes. Beyond end-to-end benchmarking, Sketch2Code supports interactive agent evaluation that mimics real-world design workflows, where a VLM-based agent iteratively refines its generations by communicating with a simulated user, either passively receiving feedback instructions or proactively asking clarification questions. We comprehensively analyze ten commercial and open-source models, showing that Sketch2Code is challenging for existing VLMs; even the most capable models struggle to accurately interpret sketches and formulate effective questions that lead to steady improvement. Nevertheless, a user study with UI/UX experts reveals a significant preference for proactive question-asking over passive feedback reception, highlighting the need to develop more effective paradigms for multi-turn conversational agents.
comment: preprint, 9 pages
☆ Building A Coding Assistant via the Retrieval-Augmented Language Model
Pretrained language models have shown strong effectiveness in code-related tasks, such as code retrieval, code generation, code summarization, and code completion tasks. In this paper, we propose COde assistaNt viA retrieval-augmeNted language model (CONAN), which aims to build a code assistant by mimicking the knowledge-seeking behaviors of humans during coding. Specifically, it consists of a code structure aware retriever (CONAN-R) and a dual-view code representation-based retrieval-augmented generation model (CONAN-G). CONAN-R pretrains CodeT5 using Code-Documentation Alignment and Masked Entity Prediction tasks to make language models code structure-aware and learn effective representations for code snippets and documentation. Then CONAN-G designs a dual-view code representation mechanism for implementing a retrieval-augmented code generation model. CONAN-G regards the code documentation descriptions as prompts, which help language models better understand the code semantics. Our experiments show that CONAN achieves convincing performance on different code generation tasks and significantly outperforms previous retrieval augmented code generation models. Our further analyses show that CONAN learns tailored representations for both code snippets and documentation by aligning code-documentation data pairs and capturing structural semantics by masking and predicting entities in the code data. Additionally, the retrieved code snippets and documentation provide necessary information from both program language and natural language to assist the code generation process. CONAN can also be used as an assistant for Large Language Models (LLMs), providing LLMs with external knowledge in shorter code document lengths to improve their effectiveness on various code tasks. It shows the ability of CONAN to extract necessary information and help filter out the noise from retrieved code documents.
☆ On Creating an English-Thai Code-switched Machine Translation in Medical Domain
Machine translation (MT) in the medical domain plays a pivotal role in enhancing healthcare quality and disseminating medical knowledge. Despite advancements in English-Thai MT technology, common MT approaches often underperform in the medical field due to their inability to precisely translate medical terminologies. Our research prioritizes not merely improving translation accuracy but also maintaining medical terminology in English within the translated text through code-switched (CS) translation. We developed a method to produce CS medical translation data, fine-tuned a CS translation model with this data, and evaluated its performance against strong baselines, such as Google Neural Machine Translation (NMT) and GPT-3.5/GPT-4. Our model demonstrated competitive performance in automatic metrics and was highly favored in human preference evaluations. Our evaluation result also shows that medical professionals significantly prefer CS translations that maintain critical English terms accurately, even if it slightly compromises fluency. Our code and test set are publicly available https://github.com/preceptorai-org/NLLB_CS_EM_NLP2024.
Pre-training Distillation for Large Language Models: A Design Space Exploration
Knowledge distillation (KD) aims to transfer knowledge from a large teacher model to a smaller student model. Previous work applying KD in the field of large language models (LLMs) typically focused on the post-training phase, where the student LLM learns directly from instructions and corresponding responses generated by the teacher model. In this paper, we extend KD to the pre-training phase of LLMs, named pre-training distillation (PD). We first conduct a preliminary experiment using GLM-4-9B as the teacher LLM to distill a 1.9B parameter student LLM, validating the effectiveness of PD. Considering the key impact factors of distillation, we systematically explore the design space of pre-training distillation across four aspects: logits processing, loss selection, scaling law, and offline or online logits. We conduct extensive experiments to explore the design space of pre-training distillation and find better configurations and interesting conclusions, such as larger student LLMs generally benefiting more from pre-training distillation, while a larger teacher LLM does not necessarily guarantee better results. We hope our exploration of the design space will inform future practices in pre-training distillation.
☆ Compute-Constrained Data Selection
Data selection can reduce the amount of training data needed to finetune LLMs; however, the efficacy of data selection scales directly with its compute. Motivated by the practical challenge of compute-constrained finetuning, we consider the setting in which both the cost of selecting data and training are budgeted for. We first formalize the problem of data selection with a cost-aware utility function, and model the data selection problem as trading off initial-selection cost for training gain. We run a comprehensive sweep of experiments across multiple tasks, varying compute budget by scaling finetuning tokens, model sizes, and data selection compute. These experiments show the validity of this model in real-world experiments. Interestingly we find that many powerful data selection methods are almost never compute-optimal, and that cheaper data selection alternatives dominate both from a theoretical and empirical perspective.
☆ CoT-TL: Low-Resource Temporal Knowledge Representation of Planning Instructions Using Chain-of-Thought Reasoning IROS 2024
Autonomous agents often face the challenge of interpreting uncertain natural language instructions for planning tasks. Representing these instructions as Linear Temporal Logic (LTL) enables planners to synthesize actionable plans. We introduce CoT-TL, a data-efficient in-context learning framework for translating natural language specifications into LTL representations. CoT-TL addresses the limitations of large language models, which typically rely on extensive fine-tuning data, by extending chain-of-thought reasoning and semantic roles to align with the requirements of formal logic creation. This approach enhances the transparency and rationale behind LTL generation, fostering user trust. CoT-TL achieves state-of-the-art accuracy across three diverse datasets in low-data scenarios, outperforming existing methods without fine-tuning or intermediate translations. To improve reliability and minimize hallucinations, we incorporate model checking to validate the syntax of the generated LTL output. We further demonstrate CoT-TL's effectiveness through ablation studies and evaluations on unseen LTL structures and formulas in a new dataset. Finally, we validate CoT-TL's practicality by integrating it into a QuadCopter for multi-step drone planning based on natural language instructions.
comment: Accepted for publication in Proceedings of the 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2024), Abu Dhabi 14-18 October 2024
☆ Systematic Review: Text Processing Algorithms in Machine Learning and Deep Learning for Mental Health Detection on Social Media
The global rise in depression necessitates innovative detection methods for early intervention. Social media provides a unique opportunity to identify depression through user-generated posts. This systematic review evaluates machine learning (ML) models for depression detection on social media, focusing on biases and methodological challenges throughout the ML lifecycle. A search of PubMed, IEEE Xplore, and Google Scholar identified 47 relevant studies published after 2010. The Prediction model Risk Of Bias ASsessment Tool (PROBAST) was utilized to assess methodological quality and risk of bias. Significant biases impacting model reliability and generalizability were found. There is a predominant reliance on Twitter (63.8%) and English-language content (over 90%), with most studies focusing on users from the United States and Europe. Non-probability sampling methods (approximately 80%) limit representativeness. Only 23% of studies explicitly addressed linguistic nuances like negations, crucial for accurate sentiment analysis. Inconsistent hyperparameter tuning was observed, with only 27.7% properly tuning models. About 17% did not adequately partition data into training, validation, and test sets, risking overfitting. While 74.5% used appropriate evaluation metrics for imbalanced data, others relied on accuracy without addressing class imbalance, potentially skewing results. Reporting transparency varied, often lacking critical methodological details. These findings highlight the need to diversify data sources, standardize preprocessing protocols, ensure consistent model development practices, address class imbalance, and enhance reporting transparency. By overcoming these challenges, future research can develop more robust and generalizable ML models for depression detection on social media, contributing to improved mental health outcomes globally.
☆ Information for Conversation Generation: Proposals Utilising Knowledge Graphs ISWC 2024
LLMs are frequently used tools for conversational generation. Without additional information LLMs can generate lower quality responses due to lacking relevant content and hallucinations, as well as the perception of poor emotional capability, and an inability to maintain a consistent character. Knowledge graphs are commonly used forms of external knowledge and may provide solutions to these challenges. This paper introduces three proposals, utilizing knowledge graphs to enhance LLM generation. Firstly, dynamic knowledge graph embeddings and recommendation could allow for the integration of new information and the selection of relevant knowledge for response generation. Secondly, storing entities with emotional values as additional features may provide knowledge that is better emotionally aligned with the user input. Thirdly, integrating character information through narrative bubbles would maintain character consistency, as well as introducing a structure that would readily incorporate new information.
comment: 7 pages with citations, 1 figure, accepted to the ISWC 2024 Special Session
☆ Contamination Report for Multilingual Benchmarks
Benchmark contamination refers to the presence of test datasets in Large Language Model (LLM) pre-training or post-training data. Contamination can lead to inflated scores on benchmarks, compromising evaluation results and making it difficult to determine the capabilities of models. In this work, we study the contamination of popular multilingual benchmarks in LLMs that support multiple languages. We use the Black Box test to determine whether $7$ frequently used multilingual benchmarks are contaminated in $7$ popular open and closed LLMs and find that almost all models show signs of being contaminated with almost all the benchmarks we test. Our findings can help the community determine the best set of benchmarks to use for multilingual evaluation.
comment: 11 pages, 2 tables
☆ RM-Bench: Benchmarking Reward Models of Language Models with Subtlety and Style
Reward models are critical in techniques like Reinforcement Learning from Human Feedback (RLHF) and Inference Scaling Laws, where they guide language model alignment and select optimal responses. Despite their importance, existing reward model benchmarks often evaluate models by asking them to distinguish between responses generated by models of varying power. However, this approach fails to assess reward models on subtle but critical content changes and variations in style, resulting in a low correlation with policy model performance. To this end, we introduce RM-Bench, a novel benchmark designed to evaluate reward models based on their sensitivity to subtle content differences and resistance to style biases. Extensive experiments demonstrate that RM-Bench strongly correlates with policy model performance, making it a reliable reference for selecting reward models to align language models effectively. We evaluate nearly 40 reward models on RM-Bench. Our results reveal that even state-of-the-art models achieve an average performance of only 46.6%, which falls short of random-level accuracy (50%) when faced with style bias interference. These findings highlight the significant room for improvement in current reward models. Related code and data are available at https://github.com/THU-KEG/RM-Bench.
☆ MagicPIG: LSH Sampling for Efficient LLM Generation
Large language models (LLMs) with long context windows have gained significant attention. However, the KV cache, stored to avoid re-computation, becomes a bottleneck. Various dynamic sparse or TopK-based attention approximation methods have been proposed to leverage the common insight that attention is sparse. In this paper, we first show that TopK attention itself suffers from quality degradation in certain downstream tasks because attention is not always as sparse as expected. Rather than selecting the keys and values with the highest attention scores, sampling with theoretical guarantees can provide a better estimation for attention output. To make the sampling-based approximation practical in LLM generation, we propose MagicPIG, a heterogeneous system based on Locality Sensitive Hashing (LSH). MagicPIG significantly reduces the workload of attention computation while preserving high accuracy for diverse tasks. MagicPIG stores the LSH hash tables and runs the attention computation on the CPU, which allows it to serve longer contexts and larger batch sizes with high approximation accuracy. MagicPIG can improve decoding throughput by $1.9\sim3.9\times$ across various GPU hardware and achieve 110ms decoding latency on a single RTX 4090 for Llama-3.1-8B-Instruct model with a context of 96k tokens. The code is available at \url{https://github.com/Infini-AI-Lab/MagicPIG}.
☆ Exploring Pretraining via Active Forgetting for Improving Cross Lingual Transfer for Decoder Language Models
Large Language Models (LLMs) demonstrate exceptional capabilities in a multitude of NLP tasks. However, the efficacy of such models to languages other than English is often limited. Prior works have shown that encoder-only models such as BERT or XLM-RoBERTa show impressive cross lingual transfer of their capabilities from English to other languages. In this work, we propose a pretraining strategy that uses active forgetting to achieve similar cross lingual transfer in decoder-only LLMs. We show that LLMs pretrained with active forgetting are highly effective when adapting to new and unseen languages. Through extensive experimentation, we find that LLMs pretrained with active forgetting are able to learn better multilingual representations which translates to better performance in many downstream tasks.
comment: 12 pages, 11 tables, 12 figures
☆ Beyond Filtering: Adaptive Image-Text Quality Enhancement for MLLM Pretraining
Multimodal large language models (MLLMs) have made significant strides by integrating visual and textual modalities. A critical factor in training MLLMs is the quality of image-text pairs within multimodal pretraining datasets. However, $\textit {de facto}$ filter-based data quality enhancement paradigms often discard a substantial portion of high-quality image data due to inadequate semantic alignment between images and texts, leading to inefficiencies in data utilization and scalability. In this paper, we propose the Adaptive Image-Text Quality Enhancer (AITQE), a model that dynamically assesses and enhances the quality of image-text pairs. AITQE employs a text rewriting mechanism for low-quality pairs and incorporates a negative sample learning strategy to improve evaluative capabilities by integrating deliberately selected low-quality samples during training. Unlike prior approaches that significantly alter text distributions, our method minimally adjusts text to preserve data volume while enhancing quality. Experimental results demonstrate that AITQE surpasses existing methods on various benchmark, effectively leveraging raw data and scaling efficiently with increasing data volumes. We hope our work will inspire future works. The code and model are available at: https://github.com/hanhuang22/AITQE.
☆ From Tokens to Materials: Leveraging Language Models for Scientific Discovery
Exploring the predictive capabilities of language models in material science is an ongoing interest. This study investigates the application of language model embeddings to enhance material property prediction in materials science. By evaluating various contextual embedding methods and pre-trained models, including Bidirectional Encoder Representations from Transformers (BERT) and Generative Pre-trained Transformers (GPT), we demonstrate that domain-specific models, particularly MatBERT significantly outperform general-purpose models in extracting implicit knowledge from compound names and material properties. Our findings reveal that information-dense embeddings from the third layer of MatBERT, combined with a context-averaging approach, offer the most effective method for capturing material-property relationships from the scientific literature. We also identify a crucial "tokenizer effect," highlighting the importance of specialized text processing techniques that preserve complete compound names while maintaining consistent token counts. These insights underscore the value of domain-specific training and tokenization in materials science applications and offer a promising pathway for accelerating the discovery and development of new materials through AI-driven approaches.
☆ GenAI Assisting Medical Training
Medical procedures such as venipuncture and cannulation are essential for nurses and require precise skills. Learning this skill, in turn, is a challenge for educators due to the number of teachers per class and the complexity of the task. The study aims to help students with skill acquisition and alleviate the educator's workload by integrating generative AI methods to provide real-time feedback on medical procedures such as venipuncture and cannulation.
comment: 2 pages, 2 figures
☆ Sparkle: Mastering Basic Spatial Capabilities in Vision Language Models Elicits Generalization to Composite Spatial Reasoning
Vision language models (VLMs) have demonstrated impressive performance across a wide range of downstream tasks. However, their proficiency in spatial reasoning remains limited, despite its crucial role in tasks involving navigation and interaction with physical environments. Specifically, much of the spatial reasoning in these tasks occurs in two-dimensional (2D) environments, and our evaluation reveals that state-of-the-art VLMs frequently generate implausible and incorrect responses to composite spatial reasoning problems, including simple pathfinding tasks that humans can solve effortlessly at a glance. To address this, we explore an effective approach to enhance 2D spatial reasoning within VLMs by training the model on basic spatial capabilities. We begin by disentangling the key components of 2D spatial reasoning: direction comprehension, distance estimation, and localization. Our central hypothesis is that mastering these basic spatial capabilities can significantly enhance a model's performance on composite spatial tasks requiring advanced spatial understanding and combinatorial problem-solving. To investigate this hypothesis, we introduce Sparkle, a framework that fine-tunes VLMs on these three basic spatial capabilities by synthetic data generation and targeted supervision to form an instruction dataset for each capability. Our experiments demonstrate that VLMs fine-tuned with Sparkle achieve significant performance gains, not only in the basic tasks themselves but also in generalizing to composite and out-of-distribution spatial reasoning tasks (e.g., improving from 13.5% to 40.0% on the shortest path problem). These findings underscore the effectiveness of mastering basic spatial capabilities in enhancing composite spatial problem-solving, offering insights for improving VLMs' spatial reasoning capabilities.
☆ Limpeh ga li gong: Challenges in Singlish Annotations
Singlish, or Colloquial Singapore English, is a language formed from oral and social communication within multicultural Singapore. In this work, we work on a fundamental Natural Language Processing (NLP) task: Parts-Of-Speech (POS) tagging of Singlish sentences. For our analysis, we build a parallel Singlish dataset containing direct English translations and POS tags, with translation and POS annotation done by native Singlish speakers. Our experiments show that automatic transition- and transformer- based taggers perform with only $\sim 80\%$ accuracy when evaluated against human-annotated POS labels, suggesting that there is indeed room for improvement on computation analysis of the language. We provide an exposition of challenges in Singlish annotation: its inconsistencies in form and semantics, the highly context-dependent particles of the language, its structural unique expressions, and the variation of the language on different mediums. Our task definition, resultant labels and results reflects the challenges in analysing colloquial languages formulated from a variety of dialects, and paves the way for future studies beyond POS tagging.
☆ A Troublemaker with Contagious Jailbreak Makes Chaos in Honest Towns
With the development of large language models, they are widely used as agents in various fields. A key component of agents is memory, which stores vital information but is susceptible to jailbreak attacks. Existing research mainly focuses on single-agent attacks and shared memory attacks. However, real-world scenarios often involve independent memory. In this paper, we propose the Troublemaker Makes Chaos in Honest Town (TMCHT) task, a large-scale, multi-agent, multi-topology text-based attack evaluation framework. TMCHT involves one attacker agent attempting to mislead an entire society of agents. We identify two major challenges in multi-agent attacks: (1) Non-complete graph structure, (2) Large-scale systems. We attribute these challenges to a phenomenon we term toxicity disappearing. To address these issues, we propose an Adversarial Replication Contagious Jailbreak (ARCJ) method, which optimizes the retrieval suffix to make poisoned samples more easily retrieved and optimizes the replication suffix to make poisoned samples have contagious ability. We demonstrate the superiority of our approach in TMCHT, with 23.51%, 18.95%, and 52.93% improvements in line topology, star topology, and 100-agent settings. Encourage community attention to the security of multi-agent systems.
☆ Pangea: A Fully Open Multilingual Multimodal LLM for 39 Languages
Despite recent advances in multimodal large language models (MLLMs), their development has predominantly focused on English- and western-centric datasets and tasks, leaving most of the world's languages and diverse cultural contexts underrepresented. This paper introduces Pangea, a multilingual multimodal LLM trained on PangeaIns, a diverse 6M instruction dataset spanning 39 languages. PangeaIns features: 1) high-quality English instructions, 2) carefully machine-translated instructions, and 3) culturally relevant multimodal tasks to ensure cross-cultural coverage. To rigorously assess models' capabilities, we introduce PangeaBench, a holistic evaluation suite encompassing 14 datasets covering 47 languages. Results show that Pangea significantly outperforms existing open-source models in multilingual settings and diverse cultural contexts. Ablation studies further reveal the importance of English data proportions, language popularity, and the number of multimodal training samples on overall performance. We fully open-source our data, code, and trained checkpoints, to facilitate the development of inclusive and robust multilingual MLLMs, promoting equity and accessibility across a broader linguistic and cultural spectrum.
comment: 52 pages, 27 figures
☆ 1-bit AI Infra: Part 1.1, Fast and Lossless BitNet b1.58 Inference on CPUs
Recent advances in 1-bit Large Language Models (LLMs), such as BitNet and BitNet b1.58, present a promising approach to enhancing the efficiency of LLMs in terms of speed and energy consumption. These developments also enable local LLM deployment across a broad range of devices. In this work, we introduce bitnet.cpp, a tailored software stack designed to unlock the full potential of 1-bit LLMs. Specifically, we develop a set of kernels to support fast and lossless inference of ternary BitNet b1.58 LLMs on CPUs. Extensive experiments demonstrate that bitnet.cpp achieves significant speedups, ranging from 2.37x to 6.17x on x86 CPUs and from 1.37x to 5.07x on ARM CPUs, across various model sizes. The code is available at https://github.com/microsoft/BitNet.
☆ A Psycholinguistic Evaluation of Language Models' Sensitivity to Argument Roles
We present a systematic evaluation of large language models' sensitivity to argument roles, i.e., who did what to whom, by replicating psycholinguistic studies on human argument role processing. In three experiments, we find that language models are able to distinguish verbs that appear in plausible and implausible contexts, where plausibility is determined through the relation between the verb and its preceding arguments. However, none of the models capture the same selective patterns that human comprehenders exhibit during real-time verb prediction. This indicates that language models' capacity to detect verb plausibility does not arise from the same mechanism that underlies human real-time sentence processing.
☆ Can Large Audio-Language Models Truly Hear? Tackling Hallucinations with Multi-Task Assessment and Stepwise Audio Reasoning
Recent advancements in large audio-language models (LALMs) have shown impressive capabilities in understanding and reasoning about audio and speech information. However, these models still face challenges, including hallucinating non-existent sound events, misidentifying the order of sound events, and incorrectly attributing sound sources, which undermine their reliability and real-world application. To systematically evaluate these issues, we propose three distinct tasks: object existence, temporal order, and object attribute within audio. These tasks assess the models' comprehension of critical audio information aspects. Our experimental results reveal limitations in these fundamental tasks, underscoring the need for better models in recognizing specific sound events, determining event sequences, and identifying sound sources. To improve performance in these areas, we introduce a multi-turn chain-of-thought approach, which demonstrates significantly improved model performance across the proposed tasks.
comment: 5 pages, 1 figure
☆ Do LLMs write like humans? Variation in grammatical and rhetorical styles
Large language models (LLMs) are capable of writing grammatical text that follows instructions, answers questions, and solves problems. As they have advanced, it has become difficult to distinguish their output from human-written text. While past research has found some differences in surface features such as word choice and punctuation, and developed classifiers to detect LLM output, none has studied the rhetorical styles of LLMs. Using several variants of Llama 3 and GPT-4o, we construct two parallel corpora of human- and LLM-written texts from common prompts. Using Douglas Biber's set of lexical, grammatical, and rhetorical features, we identify systematic differences between LLMs and humans and between different LLMs. These differences persist when moving from smaller models to larger ones, and are larger for instruction-tuned models than base models. This demonstrates that despite their advanced abilities, LLMs struggle to match human styles, and hence more advanced linguistic features can detect patterns in their behavior not previously recognized.
comment: 29 pages, 4 figures, 11 tables
☆ Analysing the Residual Stream of Language Models Under Knowledge Conflicts NeurIPS 2024
Large language models (LLMs) can store a significant amount of factual knowledge in their parameters. However, their parametric knowledge may conflict with the information provided in the context. Such conflicts can lead to undesirable model behaviour, such as reliance on outdated or incorrect information. In this work, we investigate whether LLMs can identify knowledge conflicts and whether it is possible to know which source of knowledge the model will rely on by analysing the residual stream of the LLM. Through probing tasks, we find that LLMs can internally register the signal of knowledge conflict in the residual stream, which can be accurately detected by probing the intermediate model activations. This allows us to detect conflicts within the residual stream before generating the answers without modifying the input or model parameters. Moreover, we find that the residual stream shows significantly different patterns when the model relies on contextual knowledge versus parametric knowledge to resolve conflicts. This pattern can be employed to estimate the behaviour of LLMs when conflict happens and prevent unexpected answers before producing the answers. Our analysis offers insights into how LLMs internally manage knowledge conflicts and provides a foundation for developing methods to control the knowledge selection processes.
comment: Foundation Model Interventions Workshop @ NeurIPS 2024
☆ Fine-Tuning LLMs for Reliable Medical Question-Answering Services ICDM
We present an advanced approach to medical question-answering (QA) services, using fine-tuned Large Language Models (LLMs) to improve the accuracy and reliability of healthcare information. Our study focuses on optimizing models like LLaMA-2 and Mistral, which have shown great promise in delivering precise, reliable medical answers. By leveraging comprehensive datasets, we applied fine-tuning techniques such as rsDoRA+ and ReRAG. rsDoRA+ enhances model performance through a combination of decomposed model weights, varied learning rates for low-rank matrices, and rank stabilization, leading to improved efficiency. ReRAG, which integrates retrieval on demand and question rewriting, further refines the accuracy of the responses. This approach enables healthcare providers to access fast, dependable information, aiding in more efficient decision-making and fostering greater patient trust. Our work highlights the potential of fine-tuned LLMs to significantly improve the quality and accessibility of medical information services, ultimately contributing to better healthcare outcomes for all.
comment: 8 pages, 10 figures, accepted and to be published in the proceedings of 2024 IEEE International Conference on Data Mining Workshops (ICDMW)
☆ Rolling the DICE on Idiomaticity: How LLMs Fail to Grasp Context
Human processing of idioms relies on understanding the contextual sentences in which idioms occur, as well as language-intrinsic features such as frequency and speaker-intrinsic factors like familiarity. While LLMs have shown high performance on idiomaticity detection tasks, this success may be attributed to reasoning shortcuts in existing datasets. To this end, we construct a novel, controlled contrastive dataset designed to test whether LLMs can effectively use context to disambiguate idiomatic meaning. Additionally, we explore how collocational frequency and sentence probability influence model performance. Our findings reveal that LLMs often fail to resolve idiomaticity when it is required to attend to the surrounding context, and that models perform better on sentences that have higher likelihood. The collocational frequency of expressions also impacts performance. We make our code and dataset publicly available.
☆ Surprise! Uniform Information Density Isn't the Whole Story: Predicting Surprisal Contours in Long-form Discourse EMNLP 2024
The Uniform Information Density (UID) hypothesis posits that speakers tend to distribute information evenly across linguistic units to achieve efficient communication. Of course, information rate in texts and discourses is not perfectly uniform. While these fluctuations can be viewed as theoretically uninteresting noise on top of a uniform target, another explanation is that UID is not the only functional pressure regulating information content in a language. Speakers may also seek to maintain interest, adhere to writing conventions, and build compelling arguments. In this paper, we propose one such functional pressure; namely that speakers modulate information rate based on location within a hierarchically-structured model of discourse. We term this the Structured Context Hypothesis and test it by predicting the surprisal contours of naturally occurring discourses extracted from large language models using predictors derived from discourse structure. We find that hierarchical predictors are significant predictors of a discourse's information contour and that deeply nested hierarchical predictors are more predictive than shallow ones. This work takes an initial step beyond UID to propose testable hypotheses for why the information rate fluctuates in predictable ways
comment: EMNLP 2024 (main conference)
♻ ☆ Learning Language Structures through Grounding
Language is highly structured, with syntactic and semantic structures, to some extent, agreed upon by speakers of the same language. With implicit or explicit awareness of such structures, humans can learn and use language efficiently and generalize to sentences that contain unseen words. Motivated by human language learning, in this dissertation, we consider a family of machine learning tasks that aim to learn language structures through grounding. We seek distant supervision from other data sources (i.e., grounds), including but not limited to other modalities (e.g., vision), execution results of programs, and other languages. We demonstrate the potential of this task formulation and advocate for its adoption through three schemes. In Part I, we consider learning syntactic parses through visual grounding. We propose the task of visually grounded grammar induction, present the first models to induce syntactic structures from visually grounded text and speech, and find that the visual grounding signals can help improve the parsing quality over language-only models. As a side contribution, we propose a novel evaluation metric that enables the evaluation of speech parsing without text or automatic speech recognition systems involved. In Part II, we propose two execution-aware methods to map sentences into corresponding semantic structures (i.e., programs), significantly improving compositional generalization and few-shot program synthesis. In Part III, we propose methods that learn language structures from annotations in other languages. Specifically, we propose a method that sets a new state of the art on cross-lingual word alignment. We then leverage the learned word alignments to improve the performance of zero-shot cross-lingual dependency parsing, by proposing a novel substructure-based projection method that preserves structural knowledge learned from the source language.
comment: Ph.D. Thesis
♻ ☆ BenTo: Benchmark Task Reduction with In-Context Transferability
Evaluating large language models (LLMs) is costly: it requires the generation and examination of LLM outputs on a large-scale benchmark of various tasks. This paper investigates how to efficiently reduce the tasks used to benchmark LLMs without affecting the evaluation quality. Our study reveals that task transferability and relevance provide critical information to identify the most representative subset of tasks via optimizing a facility location function. We propose a practically efficient metric for estimating the transferability between two tasks via in-context learning (ICL). By analyzing the pairwise transferability, we can reduce tasks in a modern LLM benchmark (e.g., MMLU or FLAN) to 5% while inducing only a <4% difference to the evaluation on the original benchmark. Compared to prior works, our method is training-free, gradient-free, and highly efficient requiring ICL only.
comment: https://github.com/tianyi-lab/bento
♻ ☆ Superposed Decoding: Multiple Generations from a Single Autoregressive Inference Pass NeurIPS 2024
Many applications today provide users with multiple auto-complete drafts as they type, including GitHub's code completion, Gmail's smart compose, and Apple's messaging auto-suggestions. Under the hood, language models support this by running an autoregressive inference pass to provide a draft. Consequently, providing $k$ drafts to the user requires running an expensive language model $k$ times. To alleviate the computation cost of running $k$ inference passes, we propose Superposed Decoding, a new decoding algorithm that generates $k$ drafts at the computation cost of one autoregressive inference pass. We achieve this by feeding a superposition of the most recent token embeddings from the $k$ drafts as input to the next decoding step of the language model. At every inference step we combine the $k$ drafts with the top-$k$ tokens to get $k^2$ new drafts and cache the $k$ most likely options, using an n-gram interpolation with minimal compute overhead to filter out incoherent generations. Our experiments show that $k$ drafts from Superposed Decoding are at least as coherent and factual as Nucleus Sampling and Greedy Decoding respectively, while being at least $2.44\times$ faster for $k\ge3$. In a compute-normalized setting, user evaluations demonstrably favor text generated by Superposed Decoding over Nucleus Sampling. Superposed Decoding can also be combined with other decoding strategies, resulting in universal coverage gains when scaling inference time compute. Code and more examples open-sourced at https://github.com/RAIVNLab/SuperposedDecoding.
comment: 23 pages, 16 figures, accepted at NeurIPS 2024
♻ ☆ Tokenization and Morphology in Multilingual Language Models: A Comparative Analysis of mT5 and ByT5
Morphology is a crucial factor for multilingual language modeling as it poses direct challenges for tokenization. Here, we seek to understand how tokenization influences the morphological knowledge encoded in multilingual language models. Specifically, we capture the impact of tokenization by contrasting two multilingual language models: mT5 and ByT5. The two models share the same architecture, training objective, and training data and only differ in their tokenization strategies: subword tokenization vs.\@ character-level tokenization. Probing the morphological knowledge encoded in these models on four tasks and 17 languages, our analyses show that the models learn the morphological systems of some languages better than others and that morphological information is encoded in the middle and late layers. Finally, we show that languages with more irregularities benefit more from having a higher share of the pre-training data.
comment: 8 pages
♻ ☆ MiCEval: Unveiling Multimodal Chain of Thought's Quality via Image Description and Reasoning Steps
Multimodal Chain of Thought (MCoT) is a popular prompting strategy for improving the performance of multimodal large language models (MLLMs) across a range of complex reasoning tasks. Despite its popularity, there is a notable absence of automated methods for evaluating the quality of reasoning steps in MCoT. To address this gap, we propose Multimodal Chain-of-Thought Evaluation (MiCEval), a framework designed to assess the correctness of reasoning chains by evaluating the quality of both the description and each reasoning step. The evaluation of the description component focuses on the accuracy of the image descriptions, while the reasoning step evaluates the quality of each step as it is conditionally generated based on the preceding steps. MiCEval is built upon a fine-grained dataset with annotations that rate each step according to correctness, relevance, and informativeness. Extensive experiments on four state-of-the-art MLLMs show that step-wise evaluations using MiCEval align more closely with human judgments compared to existing methods based on cosine similarity or fine-tuning approaches. MiCEval datasets and code can be found in https://github.com/alenai97/MiCEval.
comment: 40 pages
♻ ☆ Language Model Council: Democratically Benchmarking Foundation Models on Highly Subjective Tasks
As Large Language Models (LLMs) continue to evolve, the search for efficient and meaningful evaluation methods is ongoing. Many recent evaluations use LLMs as judges to score outputs from other LLMs, often relying on a single large model like GPT-4o. However, using a single LLM judge is prone to intra-model bias, and many tasks - such as those related to emotional intelligence, creative writing, and persuasiveness - may be too subjective for a single model to judge fairly. We introduce the Language Model Council (LMC), where a group of LLMs collaborate to create tests, respond to them, and evaluate each other's responses to produce a ranking in a democratic fashion. Unlike previous approaches that focus on reducing cost or bias by using a panel of smaller models, our work examines the benefits and nuances of a fully inclusive LLM evaluation system. In a detailed case study on emotional intelligence, we deploy a council of 20 recent LLMs to rank each other on open-ended responses to interpersonal conflicts. Our results show that the LMC produces rankings that are more separable and more robust, and through a user study, we show that they are more consistent with human evaluations than any individual LLM judge. Using all LLMs for judging can be costly, however, so we use Monte Carlo simulations and hand-curated sub-councils to study hypothetical council compositions and discuss the value of the incremental LLM judge.
♻ ☆ DiffNorm: Self-Supervised Normalization for Non-autoregressive Speech-to-speech Translation NeurIPS 2024
Non-autoregressive Transformers (NATs) are recently applied in direct speech-to-speech translation systems, which convert speech across different languages without intermediate text data. Although NATs generate high-quality outputs and offer faster inference than autoregressive models, they tend to produce incoherent and repetitive results due to complex data distribution (e.g., acoustic and linguistic variations in speech). In this work, we introduce DiffNorm, a diffusion-based normalization strategy that simplifies data distributions for training NAT models. After training with a self-supervised noise estimation objective, DiffNorm constructs normalized target data by denoising synthetically corrupted speech features. Additionally, we propose to regularize NATs with classifier-free guidance, improving model robustness and translation quality by randomly dropping out source information during training. Our strategies result in a notable improvement of about +7 ASR-BLEU for English-Spanish (En-Es) and +2 ASR-BLEU for English-French (En-Fr) translations on the CVSS benchmark, while attaining over 14x speedup for En-Es and 5x speedup for En-Fr translations compared to autoregressive baselines.
comment: Accepted at NeurIPS 2024
♻ ☆ Fine-grained and Explainable Factuality Evaluation for Multimodal Summarization
Multimodal summarization aims to generate a concise summary based on the input text and image. However, the existing methods potentially suffer from unfactual output. To evaluate the factuality of multimodal summarization models, we propose two fine-grained and explainable evaluation frameworks (FALLACIOUS) for different application scenarios, i.e. reference-based factuality evaluation framework and reference-free factuality evaluation framework. Notably, the reference-free factuality evaluation framework doesn't need ground truth and hence it has a wider application scenario. To evaluate the effectiveness of the proposed frameworks, we compute the correlation between our frameworks and the other metrics. The experimental results show the effectiveness of our proposed method. We will release our code and dataset via github.
♻ ☆ Pairing Analogy-Augmented Generation with Procedural Memory for Procedural Q&A
Large language models struggle to synthesize disparate pieces of information into a coherent plan when approaching a complex procedural task. In this work, we introduce a novel formalism and structure for such procedural knowledge. Based on this formalism, we present a novel procedural knowledge dataset called LCStep, which we created from LangChain tutorials. To leverage this procedural knowledge to solve new tasks, we propose analogy-augmented generation (AAG), which draws inspiration from the human ability to assimilate past experiences to solve unfamiliar problems. AAG uses a custom procedure memory store to retrieve and adapt specialized domain knowledge to answer new procedural tasks. We demonstrate that AAG outperforms few-shot and RAG baselines on LCStep, RecipeNLG, and CHAMP datasets under a pairwise LLM-based evaluation, corroborated by human evaluation in the case of RecipeNLG.
♻ ☆ Interactive Concept Learning for Uncovering Latent Themes in Large Text Collections ACL
Experts across diverse disciplines are often interested in making sense of large text collections. Traditionally, this challenge is approached either by noisy unsupervised techniques such as topic models, or by following a manual theme discovery process. In this paper, we expand the definition of a theme to account for more than just a word distribution, and include generalized concepts deemed relevant by domain experts. Then, we propose an interactive framework that receives and encodes expert feedback at different levels of abstraction. Our framework strikes a balance between automation and manual coding, allowing experts to maintain control of their study while reducing the manual effort required.
comment: Accepted to Findings of ACL: ACL 2023
♻ ☆ Faster Cascades via Speculative Decoding
Cascades and speculative decoding are two common approaches to improving language models' inference efficiency. Both approaches involve interleaving models of different sizes, but via fundamentally distinct mechanisms: cascades employ a deferral rule that invokes the larger model only for "hard" inputs, while speculative decoding uses speculative execution to primarily invoke the larger model in parallel verification mode. These mechanisms offer different benefits: empirically, cascades offer better cost-quality trade-offs, often even outperforming the large model, while theoretically, speculative decoding offers a guarantee of quality-neutrality. In this paper, we leverage the best of both these approaches by designing new speculative cascading techniques that implement their deferral rule through speculative execution. We characterize the optimal deferral rule for our speculative cascades, and employ a plug-in approximation to the optimal rule. Experiments with Gemma and T5 models on a range of language benchmarks show that our approach yields better cost quality trade-offs than cascading and speculative decoding baselines.
♻ ☆ BPO: Staying Close to the Behavior LLM Creates Better Online LLM Alignment EMNLP 2024
Direct alignment from preferences (DAP) has emerged as a promising paradigm for aligning large language models (LLMs) to human desiderata from pre-collected, offline preference datasets. While recent studies indicate that existing offline DAP methods can directly benefit from online training samples, we highlight the need to develop specific online DAP algorithms to fully harness the power of online training. Specifically, we identify that the learned LLM should adhere to the proximity of the behavior LLM, which collects the training samples. To this end, we propose online Preference Optimization in proximity to the Behavior LLM (BPO), emphasizing the importance of constructing a proper trust region for LLM alignment. We conduct extensive experiments to validate the effectiveness and applicability of our approach by integrating it with various DAP methods, resulting in significant performance improvements across a wide range of tasks when training with the same amount of preference data. Even when only introducing one additional data collection phase, our online BPO improves its offline DAP baseline from 72.0% to 80.2% on TL;DR and from 82.2% to 89.1% on Anthropic Helpfulness in terms of win rate against human reference text.
comment: Wenda Xu and Jiachen Li contributed equally. Accepted by EMNLP 2024
♻ ☆ Are Large Language Models Chameleons? An Attempt to Simulate Social Surveys
Can large language models (LLMs) simulate social surveys? To answer this question, we conducted millions of simulations in which LLMs were asked to answer subjective questions. A comparison of different LLM responses with the European Social Survey (ESS) data suggests that the effect of prompts on bias and variability is fundamental, highlighting major cultural, age, and gender biases. We further discussed statistical methods for measuring the difference between LLM answers and survey data and proposed a novel measure inspired by Jaccard similarity, as LLM-generated responses are likely to have a smaller variance. Our experiments also reveal that it is important to analyze the robustness and variability of prompts before using LLMs to simulate social surveys, as their imitation abilities are approximate at best.
comment: 17 pages
♻ ☆ IoT-Based Preventive Mental Health Using Knowledge Graphs and Standards for Better Well-Being
Sustainable Development Goals (SDGs) give the UN a road map for development with Agenda 2030 as a target. SDG3 "Good Health and Well-Being" ensures healthy lives and promotes well-being for all ages. Digital technologies can support SDG3. Burnout and even depression could be reduced by encouraging better preventive health. Due to the lack of patient knowledge and focus to take care of their health, it is necessary to help patients before it is too late. New trends such as positive psychology and mindfulness are highly encouraged in the USA. Digital Twins (DTs) can help with the continuous monitoring of emotion using physiological signals (e.g., collected via wearables). DTs facilitate monitoring and provide constant health insight to improve quality of life and well-being with better personalization. Healthcare DTs challenges are standardizing data formats, communication protocols, and data exchange mechanisms. As an example, ISO has the ISO/IEC JTC 1/SC 41 Internet of Things (IoT) and DTs Working Group, with standards such as "ISO/IEC 21823-3:2021 IoT - Interoperability for IoT Systems - Part 3 Semantic interoperability", "ISO/IEC CD 30178 - IoT - Data format, value and coding". To achieve those data integration and knowledge challenges, we designed the Mental Health Knowledge Graph (ontology and dataset) to boost mental health. As an example, explicit knowledge is described such as chocolate contains magnesium which is recommended for depression. The Knowledge Graph (KG) acquires knowledge from ontology-based mental health projects classified within the LOV4IoT ontology catalog (Emotion, Depression, and Mental Health). Furthermore, the KG is mapped to standards when possible. Standards from ETSI SmartM2M can be used such as SAREF4EHAW to represent medical devices and sensors, but also ITU/WHO, ISO, W3C, NIST, and IEEE standards relevant to mental health can be considered.
comment: 20 pages, Book chapter, Smart Technologies for Achieving Good Health and Well-Being: Towards Sustainable Development Goal, Taylor & Francis
♻ ☆ Toxicity Detection is NOT all you Need: Measuring the Gaps to Supporting Volunteer Content Moderators
Extensive efforts in automated approaches for content moderation have been focused on developing models to identify toxic, offensive, and hateful content with the aim of lightening the load for moderators. Yet, it remains uncertain whether improvements on those tasks have truly addressed moderators' needs in accomplishing their work. In this paper, we surface gaps between past research efforts that have aimed to provide automation for aspects of content moderation and the needs of volunteer content moderators, regarding identifying violations of various moderation rules. To do so, we conduct a model review on Hugging Face to reveal the availability of models to cover various moderation rules and guidelines from three exemplar forums. We further put state-of-the-art LLMs to the test, evaluating how well these models perform in flagging violations of platform rules from one particular forum. Finally, we conduct a user survey study with volunteer moderators to gain insight into their perspectives on useful moderation models. Overall, we observe a non-trivial gap, as missing developed models and LLMs exhibit moderate to low performance on a significant portion of the rules. Moderators' reports provide guides for future work on developing moderation assistant models.
♻ ☆ The First VoicePrivacy Attacker Challenge Evaluation Plan
The First VoicePrivacy Attacker Challenge is a new kind of challenge organized as part of the VoicePrivacy initiative and supported by ICASSP 2025 as the SP Grand Challenge It focuses on developing attacker systems against voice anonymization, which will be evaluated against a set of anonymization systems submitted to the VoicePrivacy 2024 Challenge. Training, development, and evaluation datasets are provided along with a baseline attacker system. Participants shall develop their attacker systems in the form of automatic speaker verification systems and submit their scores on the development and evaluation data to the organizers. To do so, they can use any additional training data and models, provided that they are openly available and declared before the specified deadline. The metric for evaluation is equal error rate (EER). Results will be presented at the ICASSP 2025 special session to which 5 selected top-ranked participants will be invited to submit and present their challenge systems.
♻ ☆ A Survey on Knowledge Distillation of Large Language Models
In the era of Large Language Models (LLMs), Knowledge Distillation (KD) emerges as a pivotal methodology for transferring advanced capabilities from leading proprietary LLMs, such as GPT-4, to their open-source counterparts like LLaMA and Mistral. Additionally, as open-source LLMs flourish, KD plays a crucial role in both compressing these models, and facilitating their self-improvement by employing themselves as teachers. This paper presents a comprehensive survey of KD's role within the realm of LLM, highlighting its critical function in imparting advanced knowledge to smaller models and its utility in model compression and self-improvement. Our survey is meticulously structured around three foundational pillars: \textit{algorithm}, \textit{skill}, and \textit{verticalization} -- providing a comprehensive examination of KD mechanisms, the enhancement of specific cognitive abilities, and their practical implications across diverse fields. Crucially, the survey navigates the intricate interplay between data augmentation (DA) and KD, illustrating how DA emerges as a powerful paradigm within the KD framework to bolster LLMs' performance. By leveraging DA to generate context-rich, skill-specific training data, KD transcends traditional boundaries, enabling open-source models to approximate the contextual adeptness, ethical alignment, and deep semantic insights characteristic of their proprietary counterparts. This work aims to provide an insightful guide for researchers and practitioners, offering a detailed overview of current methodologies in KD and proposing future research directions. Importantly, we firmly advocate for compliance with the legal terms that regulate the use of LLMs, ensuring ethical and lawful application of KD of LLMs. An associated Github repository is available at https://github.com/Tebmer/Awesome-Knowledge-Distillation-of-LLMs.
comment: 43 pages
♻ ☆ RACCooN: A Versatile Instructional Video Editing Framework with Auto-Generated Narratives
Recent video generative models primarily rely on carefully written text prompts for specific tasks, like inpainting or style editing. They require labor-intensive textual descriptions for input videos, hindering their flexibility to adapt personal/raw videos to user specifications. This paper proposes RACCooN, a versatile and user-friendly video-to-paragraph-to-video generative framework that supports multiple video editing capabilities such as removal, addition, and modification, through a unified pipeline. RACCooN consists of two principal stages: Video-to-Paragraph (V2P) and Paragraph-to-Video (P2V). In the V2P stage, we automatically describe video scenes in well-structured natural language, capturing both the holistic context and focused object details. Subsequently, in the P2V stage, users can optionally refine these descriptions to guide the video diffusion model, enabling various modifications to the input video, such as removing, changing subjects, and/or adding new objects. The proposed approach stands out from other methods through several significant contributions: (1) RACCooN suggests a multi-granular spatiotemporal pooling strategy to generate well-structured video descriptions, capturing both the broad context and object details without requiring complex human annotations, simplifying precise video content editing based on text for users. (2) Our video generative model incorporates auto-generated narratives or instructions to enhance the quality and accuracy of the generated content. (3) RACCooN also plans to imagine new objects in a given video, so users simply prompt the model to receive a detailed video editing plan for complex video editing. The proposed framework demonstrates impressive versatile capabilities in video-to-paragraph generation, video content editing, and can be incorporated into other SoTA video generative models for further enhancement.
comment: The first two authors contribute equally. Project Page: https://raccoon-mllm-gen.github.io/
♻ ☆ Exploring the Potential of Large Language Models for Heterophilic Graphs
Large language models (LLMs) have presented significant opportunities to enhance various machine learning applications, including graph neural networks (GNNs). By leveraging the vast open-world knowledge within LLMs, we can more effectively interpret and utilize textual data to better characterize heterophilic graphs, where neighboring nodes often have different labels. However, existing approaches for heterophilic graphs overlook the rich textual data associated with nodes, which could unlock deeper insights into their heterophilic contexts. In this work, we explore the potential of LLMs for modeling heterophilic graphs and propose a novel two-stage framework: LLM-enhanced edge discriminator and LLM-guided edge reweighting. In the first stage, we fine-tune the LLM to better identify homophilic and heterophilic edges based on the textual content of their nodes. In the second stage, we adaptively manage message propagation in GNNs for different edge types based on node features, structures, and heterophilic or homophilic characteristics. To cope with the computational demands when deploying LLMs in practical scenarios, we further explore model distillation techniques to fine-tune smaller, more efficient models that maintain competitive performance. Extensive experiments validate the effectiveness of our framework, demonstrating the feasibility of using LLMs to enhance node classification on heterophilic graphs.
comment: Under review
♻ ☆ Do Large Language Models Need a Content Delivery Network?
As the use of large language models (LLMs) expands rapidly, so does the range of knowledge needed to supplement various LLM queries. Thus, enabling flexible and efficient injection of new knowledge in LLM inference is critical. Three high-level options exist: (i) embedding the knowledge in LLM's weights (i.e., fine-tuning), (ii) including the knowledge as a part of LLM's text input (i.e., in-context learning), or (iii) injecting the KV caches of the new knowledge to LLM during prefill. This paper argues that, although fine-tuning and in-context learning are popular, using KV caches as the medium of knowledge could simultaneously enable more modular management of knowledge injection and more efficient LLM serving with low cost and fast response. To realize these benefits, we envision a Knowledge Delivery Network (KDN), a new system component in LLM services that dynamically optimizes the storage, transfer, and composition of KV cache across LLM engines and other compute and storage resources. We believe that, just like content delivery networks (CDNs), such as Akamai, enabled the success of the Internet ecosystem through their efficient data delivery, KDNs will be critical to the success of LLM applications through their efficient knowledge delivery. We have open-sourced a KDN prototype at https://github.com/LMCache/LMCache.
♻ ☆ MathGAP: Out-of-Distribution Evaluation on Problems with Arbitrarily Complex Proofs
Large language models (LLMs) can solve arithmetic word problems with high accuracy, but little is known about how well they generalize to problems that are more complex than the ones on which they have been trained. Empirical investigations of such questions are impeded by two major flaws of current evaluations: (i) much of the evaluation data is contaminated, in the sense that it has already been seen during training, and (ii) benchmark datasets do not capture how problem proofs may be arbitrarily complex in various ways. As a step towards addressing these issues, we present a framework for evaluating LLMs on problems with arbitrarily complex arithmetic proofs, called MathGAP. MathGAP generates problems that follow fixed proof specifications -- along with chain-of-thought reasoning annotations -- enabling systematic studies on generalization with respect to arithmetic proof complexity. We apply MathGAP to analyze how in-context learning interacts with generalization to problems that have more complex proofs. We find that among the models tested, most show a significant decrease in performance as proofs get deeper and wider. This effect is more pronounced in complex, nonlinear proof structures, which are challenging even for GPT-4o. Surprisingly, providing in-context examples from the same distribution as the test set is not always beneficial for performance. In particular, zero-shot prompting as well as demonstrating a diverse range of examples that are less complex than the test data sometimes yield similar or higher accuracies.
comment: Preprint
♻ ☆ PromptExp: Multi-granularity Prompt Explanation of Large Language Models
Large Language Models excel in tasks like natural language understanding and text generation. Prompt engineering plays a critical role in leveraging LLM effectively. However, LLMs black-box nature hinders its interpretability and effective prompting engineering. A wide range of model explanation approaches have been developed for deep learning models, However, these local explanations are designed for single-output tasks like classification and regression,and cannot be directly applied to LLMs, which generate sequences of tokens. Recent efforts in LLM explanation focus on natural language explanations, but they are prone to hallucinations and inaccuracies. To address this, we introduce OurTool, a framework for multi-granularity prompt explanations by aggregating token-level insights. OurTool introduces two token-level explanation approaches: 1.an aggregation-based approach combining local explanation techniques, and 2. a perturbation-based approach with novel techniques to evaluate token masking impact. OurTool supports both white-box and black-box explanations and extends explanations to higher granularity levels, enabling flexible analysis. We evaluate OurTool in case studies such as sentiment analysis, showing the perturbation-based approach performs best using semantic similarity to assess perturbation impact. Furthermore, we conducted a user study to confirm OurTool's accuracy and practical value, and demonstrate its potential to enhance LLM interpretability.
comment: 11 pages
♻ ☆ AssistantBench: Can Web Agents Solve Realistic and Time-Consuming Tasks?
Language agents, built on top of language models (LMs), are systems that can interact with complex environments, such as the open web. In this work, we examine whether such agents can perform realistic and time-consuming tasks on the web, e.g., monitoring real-estate markets or locating relevant nearby businesses. We introduce AssistantBench, a challenging new benchmark consisting of 214 realistic tasks that can be automatically evaluated, covering different scenarios and domains. We find that AssistantBench exposes the limitations of current systems, including language models and retrieval-augmented language models, as no model reaches an accuracy of more than 26 points. While closed-book LMs perform well in terms of accuracy, they exhibit low precision and tend to hallucinate facts. State-of-the-art web agents reach a score of near zero. Additionally, we introduce SeePlanAct (SPA), a new web agent that significantly outperforms previous agents, and an ensemble of SPA and closed-book models reaches the best overall performance. Moreover, we analyze failures of current systems and highlight that open web navigation remains a major challenge.
♻ ☆ On the Role of Context in Reading Time Prediction EMNLP 2024
We present a new perspective on how readers integrate context during real-time language comprehension. Our proposals build on surprisal theory, which posits that the processing effort of a linguistic unit (e.g., a word) is an affine function of its in-context information content. We first observe that surprisal is only one out of many potential ways that a contextual predictor can be derived from a language model. Another one is the pointwise mutual information (PMI) between a unit and its context, which turns out to yield the same predictive power as surprisal when controlling for unigram frequency. Moreover, both PMI and surprisal are correlated with frequency. This means that neither PMI nor surprisal contains information about context alone. In response to this, we propose a technique where we project surprisal onto the orthogonal complement of frequency, yielding a new contextual predictor that is uncorrelated with frequency. Our experiments show that the proportion of variance in reading times explained by context is a lot smaller when context is represented by the orthogonalized predictor. From an interpretability standpoint, this indicates that previous studies may have overstated the role that context has in predicting reading times.
comment: EMNLP 2024
♻ ☆ Aligning Translation-Specific Understanding to General Understanding in Large Language Models EMNLP2024
Large Language models (LLMs) have exhibited remarkable abilities in understanding complex texts, offering a promising path towards human-like translation performance. However, this study reveals the misalignment between the translation-specific understanding and the general understanding inside LLMs. This understanding misalignment leads to LLMs mistakenly or literally translating some complicated concepts that they accurately comprehend in the general scenarios (e.g., QA). To align the translation-specific understanding to the general one, we propose a novel translation process, DUAT (Difficult words Understanding Aligned Translation), explicitly incorporating the general understanding on the complicated content incurring inconsistent understanding to guide the translation. Specifically, DUAT performs cross-lingual interpretation for the difficult-to-translate words and enhances the translation with the generated interpretations. Furthermore, we reframe the external tools to improve DUAT in detecting difficult words and generating helpful interpretations. We conduct experiments on the self-constructed benchmark Challenge-WMT, consisting of samples that are prone to mistranslation. Human evaluation results on high-resource and low-resource language pairs indicate that DUAT significantly facilitates the understanding alignment, which improves the translation quality (up to +3.85 COMET) and reduces the literality of the translation by -25% to -51%.
comment: EMNLP2024 (Main)
♻ ☆ Du-IN: Discrete units-guided mask modeling for decoding speech from Intracranial Neural signals
Invasive brain-computer interfaces with Electrocorticography (ECoG) have shown promise for high-performance speech decoding in medical applications, but less damaging methods like intracranial stereo-electroencephalography (sEEG) remain underexplored. With rapid advances in representation learning, leveraging abundant recordings to enhance speech decoding is increasingly attractive. However, popular methods often pre-train temporal models based on brain-level tokens, overlooking that brain activities in different regions are highly desynchronized during tasks. Alternatively, they pre-train spatial-temporal models based on channel-level tokens but fail to evaluate them on challenging tasks like speech decoding, which requires intricate processing in specific language-related areas. To address this issue, we collected a well-annotated Chinese word-reading sEEG dataset targeting language-related brain networks from 12 subjects. Using this benchmark, we developed the Du-IN model, which extracts contextual embeddings based on region-level tokens through discrete codex-guided mask modeling. Our model achieves state-of-the-art performance on the 61-word classification task, surpassing all baselines. Model comparisons and ablation studies reveal that our design choices, including (i) temporal modeling based on region-level tokens by utilizing 1D depthwise convolution to fuse channels in the lateral sensorimotor cortex (vSMC) and superior temporal gyrus (STG) and (ii) self-supervision through discrete codex-guided mask modeling, significantly contribute to this performance. Overall, our approach -- inspired by neuroscience findings and capitalizing on region-level representations from specific brain regions -- is suitable for invasive brain modeling and represents a promising neuro-inspired AI approach in brain-computer interfaces.
♻ ☆ A Bi-consolidating Model for Joint Relational Triple Extraction
Current methods to extract relational triples directly make a prediction based on a possible entity pair in a raw sentence without depending on entity recognition. The task suffers from a serious semantic overlapping problem, in which several relation triples may share one or two entities in a sentence. In this paper, based on a two-dimensional sentence representation, a bi-consolidating model is proposed to address this problem by simultaneously reinforcing the local and global semantic features relevant to a relation triple. This model consists of a local consolidation component and a global consolidation component. The first component uses a pixel difference convolution to enhance semantic information of a possible triple representation from adjacent regions and mitigate noise in neighbouring neighbours. The second component strengthens the triple representation based a channel attention and a spatial attention, which has the advantage to learn remote semantic dependencies in a sentence. They are helpful to improve the performance of both entity identification and relation type classification in relation triple extraction. After evaluated on several publish datasets, the bi-consolidating model achieves competitive performance. Analytical experiments demonstrate the effectiveness of our model for relational triple extraction and give motivation for other natural language processing tasks.
♻ ☆ Latent Skill Discovery for Chain-of-Thought Reasoning
Chain-of-thought (CoT) prompting is a popular in-context learning (ICL) approach for large language models (LLMs), especially when tackling complex reasoning tasks. Traditional ICL approaches construct prompts using examples that contain questions similar to the input question. However, CoT prompting, which includes crucial intermediate reasoning steps (rationales) within its examples, necessitates selecting examples based on these rationales rather than the questions themselves. Existing methods require human experts or pre-trained LLMs to describe the skill, a high-level abstraction of rationales, to guide the selection. These methods, however, are often costly and difficult to scale. Instead, this paper introduces a new approach named Latent Reasoning Skills (LaRS) that employs unsupervised learning to create a latent space representation of rationales, with a latent variable called a reasoning skill. Concurrently, LaRS learns a reasoning policy to determine the required reasoning skill for a given question. Then the ICL examples are selected by aligning the reasoning skills between past examples and the question. This approach is theoretically grounded and compute-efficient, eliminating the need for auxiliary LLM inference or manual prompt design. Empirical results demonstrate that LaRS consistently outperforms SOTA skill-based selection methods, processing example banks four times faster, reducing LLM inferences during the selection stage by half, and showing greater robustness to sub-optimal example banks.
Computer Vision and Pattern Recognition 82
☆ PlaneSAM: Multimodal Plane Instance Segmentation Using the Segment Anything Model
Plane instance segmentation from RGB-D data is a crucial research topic for many downstream tasks. However, most existing deep-learning-based methods utilize only information within the RGB bands, neglecting the important role of the depth band in plane instance segmentation. Based on EfficientSAM, a fast version of SAM, we propose a plane instance segmentation network called PlaneSAM, which can fully integrate the information of the RGB bands (spectral bands) and the D band (geometric band), thereby improving the effectiveness of plane instance segmentation in a multimodal manner. Specifically, we use a dual-complexity backbone, with primarily the simpler branch learning D-band features and primarily the more complex branch learning RGB-band features. Consequently, the backbone can effectively learn D-band feature representations even when D-band training data is limited in scale, retain the powerful RGB-band feature representations of EfficientSAM, and allow the original backbone branch to be fine-tuned for the current task. To enhance the adaptability of our PlaneSAM to the RGB-D domain, we pretrain our dual-complexity backbone using the segment anything task on large-scale RGB-D data through a self-supervised pretraining strategy based on imperfect pseudo-labels. To support the segmentation of large planes, we optimize the loss function combination ratio of EfficientSAM. In addition, Faster R-CNN is used as a plane detector, and its predicted bounding boxes are fed into our dual-complexity network as prompts, thereby enabling fully automatic plane instance segmentation. Experimental results show that the proposed PlaneSAM sets a new SOTA performance on the ScanNet dataset, and outperforms previous SOTA approaches in zero-shot transfer on the 2D-3D-S, Matterport3D, and ICL-NUIM RGB-D datasets, while only incurring a 10% increase in computational overhead compared to EfficientSAM.
comment: submitted to Information Fusion
☆ Large Body Language Models
As virtual agents become increasingly prevalent in human-computer interaction, generating realistic and contextually appropriate gestures in real-time remains a significant challenge. While neural rendering techniques have made substantial progress with static scripts, their applicability to human-computer interactions remains limited. To address this, we introduce Large Body Language Models (LBLMs) and present LBLM-AVA, a novel LBLM architecture that combines a Transformer-XL large language model with a parallelized diffusion model to generate human-like gestures from multimodal inputs (text, audio, and video). LBLM-AVA incorporates several key components enhancing its gesture generation capabilities, such as multimodal-to-pose embeddings, enhanced sequence-to-sequence mapping with redefined attention mechanisms, a temporal smoothing module for gesture sequence coherence, and an attention-based refinement module for enhanced realism. The model is trained on our large-scale proprietary open-source dataset Allo-AVA. LBLM-AVA achieves state-of-the-art performance in generating lifelike and contextually appropriate gestures with a 30% reduction in Fr\'echet Gesture Distance (FGD), and a 25% improvement in Fr\'echet Inception Distance compared to existing approaches.
☆ Gradient-Free Supervised Learning using Spike-Timing-Dependent Plasticity for Image Recognition
An approach to supervised learning in spiking neural networks is presented using a gradient-free method combined with spike-timing-dependent plasticity for image recognition. The proposed network architecture is scalable to multiple layers, enabling the development of more complex and deeper SNN models. The effectiveness of this method is demonstrated by its application to the MNIST dataset, showing good learning accuracy. The proposed method provides a robust and efficient alternative to the backpropagation-based method in supervised learning.
☆ Efficient Neural Network Training via Subset Pretraining
In training neural networks, it is common practice to use partial gradients computed over batches, mostly very small subsets of the training set. This approach is motivated by the argument that such a partial gradient is close to the true one, with precision growing only with the square root of the batch size. A theoretical justification is with the help of stochastic approximation theory. However, the conditions for the validity of this theory are not satisfied in the usual learning rate schedules. Batch processing is also difficult to combine with efficient second-order optimization methods. This proposal is based on another hypothesis: the loss minimum of the training set can be expected to be well-approximated by the minima of its subsets. Such subset minima can be computed in a fraction of the time necessary for optimizing over the whole training set. This hypothesis has been tested with the help of the MNIST, CIFAR-10, and CIFAR-100 image classification benchmarks, optionally extended by training data augmentation. The experiments have confirmed that results equivalent to conventional training can be reached. In summary, even small subsets are representative if the overdetermination ratio for the given model parameter set sufficiently exceeds unity. The computing expense can be reduced to a tenth or less.
comment: To appear in KDIR 2024
☆ TIPS: Text-Image Pretraining with Spatial Awareness
While image-text representation learning has become very popular in recent years, existing models tend to lack spatial awareness and have limited direct applicability for dense understanding tasks. For this reason, self-supervised image-only pretraining is still the go-to method for many dense vision applications (e.g. depth estimation, semantic segmentation), despite the lack of explicit supervisory signals. In this paper, we close this gap between image-text and self-supervised learning, by proposing a novel general-purpose image-text model, which can be effectively used off-the-shelf for dense and global vision tasks. Our method, which we refer to as Text-Image Pretraining with Spatial awareness (TIPS), leverages two simple and effective insights. First, on textual supervision: we reveal that replacing noisy web image captions by synthetically generated textual descriptions boosts dense understanding performance significantly, due to a much richer signal for learning spatially aware representations. We propose an adapted training method that combines noisy and synthetic captions, resulting in improvements across both dense and global understanding tasks. Second, on the learning technique: we propose to combine contrastive image-text learning with self-supervised masked image modeling, to encourage spatial coherence, unlocking substantial enhancements for downstream applications. Building on these two ideas, we scale our model using the transformer architecture, trained on a curated set of public images. Our experiments are conducted on 8 tasks involving 16 datasets in total, demonstrating strong off-the-shelf performance on both dense and global understanding, for several image-only and image-text tasks.
☆ Allo-AVA: A Large-Scale Multimodal Conversational AI Dataset for Allocentric Avatar Gesture Animation
The scarcity of high-quality, multimodal training data severely hinders the creation of lifelike avatar animations for conversational AI in virtual environments. Existing datasets often lack the intricate synchronization between speech, facial expressions, and body movements that characterize natural human communication. To address this critical gap, we introduce Allo-AVA, a large-scale dataset specifically designed for text and audio-driven avatar gesture animation in an allocentric (third person point-of-view) context. Allo-AVA consists of $\sim$1,250 hours of diverse video content, complete with audio, transcripts, and extracted keypoints. Allo-AVA uniquely maps these keypoints to precise timestamps, enabling accurate replication of human movements (body and facial gestures) in synchronization with speech. This comprehensive resource enables the development and evaluation of more natural, context-aware avatar animation models, potentially transforming applications ranging from virtual reality to digital assistants.
☆ SINGAPO: Single Image Controlled Generation of Articulated Parts in Object
We address the challenge of creating 3D assets for household articulated objects from a single image. Prior work on articulated object creation either requires multi-view multi-state input, or only allows coarse control over the generation process. These limitations hinder the scalability and practicality for articulated object modeling. In this work, we propose a method to generate articulated objects from a single image. Observing the object in resting state from an arbitrary view, our method generates an articulated object that is visually consistent with the input image. To capture the ambiguity in part shape and motion posed by a single view of the object, we design a diffusion model that learns the plausible variations of objects in terms of geometry and kinematics. To tackle the complexity of generating structured data with attributes in multiple domains, we design a pipeline that produces articulated objects from high-level structure to geometric details in a coarse-to-fine manner, where we use a part connectivity graph and part abstraction as proxies. Our experiments show that our method outperforms the state-of-the-art in articulated object creation by a large margin in terms of the generated object realism, resemblance to the input image, and reconstruction quality.
comment: Project page: https://3dlg-hcvc.github.io/singapo
☆ GenGMM: Generalized Gaussian-Mixture-based Domain Adaptation Model for Semantic Segmentation
Domain adaptive semantic segmentation is the task of generating precise and dense predictions for an unlabeled target domain using a model trained on a labeled source domain. While significant efforts have been devoted to improving unsupervised domain adaptation for this task, it is crucial to note that many models rely on a strong assumption that the source data is entirely and accurately labeled, while the target data is unlabeled. In real-world scenarios, however, we often encounter partially or noisy labeled data in source and target domains, referred to as Generalized Domain Adaptation (GDA). In such cases, we suggest leveraging weak or unlabeled data from both domains to narrow the gap between them, resulting in effective adaptation. We introduce the Generalized Gaussian-mixture-based (GenGMM) domain adaptation model, which harnesses the underlying data distribution in both domains to refine noisy weak and pseudo labels. The experiments demonstrate the effectiveness of our approach.
☆ AlignVSR: Audio-Visual Cross-Modal Alignment for Visual Speech Recognition
Visual Speech Recognition (VSR) aims to recognize corresponding text by analyzing visual information from lip movements. Due to the high variability and weak information of lip movements, VSR tasks require effectively utilizing any information from any source and at any level. In this paper, we propose a VSR method based on audio-visual cross-modal alignment, named AlignVSR. The method leverages the audio modality as an auxiliary information source and utilizes the global and local correspondence between the audio and visual modalities to improve visual-to-text inference. Specifically, the method first captures global alignment between video and audio through a cross-modal attention mechanism from video frames to a bank of audio units. Then, based on the temporal correspondence between audio and video, a frame-level local alignment loss is introduced to refine the global alignment, improving the utility of the audio information. Experimental results on the LRS2 and CNVSRC.Single datasets consistently show that AlignVSR outperforms several mainstream VSR methods, demonstrating its superior and robust performance.
☆ HaHeAE: Learning Generalisable Joint Representations of Human Hand and Head Movements in Extended Reality
Human hand and head movements are the most pervasive input modalities in extended reality (XR) and are significant for a wide range of applications. However, prior works on hand and head modelling in XR only explored a single modality or focused on specific applications. We present HaHeAE - a novel self-supervised method for learning generalisable joint representations of hand and head movements in XR. At the core of our method is an autoencoder (AE) that uses a graph convolutional network-based semantic encoder and a diffusion-based stochastic encoder to learn the joint semantic and stochastic representations of hand-head movements. It also features a diffusion-based decoder to reconstruct the original signals. Through extensive evaluations on three public XR datasets, we show that our method 1) significantly outperforms commonly used self-supervised methods by up to 74.0% in terms of reconstruction quality and is generalisable across users, activities, and XR environments, 2) enables new applications, including interpretable hand-head cluster identification and variable hand-head movement generation, and 3) can serve as an effective feature extractor for downstream tasks. Together, these results demonstrate the effectiveness of our method and underline the potential of self-supervised methods for jointly modelling hand-head behaviours in extended reality.
☆ AttentionPainter: An Efficient and Adaptive Stroke Predictor for Scene Painting
Stroke-based Rendering (SBR) aims to decompose an input image into a sequence of parameterized strokes, which can be rendered into a painting that resembles the input image. Recently, Neural Painting methods that utilize deep learning and reinforcement learning models to predict the stroke sequences have been developed, but suffer from longer inference time or unstable training. To address these issues, we propose AttentionPainter, an efficient and adaptive model for single-step neural painting. First, we propose a novel scalable stroke predictor, which predicts a large number of stroke parameters within a single forward process, instead of the iterative prediction of previous Reinforcement Learning or auto-regressive methods, which makes AttentionPainter faster than previous neural painting methods. To further increase the training efficiency, we propose a Fast Stroke Stacking algorithm, which brings 13 times acceleration for training. Moreover, we propose Stroke-density Loss, which encourages the model to use small strokes for detailed information, to help improve the reconstruction quality. Finally, we propose a new stroke diffusion model for both conditional and unconditional stroke-based generation, which denoises in the stroke parameter space and facilitates stroke-based inpainting and editing applications helpful for human artists design. Extensive experiments show that AttentionPainter outperforms the state-of-the-art neural painting methods.
☆ Joker: Conditional 3D Head Synthesis with Extreme Facial Expressions
We introduce Joker, a new method for the conditional synthesis of 3D human heads with extreme expressions. Given a single reference image of a person, we synthesize a volumetric human head with the reference identity and a new expression. We offer control over the expression via a 3D morphable model (3DMM) and textual inputs. This multi-modal conditioning signal is essential since 3DMMs alone fail to define subtle emotional changes and extreme expressions, including those involving the mouth cavity and tongue articulation. Our method is built upon a 2D diffusion-based prior that generalizes well to out-of-domain samples, such as sculptures, heavy makeup, and paintings while achieving high levels of expressiveness. To improve view consistency, we propose a new 3D distillation technique that converts predictions of our 2D prior into a neural radiance field (NeRF). Both the 2D prior and our distillation technique produce state-of-the-art results, which are confirmed by our extensive evaluations. Also, to the best of our knowledge, our method is the first to achieve view-consistent extreme tongue articulation.
comment: Project Page: https://malteprinzler.github.io/projects/joker/
☆ FrugalNeRF: Fast Convergence for Few-shot Novel View Synthesis without Learned Priors
Neural Radiance Fields (NeRF) face significant challenges in few-shot scenarios, primarily due to overfitting and long training times for high-fidelity rendering. Existing methods, such as FreeNeRF and SparseNeRF, use frequency regularization or pre-trained priors but struggle with complex scheduling and bias. We introduce FrugalNeRF, a novel few-shot NeRF framework that leverages weight-sharing voxels across multiple scales to efficiently represent scene details. Our key contribution is a cross-scale geometric adaptation scheme that selects pseudo ground truth depth based on reprojection errors across scales. This guides training without relying on externally learned priors, enabling full utilization of the training data. It can also integrate pre-trained priors, enhancing quality without slowing convergence. Experiments on LLFF, DTU, and RealEstate-10K show that FrugalNeRF outperforms other few-shot NeRF methods while significantly reducing training time, making it a practical solution for efficient and accurate 3D scene reconstruction.
comment: Project page: https://linjohnss.github.io/frugalnerf/
☆ MvDrag3D: Drag-based Creative 3D Editing via Multi-view Generation-Reconstruction Priors
Drag-based editing has become popular in 2D content creation, driven by the capabilities of image generative models. However, extending this technique to 3D remains a challenge. Existing 3D drag-based editing methods, whether employing explicit spatial transformations or relying on implicit latent optimization within limited-capacity 3D generative models, fall short in handling significant topology changes or generating new textures across diverse object categories. To overcome these limitations, we introduce MVDrag3D, a novel framework for more flexible and creative drag-based 3D editing that leverages multi-view generation and reconstruction priors. At the core of our approach is the usage of a multi-view diffusion model as a strong generative prior to perform consistent drag editing over multiple rendered views, which is followed by a reconstruction model that reconstructs 3D Gaussians of the edited object. While the initial 3D Gaussians may suffer from misalignment between different views, we address this via view-specific deformation networks that adjust the position of Gaussians to be well aligned. In addition, we propose a multi-view score function that distills generative priors from multiple views to further enhance the view consistency and visual quality. Extensive experiments demonstrate that MVDrag3D provides a precise, generative, and flexible solution for 3D drag-based editing, supporting more versatile editing effects across various object categories and 3D representations.
comment: 16 pages, 10 figures, conference
☆ SAM2Long: Enhancing SAM 2 for Long Video Segmentation with a Training-Free Memory Tree
The Segment Anything Model 2 (SAM 2) has emerged as a powerful foundation model for object segmentation in both images and videos, paving the way for various downstream video applications. The crucial design of SAM 2 for video segmentation is its memory module, which prompts object-aware memories from previous frames for current frame prediction. However, its greedy-selection memory design suffers from the "error accumulation" problem, where an errored or missed mask will cascade and influence the segmentation of the subsequent frames, which limits the performance of SAM 2 toward complex long-term videos. To this end, we introduce SAM2Long, an improved training-free video object segmentation strategy, which considers the segmentation uncertainty within each frame and chooses the video-level optimal results from multiple segmentation pathways in a constrained tree search manner. In practice, we maintain a fixed number of segmentation pathways throughout the video. For each frame, multiple masks are proposed based on the existing pathways, creating various candidate branches. We then select the same fixed number of branches with higher cumulative scores as the new pathways for the next frame. After processing the final frame, the pathway with the highest cumulative score is chosen as the final segmentation result. Benefiting from its heuristic search design, SAM2Long is robust toward occlusions and object reappearances, and can effectively segment and track objects for complex long-term videos. Notably, SAM2Long achieves an average improvement of 3.0 points across all 24 head-to-head comparisons, with gains of up to 5.3 points in J&F on long-term video object segmentation benchmarks such as SA-V and LVOS. The code is released at https://github.com/Mark12Ding/SAM2Long.
comment: Project page: https://mark12ding.github.io/project/SAM2Long/
☆ xGen-MM-Vid (BLIP-3-Video): You Only Need 32 Tokens to Represent a Video Even in VLMs
We present xGen-MM-Vid (BLIP-3-Video): a multimodal language model for videos, particularly designed to efficiently capture temporal information over multiple frames. BLIP-3-Video takes advantage of the 'temporal encoder' in addition to the conventional visual tokenizer, which maps a sequence of tokens over multiple frames into a compact set of visual tokens. This enables BLIP3-Video to use much fewer visual tokens than its competing models (e.g., 32 vs. 4608 tokens). We explore different types of temporal encoders, including learnable spatio-temporal pooling as well as sequential models like Token Turing Machines. We experimentally confirm that BLIP-3-Video obtains video question-answering accuracies comparable to much larger state-of-the-art models (e.g., 34B), while being much smaller (i.e., 4B) and more efficient by using fewer visual tokens. The project website is at https://www.salesforceairesearch.com/opensource/xGen-MM-Vid/index.html
☆ 3DGS-Enhancer: Enhancing Unbounded 3D Gaussian Splatting with View-consistent 2D Diffusion Priors NeurIPS 2024
Novel-view synthesis aims to generate novel views of a scene from multiple input images or videos, and recent advancements like 3D Gaussian splatting (3DGS) have achieved notable success in producing photorealistic renderings with efficient pipelines. However, generating high-quality novel views under challenging settings, such as sparse input views, remains difficult due to insufficient information in under-sampled areas, often resulting in noticeable artifacts. This paper presents 3DGS-Enhancer, a novel pipeline for enhancing the representation quality of 3DGS representations. We leverage 2D video diffusion priors to address the challenging 3D view consistency problem, reformulating it as achieving temporal consistency within a video generation process. 3DGS-Enhancer restores view-consistent latent features of rendered novel views and integrates them with the input views through a spatial-temporal decoder. The enhanced views are then used to fine-tune the initial 3DGS model, significantly improving its rendering performance. Extensive experiments on large-scale datasets of unbounded scenes demonstrate that 3DGS-Enhancer yields superior reconstruction performance and high-fidelity rendering results compared to state-of-the-art methods. The project webpage is https://xiliu8006.github.io/3DGS-Enhancer-project .
comment: Accepted by NeurIPS 2024 Spotlight
☆ Agent-to-Sim: Learning Interactive Behavior Models from Casual Longitudinal Videos
We present Agent-to-Sim (ATS), a framework for learning interactive behavior models of 3D agents from casual longitudinal video collections. Different from prior works that rely on marker-based tracking and multiview cameras, ATS learns natural behaviors of animal and human agents non-invasively through video observations recorded over a long time-span (e.g., a month) in a single environment. Modeling 3D behavior of an agent requires persistent 3D tracking (e.g., knowing which point corresponds to which) over a long time period. To obtain such data, we develop a coarse-to-fine registration method that tracks the agent and the camera over time through a canonical 3D space, resulting in a complete and persistent spacetime 4D representation. We then train a generative model of agent behaviors using paired data of perception and motion of an agent queried from the 4D reconstruction. ATS enables real-to-sim transfer from video recordings of an agent to an interactive behavior simulator. We demonstrate results on pets (e.g., cat, dog, bunny) and human given monocular RGBD videos captured by a smartphone.
comment: Project page: https://gengshan-y.github.io/agent2sim-www/
☆ Elucidating the design space of language models for image generation
The success of autoregressive (AR) language models in text generation has inspired the computer vision community to adopt Large Language Models (LLMs) for image generation. However, considering the essential differences between text and image modalities, the design space of language models for image generation remains underexplored. We observe that image tokens exhibit greater randomness compared to text tokens, which presents challenges when training with token prediction. Nevertheless, AR models demonstrate their potential by effectively learning patterns even from a seemingly suboptimal optimization problem. Our analysis also reveals that while all models successfully grasp the importance of local information in image generation, smaller models struggle to capture the global context. In contrast, larger models showcase improved capabilities in this area, helping to explain the performance gains achieved when scaling up model size. We further elucidate the design space of language models for vision generation, including tokenizer choice, model choice, model scalability, vocabulary design, and sampling strategy through extensive comparative experiments. Our work is the first to analyze the optimization behavior of language models in vision generation, and we believe it can inspire more effective designs when applying LMs to other domains. Finally, our elucidated language model for image generation, termed as ELM, achieves state-of-the-art performance on the ImageNet 256*256 benchmark. The code is available at https://github.com/Pepperlll/LMforImageGeneration.git.
comment: Project page: https://pepper-lll.github.io/LMforImageGeneration/
☆ Revisiting Deep Feature Reconstruction for Logical and Structural Industrial Anomaly Detection
Industrial anomaly detection is crucial for quality control and predictive maintenance, but it presents challenges due to limited training data, diverse anomaly types, and external factors that alter object appearances. Existing methods commonly detect structural anomalies, such as dents and scratches, by leveraging multi-scale features from image patches extracted through deep pre-trained networks. However, significant memory and computational demands often limit their practical application. Additionally, detecting logical anomalies-such as images with missing or excess elements-requires an understanding of spatial relationships that traditional patch-based methods fail to capture. In this work, we address these limitations by focusing on Deep Feature Reconstruction (DFR), a memory- and compute-efficient approach for detecting structural anomalies. We further enhance DFR into a unified framework, called ULSAD, which is capable of detecting both structural and logical anomalies. Specifically, we refine the DFR training objective to improve performance in structural anomaly detection, while introducing an attention-based loss mechanism using a global autoencoder-like network to handle logical anomaly detection. Our empirical evaluation across five benchmark datasets demonstrates the performance of ULSAD in detecting and localizing both structural and logical anomalies, outperforming eight state-of-the-art methods. An extensive ablation study further highlights the contribution of each component to the overall performance improvement. Our code is available at https://github.com/sukanyapatra1997/ULSAD-2024.git
comment: Accepted in Transactions on Machine Learning Research (TMLR). Link to OpenReview: https://openreview.net/forum?id=kdTC4ktHPD
☆ MoRE: Multi-Modal Contrastive Pre-training with Transformers on X-Rays, ECGs, and Diagnostic Report
In this paper, we introduce a novel Multi-Modal Contrastive Pre-training Framework that synergistically combines X-rays, electrocardiograms (ECGs), and radiology/cardiology reports. Our approach leverages transformers to encode these diverse modalities into a unified representation space, aiming to enhance diagnostic accuracy and facilitate comprehensive patient assessments. We utilize LoRA-Peft to significantly reduce trainable parameters in the LLM and incorporate recent linear attention dropping strategy in the Vision Transformer(ViT) for smoother attention. Furthermore, we provide novel multimodal attention explanations and retrieval for our model. To the best of our knowledge, we are the first to propose an integrated model that combines X-ray, ECG, and Radiology/Cardiology Report with this approach. By utilizing contrastive loss, MoRE effectively aligns modality-specific features into a coherent embedding, which supports various downstream tasks such as zero-shot classification and multimodal retrieval. Employing our proposed methodology, we achieve state-of-the-art (SOTA) on the Mimic-IV, CheXpert, Edema Severity, and PtbXl downstream datasets, surpassing existing multimodal approaches. Our proposed framework shows significant improvements in capturing intricate inter-modal relationships and its robustness in medical diagnosis that establishes a framework for future research in multimodal learning in the healthcare sector.
comment: 10 pages, 5 figures, 9 tables. Supplementary detail in Appendix. Code made available in Github for reproducibility
☆ Deep Radiomics Detection of Clinically Significant Prostate Cancer on Multicenter MRI: Initial Comparison to PI-RADS Assessment
Objective: To develop and evaluate a deep radiomics model for clinically significant prostate cancer (csPCa, grade group >= 2) detection and compare its performance to Prostate Imaging Reporting and Data System (PI-RADS) assessment in a multicenter cohort. Materials and Methods: This retrospective study analyzed biparametric (T2W and DW) prostate MRI sequences of 615 patients (mean age, 63.1 +/- 7 years) from four datasets acquired between 2010 and 2020: PROSTATEx challenge, Prostate158 challenge, PCaMAP trial, and an in-house (NTNU/St. Olavs Hospital) dataset. With expert annotations as ground truth, a deep radiomics model was trained, including nnU-Net segmentation of the prostate gland, voxel-wise radiomic feature extraction, extreme gradient boost classification, and post-processing of tumor probability maps into csPCa detection maps. Training involved 5-fold cross-validation using the PROSTATEx (n=199), Prostate158 (n=138), and PCaMAP (n=78) datasets, and testing on the in-house (n=200) dataset. Patient- and lesion-level performance were compared to PI-RADS using area under ROC curve (AUROC [95% CI]), sensitivity, and specificity analysis. Results: On the test data, the radiologist achieved a patient-level AUROC of 0.94 [0.91-0.98] with 94% (75/80) sensitivity and 77% (92/120) specificity at PI-RADS >= 3. The deep radiomics model at a tumor probability cut-off >= 0.76 achieved 0.91 [0.86-0.95] AUROC with 90% (72/80) sensitivity and 73% (87/120) specificity, not significantly different (p = 0.068) from PI-RADS. On the lesion level, PI-RADS cut-off >= 3 had 84% (91/108) sensitivity at 0.2 (40/200) false positives per patient, while deep radiomics attained 68% (73/108) sensitivity at the same false positive rate. Conclusion: Deep radiomics machine learning model achieved comparable performance to PI-RADS assessment in csPCa detection at the patient-level but not at the lesion-level.
comment: 20 pages, 4 figures, 4 tables
☆ LLaVA-KD: A Framework of Distilling Multimodal Large Language Models
The success of Large Language Models (LLM) has led researchers to explore Multimodal Large Language Models (MLLM) for unified visual and linguistic understanding. However, the increasing model size and computational complexity of MLLM limit their use in resource-constrained environments. Small-scale MLLM (s-MLLM) aims to retain the capabilities of the large-scale model (l-MLLM) while reducing computational demands, but resulting in a significant decline in performance. To address the aforementioned issues, we propose a novel LLaVA-KD framework to transfer knowledge from l-MLLM to s-MLLM. Specifically, we introduce Multimodal Distillation (MDist) to minimize the divergence between the visual-textual output distributions of l-MLLM and s-MLLM, and Relation Distillation (RDist) to transfer l-MLLM's ability to model correlations between visual features. Additionally, we propose a three-stage training scheme to fully exploit the potential of s-MLLM: 1) Distilled Pre-Training to align visual-textual representations, 2) Supervised Fine-Tuning to equip the model with multimodal understanding, and 3) Distilled Fine-Tuning to further transfer l-MLLM capabilities. Our approach significantly improves performance without altering the small model's architecture. Extensive experiments and ablation studies validate the effectiveness of each proposed component. Code will be available at https://github.com/caiyuxuan1120/LLaVA-KD.
comment: Under review
☆ Managing Bandwidth: The Key to Cloud-Assisted Autonomous Driving
Prevailing wisdom asserts that one cannot rely on the cloud for critical real-time control systems like self-driving cars. We argue that we can, and must. Following the trends of increasing model sizes, improvements in hardware, and evolving mobile networks, we identify an opportunity to offload parts of time-sensitive and latency-critical compute to the cloud. Doing so requires carefully allocating bandwidth to meet strict latency SLOs, while maximizing benefit to the car.
comment: 6 pages
☆ Improve Vision Language Model Chain-of-thought Reasoning
Chain-of-thought (CoT) reasoning in vision language models (VLMs) is crucial for improving interpretability and trustworthiness. However, current training recipes lack robust CoT reasoning data, relying on datasets dominated by short annotations with minimal rationales. In this work, we show that training VLM on short answers does not generalize well to reasoning tasks that require more detailed responses. To address this, we propose a two-fold approach. First, we distill rationales from GPT-4o model to enrich the training data and fine-tune VLMs, boosting their CoT performance. Second, we apply reinforcement learning to further calibrate reasoning quality. Specifically, we construct positive (correct) and negative (incorrect) pairs of model-generated reasoning chains, by comparing their predictions with annotated short answers. Using this pairwise data, we apply the Direct Preference Optimization algorithm to refine the model's reasoning abilities. Our experiments demonstrate significant improvements in CoT reasoning on benchmark datasets and better generalization to direct answer prediction as well. This work emphasizes the importance of incorporating detailed rationales in training and leveraging reinforcement learning to strengthen the reasoning capabilities of VLMs.
comment: 10 pages + appendix
☆ Training Better Deep Learning Models Using Human Saliency
This work explores how human judgement about salient regions of an image can be introduced into deep convolutional neural network (DCNN) training. Traditionally, training of DCNNs is purely data-driven. This often results in learning features of the data that are only coincidentally correlated with class labels. Human saliency can guide network training using our proposed new component of the loss function that ConveYs Brain Oversight to Raise Generalization (CYBORG) and penalizes the model for using non-salient regions. This mechanism produces DCNNs achieving higher accuracy and generalization compared to using the same training data without human salience. Experimental results demonstrate that CYBORG applies across multiple network architectures and problem domains (detection of synthetic faces, iris presentation attacks and anomalies in chest X-rays), while requiring significantly less data than training without human saliency guidance. Visualizations show that CYBORG-trained models' saliency is more consistent across independent training runs than traditionally-trained models, and also in better agreement with humans. To lower the cost of collecting human annotations, we also explore using deep learning to provide automated annotations. CYBORG training of CNNs addresses important issues such as reducing the appetite for large training sets, increasing interpretability, and reducing fragility by generalizing better to new types of data.
☆ A Framework for Evaluating Predictive Models Using Synthetic Image Covariates and Longitudinal Data
We present a novel framework for synthesizing patient data with complex covariates (e.g., eye scans) paired with longitudinal observations (e.g., visual acuity over time), addressing privacy concerns in healthcare research. Our approach introduces controlled association in latent spaces generating each data modality, enabling the creation of complex covariate-longitudinal observation pairs. This framework facilitates the development of predictive models and provides openly available benchmarking datasets for healthcare research. We demonstrate our framework using optical coherence tomography (OCT) scans, though it is applicable across domains. Using 109,309 2D OCT scan slices, we trained an image generative model combining a variational autoencoder and a diffusion model. Longitudinal observations were simulated using a nonlinear mixed effect (NLME) model from a low-dimensional space of random effects. We generated 1.1M OCT scan slices paired with five sets of longitudinal observations at controlled association levels (100%, 50%, 10%, 5.26%, and 2% of between-subject variability). To assess the framework, we modeled synthetic longitudinal observations with another NLME model, computed empirical Bayes estimates of random effects, and trained a ResNet to predict these estimates from synthetic OCT scans. We then incorporated ResNet predictions into the NLME model for patient-individualized predictions. Prediction accuracy on withheld data declined as intended with reduced association between images and longitudinal measurements. Notably, in all but the 2% case, we achieved within 50% of the theoretical best possible prediction on withheld data, demonstrating our ability to detect even weak signals. This confirms the effectiveness of our framework in generating synthetic data with controlled levels of association, providing a valuable tool for healthcare research.
☆ Beyond Filtering: Adaptive Image-Text Quality Enhancement for MLLM Pretraining
Multimodal large language models (MLLMs) have made significant strides by integrating visual and textual modalities. A critical factor in training MLLMs is the quality of image-text pairs within multimodal pretraining datasets. However, $\textit {de facto}$ filter-based data quality enhancement paradigms often discard a substantial portion of high-quality image data due to inadequate semantic alignment between images and texts, leading to inefficiencies in data utilization and scalability. In this paper, we propose the Adaptive Image-Text Quality Enhancer (AITQE), a model that dynamically assesses and enhances the quality of image-text pairs. AITQE employs a text rewriting mechanism for low-quality pairs and incorporates a negative sample learning strategy to improve evaluative capabilities by integrating deliberately selected low-quality samples during training. Unlike prior approaches that significantly alter text distributions, our method minimally adjusts text to preserve data volume while enhancing quality. Experimental results demonstrate that AITQE surpasses existing methods on various benchmark, effectively leveraging raw data and scaling efficiently with increasing data volumes. We hope our work will inspire future works. The code and model are available at: https://github.com/hanhuang22/AITQE.
☆ Griffon-G: Bridging Vision-Language and Vision-Centric Tasks via Large Multimodal Models
Large Multimodal Models (LMMs) have achieved significant breakthroughs in various vision-language and vision-centric tasks based on auto-regressive modeling. However, these models typically focus on either vision-centric tasks, such as visual grounding and region description, or vision-language tasks, like image caption and multi-scenario VQAs. None of the LMMs have yet comprehensively unified both types of tasks within a single model, as seen in Large Language Models in the natural language processing field. Furthermore, even with abundant multi-task instruction-following data, directly stacking these data for universal capabilities extension remains challenging. To address these issues, we introduce a novel multi-dimension curated and consolidated multimodal dataset, named CCMD-8M, which overcomes the data barriers of unifying vision-centric and vision-language tasks through multi-level data curation and multi-task consolidation. More importantly, we present Griffon-G, a general large multimodal model that addresses both vision-centric and vision-language tasks within a single end-to-end paradigm. Griffon-G resolves the training collapse issue encountered during the joint optimization of these tasks, achieving better training efficiency. Evaluations across multimodal benchmarks, general Visual Question Answering (VQA) tasks, scene text-centric VQA tasks, document-related VQA tasks, Referring Expression Comprehension, and object detection demonstrate that Griffon-G surpasses the advanced LMMs and achieves expert-level performance in complicated vision-centric tasks.
comment: This work has been submitted to the IEEE for possible publication. Codes and data will be later released at https://github.com/jefferyZhan/Griffon
☆ Sparkle: Mastering Basic Spatial Capabilities in Vision Language Models Elicits Generalization to Composite Spatial Reasoning
Vision language models (VLMs) have demonstrated impressive performance across a wide range of downstream tasks. However, their proficiency in spatial reasoning remains limited, despite its crucial role in tasks involving navigation and interaction with physical environments. Specifically, much of the spatial reasoning in these tasks occurs in two-dimensional (2D) environments, and our evaluation reveals that state-of-the-art VLMs frequently generate implausible and incorrect responses to composite spatial reasoning problems, including simple pathfinding tasks that humans can solve effortlessly at a glance. To address this, we explore an effective approach to enhance 2D spatial reasoning within VLMs by training the model on basic spatial capabilities. We begin by disentangling the key components of 2D spatial reasoning: direction comprehension, distance estimation, and localization. Our central hypothesis is that mastering these basic spatial capabilities can significantly enhance a model's performance on composite spatial tasks requiring advanced spatial understanding and combinatorial problem-solving. To investigate this hypothesis, we introduce Sparkle, a framework that fine-tunes VLMs on these three basic spatial capabilities by synthetic data generation and targeted supervision to form an instruction dataset for each capability. Our experiments demonstrate that VLMs fine-tuned with Sparkle achieve significant performance gains, not only in the basic tasks themselves but also in generalizing to composite and out-of-distribution spatial reasoning tasks (e.g., improving from 13.5% to 40.0% on the shortest path problem). These findings underscore the effectiveness of mastering basic spatial capabilities in enhancing composite spatial problem-solving, offering insights for improving VLMs' spatial reasoning capabilities.
☆ Metric as Transform: Exploring beyond Affine Transform for Interpretable Neural Network
Artificial Neural Networks of varying architectures are generally paired with affine transformation at the core. However, we find dot product neurons with global influence less interpretable as compared to local influence of euclidean distance (as used in Radial Basis Function Network). In this work, we explore the generalization of dot product neurons to $l^p$-norm, metrics, and beyond. We find that metrics as transform performs similarly to affine transform when used in MultiLayer Perceptron or Convolutional Neural Network. Moreover, we explore various properties of Metrics, compare it with Affine, and present multiple cases where metrics seem to provide better interpretability. We develop an interpretable local dictionary based Neural Networks and use it to understand and reject adversarial examples.
comment: 22 pages, 20 figures, 3 tables
☆ Pangea: A Fully Open Multilingual Multimodal LLM for 39 Languages
Despite recent advances in multimodal large language models (MLLMs), their development has predominantly focused on English- and western-centric datasets and tasks, leaving most of the world's languages and diverse cultural contexts underrepresented. This paper introduces Pangea, a multilingual multimodal LLM trained on PangeaIns, a diverse 6M instruction dataset spanning 39 languages. PangeaIns features: 1) high-quality English instructions, 2) carefully machine-translated instructions, and 3) culturally relevant multimodal tasks to ensure cross-cultural coverage. To rigorously assess models' capabilities, we introduce PangeaBench, a holistic evaluation suite encompassing 14 datasets covering 47 languages. Results show that Pangea significantly outperforms existing open-source models in multilingual settings and diverse cultural contexts. Ablation studies further reveal the importance of English data proportions, language popularity, and the number of multimodal training samples on overall performance. We fully open-source our data, code, and trained checkpoints, to facilitate the development of inclusive and robust multilingual MLLMs, promoting equity and accessibility across a broader linguistic and cultural spectrum.
comment: 52 pages, 27 figures
☆ Towards Combating Frequency Simplicity-biased Learning for Domain Generalization NeurIPS 2024
Domain generalization methods aim to learn transferable knowledge from source domains that can generalize well to unseen target domains. Recent studies show that neural networks frequently suffer from a simplicity-biased learning behavior which leads to over-reliance on specific frequency sets, namely as frequency shortcuts, instead of semantic information, resulting in poor generalization performance. Despite previous data augmentation techniques successfully enhancing generalization performances, they intend to apply more frequency shortcuts, thereby causing hallucinations of generalization improvement. In this paper, we aim to prevent such learning behavior of applying frequency shortcuts from a data-driven perspective. Given the theoretical justification of models' biased learning behavior on different spatial frequency components, which is based on the dataset frequency properties, we argue that the learning behavior on various frequency components could be manipulated by changing the dataset statistical structure in the Fourier domain. Intuitively, as frequency shortcuts are hidden in the dominant and highly dependent frequencies of dataset structure, dynamically perturbating the over-reliance frequency components could prevent the application of frequency shortcuts. To this end, we propose two effective data augmentation modules designed to collaboratively and adaptively adjust the frequency characteristic of the dataset, aiming to dynamically influence the learning behavior of the model and ultimately serving as a strategy to mitigate shortcut learning. Code is available at AdvFrequency (https://github.com/C0notSilly/AdvFrequency).
comment: Accepted by NeurIPS 2024
☆ An Explainable Contrastive-based Dilated Convolutional Network with Transformer for Pediatric Pneumonia Detection
Pediatric pneumonia remains a significant global threat, posing a larger mortality risk than any other communicable disease. According to UNICEF, it is a leading cause of mortality in children under five and requires prompt diagnosis. Early diagnosis using chest radiographs is the prevalent standard, but limitations include low radiation levels in unprocessed images and data imbalance issues. This necessitates the development of efficient, computer-aided diagnosis techniques. To this end, we propose a novel EXplainable Contrastive-based Dilated Convolutional Network with Transformer (XCCNet) for pediatric pneumonia detection. XCCNet harnesses the spatial power of dilated convolutions and the global insights from contrastive-based transformers for effective feature refinement. A robust chest X-ray processing module tackles low-intensity radiographs, while adversarial-based data augmentation mitigates the skewed distribution of chest X-rays in the dataset. Furthermore, we actively integrate an explainability approach through feature visualization, directly aligning it with the attention region that pinpoints the presence of pneumonia or normality in radiographs. The efficacy of XCCNet is comprehensively assessed on four publicly available datasets. Extensive performance evaluation demonstrates the superiority of XCCNet compared to state-of-the-art methods.
☆ Multimodal Flare Forecasting with Deep Learning
Solar flare forecasting mainly relies on photospheric magnetograms and associated physical features to predict forthcoming flares. However, it is believed that flare initiation mechanisms often originate in the chromosphere and the lower corona. In this study, we employ deep learning as a purely data-driven approach to compare the predictive capabilities of chromospheric and coronal UV and EUV emissions across different wavelengths with those of photospheric line-of-sight magnetograms. Our findings indicate that individual EUV wavelengths can provide discriminatory power comparable or better to that of line-of-sight magnetograms. Moreover, we identify simple multimodal neural network architectures that consistently outperform single-input models, showing complementarity between the flare precursors that can be extracted from the distinct layers of the solar atmosphere. To mitigate potential biases from known misattributions in Active Region flare catalogs, our models are trained and evaluated using full-disk images and a comprehensive flare event catalog at the full-disk level. We introduce a deep-learning architecture suited for extracting temporal features from full-disk videos.
☆ Increasing Interpretability of Neural Networks By Approximating Human Visual Saliency
Understanding specifically where a model focuses on within an image is critical for human interpretability of the decision-making process. Deep learning-based solutions are prone to learning coincidental correlations in training datasets, causing over-fitting and reducing the explainability. Recent advances have shown that guiding models to human-defined regions of saliency within individual images significantly increases performance and interpretability. Human-guided models also exhibit greater generalization capabilities, as coincidental dataset features are avoided. Results show that models trained with saliency incorporation display an increase in interpretability of up to 30% over models trained without saliency information. The collection of this saliency information, however, can be costly, laborious and in some cases infeasible. To address this limitation, we propose a combination strategy of saliency incorporation and active learning to reduce the human annotation data required by 80% while maintaining the interpretability and performance increase from human saliency. Extensive experimentation outlines the effectiveness of the proposed approach across five public datasets and six active learning criteria.
☆ LMHaze: Intensity-aware Image Dehazing with a Large-scale Multi-intensity Real Haze Dataset
Image dehazing has drawn a significant attention in recent years. Learning-based methods usually require paired hazy and corresponding ground truth (haze-free) images for training. However, it is difficult to collect real-world image pairs, which prevents developments of existing methods. Although several works partially alleviate this issue by using synthetic datasets or small-scale real datasets. The haze intensity distribution bias and scene homogeneity in existing datasets limit the generalization ability of these methods, particularly when encountering images with previously unseen haze intensities. In this work, we present LMHaze, a large-scale, high-quality real-world dataset. LMHaze comprises paired hazy and haze-free images captured in diverse indoor and outdoor environments, spanning multiple scenarios and haze intensities. It contains over 5K high-resolution image pairs, surpassing the size of the biggest existing real-world dehazing dataset by over 25 times. Meanwhile, to better handle images with different haze intensities, we propose a mixture-of-experts model based on Mamba (MoE-Mamba) for dehazing, which dynamically adjusts the model parameters according to the haze intensity. Moreover, with our proposed dataset, we conduct a new large multimodal model (LMM)-based benchmark study to simulate human perception for evaluating dehazed images. Experiments demonstrate that LMHaze dataset improves the dehazing performance in real scenarios and our dehazing method provides better results compared to state-of-the-art methods.
☆ Final Report for CHESS: Cloud, High-Performance Computing, and Edge for Science and Security
Automating the theory-experiment cycle requires effective distributed workflows that utilize a computing continuum spanning lab instruments, edge sensors, computing resources at multiple facilities, data sets distributed across multiple information sources, and potentially cloud. Unfortunately, the obvious methods for constructing continuum platforms, orchestrating workflow tasks, and curating datasets over time fail to achieve scientific requirements for performance, energy, security, and reliability. Furthermore, achieving the best use of continuum resources depends upon the efficient composition and execution of workflow tasks, i.e., combinations of numerical solvers, data analytics, and machine learning. Pacific Northwest National Laboratory's LDRD "Cloud, High-Performance Computing (HPC), and Edge for Science and Security" (CHESS) has developed a set of interrelated capabilities for enabling distributed scientific workflows and curating datasets. This report describes the results and successes of CHESS from the perspective of open science.
☆ Integrated Image-Text Based on Semi-supervised Learning for Small Sample Instance Segmentation
Small sample instance segmentation is a very challenging task, and many existing methods follow the training strategy of meta-learning which pre-train models on support set and fine-tune on query set. The pre-training phase, which is highly task related, requires a significant amount of additional training time and the selection of datasets with close proximity to ensure effectiveness. The article proposes a novel small sample instance segmentation solution from the perspective of maximizing the utilization of existing information without increasing annotation burden and training costs. The proposed method designs two modules to address the problems encountered in small sample instance segmentation. First, it helps the model fully utilize unlabeled data by learning to generate pseudo labels, increasing the number of available samples. Second, by integrating the features of text and image, more accurate classification results can be obtained. These two modules are suitable for box-free and box-dependent frameworks. In the way, the proposed method not only improves the performance of small sample instance segmentation, but also greatly reduce reliance on pre-training. We have conducted experiments in three datasets from different scenes: on land, underwater and under microscope. As evidenced by our experiments, integrated image-text corrects the confidence of classification, and pseudo labels help the model obtain preciser masks. All the results demonstrate the effectiveness and superiority of our method.
☆ Label Filling via Mixed Supervision for Medical Image Segmentation from Noisy Annotations
The success of medical image segmentation usually requires a large number of high-quality labels. But since the labeling process is usually affected by the raters' varying skill levels and characteristics, the estimated masks provided by different raters usually suffer from high inter-rater variability. In this paper, we propose a simple yet effective Label Filling framework, termed as LF-Net, predicting the groundtruth segmentation label given only noisy annotations during training. The fundamental idea of label filling is to supervise the segmentation model by a subset of pixels with trustworthy labels, meanwhile filling labels of other pixels by mixed supervision. More concretely, we propose a qualified majority voting strategy, i.e., a threshold voting scheme is designed to model agreement among raters and the majority-voted labels of the selected subset of pixels are regarded as supervision. To fill labels of other pixels, two types of mixed auxiliary supervision are proposed: a soft label learned from intrinsic structures of noisy annotations, and raters' characteristics labels which propagate individual rater's characteristics information. LF-Net has two main advantages. 1) Training with trustworthy pixels incorporates training with confident supervision, guiding the direction of groundtruth label learning. 2) Two types of mixed supervision prevent over-fitting issues when the network is supervised by a subset of pixels, and guarantee high fidelity with the true label. Results on five datasets of diverse imaging modalities show that our LF-Net boosts segmentation accuracy in all datasets compared with state-of-the-art methods, with even a 7% improvement in DSC for MS lesion segmentation.
☆ Domain-Adaptive Neural Posterior Estimation for Strong Gravitational Lens Analysis
Modeling strong gravitational lenses is prohibitively expensive for modern and next-generation cosmic survey data. Neural posterior estimation (NPE), a simulation-based inference (SBI) approach, has been studied as an avenue for efficient analysis of strong lensing data. However, NPE has not been demonstrated to perform well on out-of-domain target data -- e.g., when trained on simulated data and then applied to real, observational data. In this work, we perform the first study of the efficacy of NPE in combination with unsupervised domain adaptation (UDA). The source domain is noiseless, and the target domain has noise mimicking modern cosmology surveys. We find that combining UDA and NPE improves the accuracy of the inference by 1-2 orders of magnitude and significantly improves the posterior coverage over an NPE model without UDA. We anticipate that this combination of approaches will help enable future applications of NPE models to real observational data.
comment: 20 pages, 2 figures, 2 tables
☆ Benchmarking Pathology Foundation Models: Adaptation Strategies and Scenarios
In computational pathology, several foundation models have recently emerged and demonstrated enhanced learning capability for analyzing pathology images. However, adapting these models to various downstream tasks remains challenging, particularly when faced with datasets from different sources and acquisition conditions, as well as limited data availability. In this study, we benchmark four pathology-specific foundation models across 14 datasets and two scenarios-consistency assessment and flexibility assessment-addressing diverse adaptation scenarios and downstream tasks. In the consistency assessment scenario, involving five fine-tuning methods, we found that the parameter-efficient fine-tuning approach was both efficient and effective for adapting pathology-specific foundation models to diverse datasets within the same downstream task. In the flexibility assessment scenario under data-limited environments, utilizing five few-shot learning methods, we observed that the foundation models benefited more from the few-shot learning methods that involve modification during the testing phase only. These findings provide insights that could guide the deployment of pathology-specific foundation models in real clinical settings, potentially improving the accuracy and reliability of pathology image analysis. The code for this study is available at: https://github.com/QuIIL/BenchmarkingPathologyFoundationModels.
☆ Improving the Multi-label Atomic Activity Recognition by Robust Visual Feature and Advanced Attention @ ROAD++ Atomic Activity Recognition 2024
Road++ Track3 proposes a multi-label atomic activity recognition task in traffic scenarios, which can be standardized as a 64-class multi-label video action recognition task. In the multi-label atomic activity recognition task, the robustness of visual feature extraction remains a key challenge, which directly affects the model performance and generalization ability. To cope with these issues, our team optimized three aspects: data processing, model and post-processing. Firstly, the appropriate resolution and video sampling strategy are selected, and a fixed sampling strategy is set on the validation and test sets. Secondly, in terms of model training, the team selects a variety of visual backbone networks for feature extraction, and then introduces the action-slot model, which is trained on the training and validation sets, and reasoned on the test set. Finally, for post-processing, the team combined the strengths and weaknesses of different models for weighted fusion, and the final mAP on the test set was 58%, which is 4% higher than the challenge baseline.
☆ Few-shot target-driven instance detection based on open-vocabulary object detection models
Current large open vision models could be useful for one and few-shot object recognition. Nevertheless, gradient-based re-training solutions are costly. On the other hand, open-vocabulary object detection models bring closer visual and textual concepts in the same latent space, allowing zero-shot detection via prompting at small computational cost. We propose a lightweight method to turn the latter into a one-shot or few-shot object recognition models without requiring textual descriptions. Our experiments on the TEgO dataset using the YOLO-World model as a base show that performance increases with the model size, the number of examples and the use of image augmentation.
☆ START: A Generalized State Space Model with Saliency-Driven Token-Aware Transformation NeurIPS2024
Domain Generalization (DG) aims to enable models to generalize to unseen target domains by learning from multiple source domains. Existing DG methods primarily rely on convolutional neural networks (CNNs), which inherently learn texture biases due to their limited receptive fields, making them prone to overfitting source domains. While some works have introduced transformer-based methods (ViTs) for DG to leverage the global receptive field, these methods incur high computational costs due to the quadratic complexity of self-attention. Recently, advanced state space models (SSMs), represented by Mamba, have shown promising results in supervised learning tasks by achieving linear complexity in sequence length during training and fast RNN-like computation during inference. Inspired by this, we investigate the generalization ability of the Mamba model under domain shifts and find that input-dependent matrices within SSMs could accumulate and amplify domain-specific features, thus hindering model generalization. To address this issue, we propose a novel SSM-based architecture with saliency-based token-aware transformation (namely START), which achieves state-of-the-art (SOTA) performances and offers a competitive alternative to CNNs and ViTs. Our START can selectively perturb and suppress domain-specific features in salient tokens within the input-dependent matrices of SSMs, thus effectively reducing the discrepancy between different domains. Extensive experiments on five benchmarks demonstrate that START outperforms existing SOTA DG methods with efficient linear complexity. Our code is available at https://github.com/lingeringlight/START.
comment: Accepted by NeurIPS2024. The code is available at https://github.com/lingeringlight/START
☆ Multispectral Texture Synthesis using RGB Convolutional Neural Networks
State-of-the-art RGB texture synthesis algorithms rely on style distances that are computed through statistics of deep features. These deep features are extracted by classification neural networks that have been trained on large datasets of RGB images. Extending such synthesis methods to multispectral images is not straightforward, since the pre-trained networks are designed for and have been trained on RGB images. In this work, we propose two solutions to extend these methods to multispectral imaging. Neither of them require additional training of the neural network from which the second order neural statistics are extracted. The first one consists in optimizing over batches of random triplets of spectral bands throughout training. The second one projects multispectral pixels onto a 3 dimensional space. We further explore the benefit of a color transfer operation upstream of the projection to avoid the potentially abnormal color distributions induced by the projection. Our experiments compare the performances of the various methods through different metrics. We demonstrate that they can be used to perform exemplar-based texture synthesis, achieve good visual quality and comes close to state-of-the art methods on RGB bands.
☆ Massimo: Public Queue Monitoring and Management using Mass-Spring Model
An efficient system of a queue control and regulation in public spaces is very important in order to avoid the traffic jams and to improve the customer satisfaction. This article offers a detailed road map based on a merger of intelligent systems and creating an efficient systems of queues in public places. Through the utilization of different technologies i.e. computer vision, machine learning algorithms, deep learning our system provide accurate information about the place is crowded or not and the necessary efforts to be taken.
comment: 8 pages, 6 figures, 3 algorithms, 3 tables
☆ 3D-GANTex: 3D Face Reconstruction with StyleGAN3-based Multi-View Images and 3DDFA based Mesh Generation
Geometry and texture estimation from a single face image is an ill-posed problem since there is very little information to work with. The problem further escalates when the face is rotated at a different angle. This paper tries to tackle this problem by introducing a novel method for texture estimation from a single image by first using StyleGAN and 3D Morphable Models. The method begins by generating multi-view faces using the latent space of GAN. Then 3DDFA trained on 3DMM estimates a 3D face mesh as well as a high-resolution texture map that is consistent with the estimated face shape. The result shows that the generated mesh is of high quality with near to accurate texture representation.
comment: 7 pages, 4 figures, 2 tables, pre-print version
☆ Visual Representation Learning Guided By Multi-modal Prior Knowledge
Despite the remarkable success of deep neural networks (DNNs) in computer vision, they fail to remain high-performing when facing distribution shifts between training and testing data. In this paper, we propose Knowledge-Guided Visual representation learning (KGV), a distribution-based learning approach leveraging multi-modal prior knowledge, to improve generalization under distribution shift. We use prior knowledge from two distinct modalities: 1) a knowledge graph (KG) with hierarchical and association relationships; and 2) generated synthetic images of visual elements semantically represented in the KG. The respective embeddings are generated from the given modalities in a common latent space, i.e., visual embeddings from original and synthetic images as well as knowledge graph embeddings (KGEs). These embeddings are aligned via a novel variant of translation-based KGE methods, where the node and relation embeddings of the KG are modeled as Gaussian distributions and translations respectively. We claim that incorporating multi-model prior knowledge enables more regularized learning of image representations. Thus, the models are able to better generalize across different data distributions. We evaluate KGV on different image classification tasks with major or minor distribution shifts, namely road sign classification across datasets from Germany, China, and Russia, image classification with the mini-ImageNet dataset and its variants, as well as the DVM-CAR dataset. The results demonstrate that KGV consistently exhibits higher accuracy and data efficiency than the baselines across all experiments.
☆ Zero-Shot Scene Reconstruction from Single Images with Deep Prior Assembly NeurIPS 2024
Large language and vision models have been leading a revolution in visual computing. By greatly scaling up sizes of data and model parameters, the large models learn deep priors which lead to remarkable performance in various tasks. In this work, we present deep prior assembly, a novel framework that assembles diverse deep priors from large models for scene reconstruction from single images in a zero-shot manner. We show that this challenging task can be done without extra knowledge but just simply generalizing one deep prior in one sub-task. To this end, we introduce novel methods related to poses, scales, and occlusion parsing which are keys to enable deep priors to work together in a robust way. Deep prior assembly does not require any 3D or 2D data-driven training in the task and demonstrates superior performance in generalizing priors to open-world scenes. We conduct evaluations on various datasets, and report analysis, numerical and visual comparisons with the latest methods to show our superiority. Project page: https://junshengzhou.github.io/DeepPriorAssembly.
comment: To appear at NeurIPS 2024. Project page: https://junshengzhou.github.io/DeepPriorAssembly
☆ A Paradigm Shift in Mouza Map Vectorization: A Human-Machine Collaboration Approach
Efficient vectorization of hand-drawn cadastral maps, such as Mouza maps in Bangladesh, poses a significant challenge due to their complex structures. Current manual digitization methods are time-consuming and labor-intensive. Our study proposes a semi-automated approach to streamline the digitization process, saving both time and human resources. Our methodology focuses on separating the plot boundaries and plot identifiers and applying our digitization methodology to convert both of them into vectorized format. To accomplish full vectorization, Convolutional Neural Network (CNN) models are utilized for pre-processing and plot number detection along with our smoothing algorithms based on the diversity of vector maps. The CNN models are trained with our own labeled dataset, generated from the maps, and smoothing algorithms are introduced from the various observations of the map's vector formats. Further human intervention remains essential for precision. We have evaluated our methods on several maps and provided both quantitative and qualitative results with user study. The result demonstrates that our methodology outperforms the existing map digitization processes significantly.
comment: 13 pages including reference, 14 figures, 4 tables
☆ Diffusion Transformer Policy
Recent large visual-language action models pretrained on diverse robot datasets have demonstrated the potential for generalizing to new environments with a few in-domain data. However, those approaches usually predict discretized or continuous actions by a small action head, which limits the ability in handling diverse action spaces. In contrast, we model the continuous action with a large multi-modal diffusion transformer, dubbed as Diffusion Transformer Policy, in which we directly denoise action chunks by a large transformer model rather than a small action head. By leveraging the scaling capability of transformers, the proposed approach can effectively model continuous end-effector actions across large diverse robot datasets, and achieve better generalization performance. Extensive experiments demonstrate Diffusion Transformer Policy pretrained on diverse robot data can generalize to different embodiments, including simulation environments like Maniskill2 and Calvin, as well as the real-world Franka arm. Specifically, without bells and whistles, the proposed approach achieves state-of-the-art performance with only a single third-view camera stream in the Calvin novel task setting (ABC->D), improving the average number of tasks completed in a row of 5 to 3.6, and the pretraining stage significantly facilitates the success sequence length on the Calvin by over 1.2. The code will be publicly available.
comment: Preprint
☆ AI-Driven Approaches for Glaucoma Detection -- A Comprehensive Review
The diagnosis of glaucoma plays a critical role in the management and treatment of this vision-threatening disease. Glaucoma is a group of eye diseases that cause blindness by damaging the optic nerve at the back of the eye. Often called "silent thief of sight", it exhibits no symptoms during the early stages. Therefore, early detection is crucial to prevent vision loss. With the rise of Artificial Intelligence (AI), particularly Deep Learning (DL) techniques, Computer-Aided Diagnosis (CADx) systems have emerged as promising tools to assist clinicians in accurately diagnosing glaucoma early. This paper aims to provide a comprehensive overview of AI techniques utilized in CADx systems for glaucoma diagnosis. Through a detailed analysis of current literature, we identify key gaps and challenges in these systems, emphasizing the need for improved safety, reliability, interpretability, and explainability. By identifying research gaps, we aim to advance the field of CADx systems especially for the early diagnosis of glaucoma, in order to prevent any potential loss of vision.
☆ MBPU: A Plug-and-Play State Space Model for Point Cloud Upsamping with Fast Point Rendering
The task of point cloud upsampling (PCU) is to generate dense and uniform point clouds from sparse input captured by 3D sensors like LiDAR, holding potential applications in real yet is still a challenging task. Existing deep learning-based methods have shown significant achievements in this field. However, they still face limitations in effectively handling long sequences and addressing the issue of shrinkage artifacts around the surface of the point cloud. Inspired by the newly proposed Mamba, in this paper, we introduce a network named MBPU built on top of the Mamba architecture, which performs well in long sequence modeling, especially for large-scale point cloud upsampling, and achieves fast convergence speed. Moreover, MBPU is an arbitrary-scale upsampling framework as the predictor of point distance in the point refinement phase. At the same time, we simultaneously predict the 3D position shift and 1D point-to-point distance as regression quantities to constrain the global features while ensuring the accuracy of local details. We also introduce a fast differentiable renderer to further enhance the fidelity of the upsampled point cloud and reduce artifacts. It is noted that, by the merits of our fast point rendering, MBPU yields high-quality upsampled point clouds by effectively eliminating surface noise. Extensive experiments have demonstrated that our MBPU outperforms other off-the-shelf methods in terms of point cloud upsampling, especially for large-scale point clouds.
☆ Focus on BEV: Self-calibrated Cycle View Transformation for Monocular Birds-Eye-View Segmentation
Birds-Eye-View (BEV) segmentation aims to establish a spatial mapping from the perspective view to the top view and estimate the semantic maps from monocular images. Recent studies have encountered difficulties in view transformation due to the disruption of BEV-agnostic features in image space. To tackle this issue, we propose a novel FocusBEV framework consisting of $(i)$ a self-calibrated cross view transformation module to suppress the BEV-agnostic image areas and focus on the BEV-relevant areas in the view transformation stage, $(ii)$ a plug-and-play ego-motion-based temporal fusion module to exploit the spatiotemporal structure consistency in BEV space with a memory bank, and $(iii)$ an occupancy-agnostic IoU loss to mitigate both semantic and positional uncertainties. Experimental evidence demonstrates that our approach achieves new state-of-the-art on two popular benchmarks,\ie, 29.2\% mIoU on nuScenes and 35.2\% mIoU on Argoverse.
☆ GReFEL: Geometry-Aware Reliable Facial Expression Learning under Bias and Imbalanced Data Distribution ACCV 2024
Reliable facial expression learning (FEL) involves the effective learning of distinctive facial expression characteristics for more reliable, unbiased and accurate predictions in real-life settings. However, current systems struggle with FEL tasks because of the variance in people's facial expressions due to their unique facial structures, movements, tones, and demographics. Biased and imbalanced datasets compound this challenge, leading to wrong and biased prediction labels. To tackle these, we introduce GReFEL, leveraging Vision Transformers and a facial geometry-aware anchor-based reliability balancing module to combat imbalanced data distributions, bias, and uncertainty in facial expression learning. Integrating local and global data with anchors that learn different facial data points and structural features, our approach adjusts biased and mislabeled emotions caused by intra-class disparity, inter-class similarity, and scale sensitivity, resulting in comprehensive, accurate, and reliable facial expression predictions. Our model outperforms current state-of-the-art methodologies, as demonstrated by extensive experiments on various datasets.
comment: ACCV 2024. Extended version of ARBEx (arXiv:2305.01486)
☆ Mitigating Object Hallucination via Concentric Causal Attention NeurIPS 2024
Recent Large Vision Language Models (LVLMs) present remarkable zero-shot conversational and reasoning capabilities given multimodal queries. Nevertheless, they suffer from object hallucination, a phenomenon where LVLMs are prone to generate textual responses not factually aligned with image inputs. Our pilot study reveals that object hallucination is closely tied with Rotary Position Encoding (RoPE), a widely adopted positional dependency modeling design in existing LVLMs. Due to the long-term decay in RoPE, LVLMs tend to hallucinate more when relevant visual cues are distant from instruction tokens in the multimodal input sequence. Additionally, we observe a similar effect when reversing the sequential order of visual tokens during multimodal alignment. Our tests indicate that long-term decay in RoPE poses challenges to LVLMs while capturing visual-instruction interactions across long distances. We propose Concentric Causal Attention (CCA), a simple yet effective positional alignment strategy that mitigates the impact of RoPE long-term decay in LVLMs by naturally reducing relative distance between visual and instruction tokens. With CCA, visual tokens can better interact with instruction tokens, thereby enhancing model's perception capability and alleviating object hallucination. Without bells and whistles, our positional alignment method surpasses existing hallucination mitigation strategies by large margins on multiple object hallucination benchmarks.
comment: To appear at NeurIPS 2024. Code is available at https://github.com/xing0047/cca-llava
♻ ☆ Learning Language Structures through Grounding
Language is highly structured, with syntactic and semantic structures, to some extent, agreed upon by speakers of the same language. With implicit or explicit awareness of such structures, humans can learn and use language efficiently and generalize to sentences that contain unseen words. Motivated by human language learning, in this dissertation, we consider a family of machine learning tasks that aim to learn language structures through grounding. We seek distant supervision from other data sources (i.e., grounds), including but not limited to other modalities (e.g., vision), execution results of programs, and other languages. We demonstrate the potential of this task formulation and advocate for its adoption through three schemes. In Part I, we consider learning syntactic parses through visual grounding. We propose the task of visually grounded grammar induction, present the first models to induce syntactic structures from visually grounded text and speech, and find that the visual grounding signals can help improve the parsing quality over language-only models. As a side contribution, we propose a novel evaluation metric that enables the evaluation of speech parsing without text or automatic speech recognition systems involved. In Part II, we propose two execution-aware methods to map sentences into corresponding semantic structures (i.e., programs), significantly improving compositional generalization and few-shot program synthesis. In Part III, we propose methods that learn language structures from annotations in other languages. Specifically, we propose a method that sets a new state of the art on cross-lingual word alignment. We then leverage the learned word alignments to improve the performance of zero-shot cross-lingual dependency parsing, by proposing a novel substructure-based projection method that preserves structural knowledge learned from the source language.
comment: Ph.D. Thesis
♻ ☆ Onboard Satellite Image Classification for Earth Observation: A Comparative Study of ViT Models
This study focuses on identifying the most effective pre-trained model for land use classification in onboard satellite processing, emphasizing achieving high accuracy, computational efficiency, and robustness against noisy data conditions commonly encountered during satellite-based inference. Through extensive experimentation, we compare the performance of traditional CNN-based, ResNet-based, and various pre-trained vision Transformer models. Our findings demonstrate that pre-trained Vision Transformer (ViT) models, particularly MobileViTV2 and EfficientViT-M2, outperform models trained from scratch in terms of accuracy and efficiency. These models achieve high performance with reduced computational requirements and exhibit greater resilience during inference under noisy conditions. While MobileViTV2 has excelled on clean validation data, EfficientViT-M2 has proved more robust when handling noise, making it the most suitable model for onboard satellite EO tasks. Our experimental results demonstrate that EfficientViT-M2 is the optimal choice for reliable and efficient RS-IC in satellite operations, achieving 98.76 % of accuracy, precision, and recall. Precisely, EfficientViT-M2 delivers the highest performance across all metrics, excels in training efficiency (1,000s) and inference time (10s), and demonstrates greater robustness (overall robustness score of 0.79). Consequently, EfficientViT-M2 consumes 63.93 % less power than MobileViTV2 (79.23 W) and 73.26 % less power than SwinTransformer (108.90 W). This highlights its significant advantage in energy efficiency.
♻ ☆ StochGradAdam: Accelerating Neural Networks Training with Stochastic Gradient Sampling
In this paper, we introduce StochGradAdam, a novel optimizer designed as an extension of the Adam algorithm, incorporating stochastic gradient sampling techniques to improve computational efficiency while maintaining robust performance. StochGradAdam optimizes by selectively sampling a subset of gradients during training, reducing the computational cost while preserving the advantages of adaptive learning rates and bias corrections found in Adam. Our experimental results, applied to image classification and segmentation tasks, demonstrate that StochGradAdam can achieve comparable or superior performance to Adam, even when using fewer gradient updates per iteration. By focusing on key gradient updates, StochGradAdam offers stable convergence and enhanced exploration of the loss landscape, while mitigating the impact of noisy gradients. The results suggest that this approach is particularly effective for large-scale models and datasets, providing a promising alternative to traditional optimization techniques for deep learning applications.
♻ ☆ Improving Neural Optimal Transport via Displacement Interpolation
Optimal Transport (OT) theory investigates the cost-minimizing transport map that moves a source distribution to a target distribution. Recently, several approaches have emerged for learning the optimal transport map for a given cost function using neural networks. We refer to these approaches as the OT Map. OT Map provides a powerful tool for diverse machine learning tasks, such as generative modeling and unpaired image-to-image translation. However, existing methods that utilize max-min optimization often experience training instability and sensitivity to hyperparameters. In this paper, we propose a novel method to improve stability and achieve a better approximation of the OT Map by exploiting displacement interpolation, dubbed Displacement Interpolation Optimal Transport Model (DIOTM). We derive the dual formulation of displacement interpolation at specific time $t$ and prove how these dual problems are related across time. This result allows us to utilize the entire trajectory of displacement interpolation in learning the OT Map. Our method improves the training stability and achieves superior results in estimating optimal transport maps. We demonstrate that DIOTM outperforms existing OT-based models on image-to-image translation tasks.
comment: 20 pages
♻ ☆ Playground v3: Improving Text-to-Image Alignment with Deep-Fusion Large Language Models
We introduce Playground v3 (PGv3), our latest text-to-image model that achieves state-of-the-art (SoTA) performance across multiple testing benchmarks, excels in graphic design abilities and introduces new capabilities. Unlike traditional text-to-image generative models that rely on pre-trained language models like T5 or CLIP text encoders, our approach fully integrates Large Language Models (LLMs) with a novel structure that leverages text conditions exclusively from a decoder-only LLM. Additionally, to enhance image captioning quality-we developed an in-house captioner, capable of generating captions with varying levels of detail, enriching the diversity of text structures. We also introduce a new benchmark CapsBench to evaluate detailed image captioning performance. Experimental results demonstrate that PGv3 excels in text prompt adherence, complex reasoning, and accurate text rendering. User preference studies indicate the super-human graphic design ability of our model for common design applications, such as stickers, posters, and logo designs. Furthermore, PGv3 introduces new capabilities, including precise RGB color control and robust multilingual understanding.
comment: Project page: https://playground.com/pg-v3
♻ ☆ PixelBytes: Catching Unified Embedding for Multimodal Generation
This report introduces PixelBytes Embedding, a novel approach for unified multimodal representation learning. Our method captures diverse inputs in a single, cohesive representation, enabling emergent properties for multimodal sequence generation, particularly for text and pixelated images. Inspired by state-of-the-art sequence models such as Image Transformers, PixelCNN, and Mamba-Bytes, PixelBytes aims to address the challenges of integrating different data types. We explore various model architectures, including Recurrent Neural Networks (RNNs), State Space Models (SSMs), and Attention-based models, focusing on bidirectional processing and our innovative PxBy embedding technique. Our experiments, conducted on a specialized PixelBytes Pok{\'e}mon dataset, demonstrate that bidirectional sequence models with PxBy embedding and convolutional layers can generate coherent multimodal sequences. This work contributes to the advancement of integrated AI models capable of understanding and generating multimodal data in a unified manner.
comment: This article is an earlier version of my work arXiv:2410.01820 "PixelBytes: Catching Unified Representation for Multimodal Generation."
♻ ☆ GazeMoDiff: Gaze-guided Diffusion Model for Stochastic Human Motion Prediction
Human motion prediction is important for many virtual and augmented reality (VR/AR) applications such as collision avoidance and realistic avatar generation. Existing methods have synthesised body motion only from observed past motion, despite the fact that human eye gaze is known to correlate strongly with body movements and is readily available in recent VR/AR headsets. We present GazeMoDiff - a novel gaze-guided denoising diffusion model to generate stochastic human motions. Our method first uses a gaze encoder and a motion encoder to extract the gaze and motion features respectively, then employs a graph attention network to fuse these features, and finally injects the gaze-motion features into a noise prediction network via a cross-attention mechanism to progressively generate multiple reasonable human motions in the future. Extensive experiments on the MoGaze and GIMO datasets demonstrate that our method outperforms the state-of-the-art methods by a large margin in terms of multi-modal final displacement error (17.3% on MoGaze and 13.3% on GIMO). We further conducted a human study (N=21) and validated that the motions generated by our method were perceived as both more precise and more realistic than those of prior methods. Taken together, these results reveal the significant information content available in eye gaze for stochastic human motion prediction as well as the effectiveness of our method in exploiting this information.
comment: Accepted at PG 2024. Link: https://zhiminghu.net/yan24_gazemodiff.html
♻ ☆ Competency-Aware Planning for Probabilistically Safe Navigation Under Perception Uncertainty
Perception-based navigation systems are useful for unmanned ground vehicle (UGV) navigation in complex terrains, where traditional depth-based navigation schemes are insufficient. However, these data-driven methods are highly dependent on their training data and can fail in surprising and dramatic ways with little warning. To ensure the safety of the vehicle and the surrounding environment, it is imperative that the navigation system is able to recognize the predictive uncertainty of the perception model and respond safely and effectively in the face of uncertainty. In an effort to enable safe navigation under perception uncertainty, we develop a probabilistic and reconstruction-based competency estimation (PaRCE) method to estimate the model's level of familiarity with an input image as a whole and with specific regions in the image. We find that the overall competency score can correctly predict correctly classified, misclassified, and out-of-distribution (OOD) samples. We also confirm that the regional competency maps can accurately distinguish between familiar and unfamiliar regions across images. We then use this competency information to develop a planning and control scheme that enables effective navigation while maintaining a low probability of error. We find that the competency-aware scheme greatly reduces the number of collisions with unfamiliar obstacles, compared to a baseline controller with no competency awareness. Furthermore, the regional competency information is very valuable in enabling efficient navigation.
♻ ☆ Decomposing and Interpreting Image Representations via Text in ViTs Beyond CLIP NeurIPS 2024
Recent work has explored how individual components of the CLIP-ViT model contribute to the final representation by leveraging the shared image-text representation space of CLIP. These components, such as attention heads and MLPs, have been shown to capture distinct image features like shape, color or texture. However, understanding the role of these components in arbitrary vision transformers (ViTs) is challenging. To this end, we introduce a general framework which can identify the roles of various components in ViTs beyond CLIP. Specifically, we (a) automate the decomposition of the final representation into contributions from different model components, and (b) linearly map these contributions to CLIP space to interpret them via text. Additionally, we introduce a novel scoring function to rank components by their importance with respect to specific features. Applying our framework to various ViT variants (e.g. DeiT, DINO, DINOv2, Swin, MaxViT), we gain insights into the roles of different components concerning particular image features. These insights facilitate applications such as image retrieval using text descriptions or reference images, visualizing token importance heatmaps, and mitigating spurious correlations. We release our code to reproduce the experiments at https://github.com/SriramB-98/vit-decompose
comment: NeurIPS 2024, 31 pages, 15 figures
♻ ☆ RACCooN: A Versatile Instructional Video Editing Framework with Auto-Generated Narratives
Recent video generative models primarily rely on carefully written text prompts for specific tasks, like inpainting or style editing. They require labor-intensive textual descriptions for input videos, hindering their flexibility to adapt personal/raw videos to user specifications. This paper proposes RACCooN, a versatile and user-friendly video-to-paragraph-to-video generative framework that supports multiple video editing capabilities such as removal, addition, and modification, through a unified pipeline. RACCooN consists of two principal stages: Video-to-Paragraph (V2P) and Paragraph-to-Video (P2V). In the V2P stage, we automatically describe video scenes in well-structured natural language, capturing both the holistic context and focused object details. Subsequently, in the P2V stage, users can optionally refine these descriptions to guide the video diffusion model, enabling various modifications to the input video, such as removing, changing subjects, and/or adding new objects. The proposed approach stands out from other methods through several significant contributions: (1) RACCooN suggests a multi-granular spatiotemporal pooling strategy to generate well-structured video descriptions, capturing both the broad context and object details without requiring complex human annotations, simplifying precise video content editing based on text for users. (2) Our video generative model incorporates auto-generated narratives or instructions to enhance the quality and accuracy of the generated content. (3) RACCooN also plans to imagine new objects in a given video, so users simply prompt the model to receive a detailed video editing plan for complex video editing. The proposed framework demonstrates impressive versatile capabilities in video-to-paragraph generation, video content editing, and can be incorporated into other SoTA video generative models for further enhancement.
comment: The first two authors contribute equally. Project Page: https://raccoon-mllm-gen.github.io/
♻ ☆ Human-Agent Joint Learning for Efficient Robot Manipulation Skill Acquisition
Employing a teleoperation system for gathering demonstrations offers the potential for more efficient learning of robot manipulation. However, teleoperating a robot arm equipped with a dexterous hand or gripper, via a teleoperation system presents inherent challenges due to the task's high dimensionality, complexity of motion, and differences between physiological structures. In this study, we introduce a novel system for joint learning between human operators and robots, that enables human operators to share control of a robot end-effector with a learned assistive agent, simplifies the data collection process, and facilitates simultaneous human demonstration collection and robot manipulation training. As data accumulates, the assistive agent gradually learns. Consequently, less human effort and attention are required, enhancing the efficiency of the data collection process. It also allows the human operator to adjust the control ratio to achieve a trade-off between manual and automated control. We conducted experiments in both simulated environments and physical real-world settings. Through user studies and quantitative evaluations, it is evident that the proposed system could enhance data collection efficiency and reduce the need for human adaptation while ensuring the collected data is of sufficient quality for downstream tasks. \textit{For more details, please refer to our webpage https://norweig1an.github.io/HAJL.github.io/.
comment: 8 pages, 6 figures
♻ ☆ CoTCoNet: An Optimized Coupled Transformer-Convolutional Network with an Adaptive Graph Reconstruction for Leukemia Detection
Swift and accurate blood smear analysis is an effective diagnostic method for leukemia and other hematological malignancies. However, manual leukocyte count and morphological evaluation using a microscope is time-consuming and prone to errors. Conventional image processing methods also exhibit limitations in differentiating cells due to the visual similarity between malignant and benign cell morphology. This limitation is further compounded by the skewed training data that hinders the extraction of reliable and pertinent features. In response to these challenges, we propose an optimized Coupled Transformer Convolutional Network (CoTCoNet) framework for the classification of leukemia, which employs a well-designed transformer integrated with a deep convolutional network to effectively capture comprehensive global features and scalable spatial patterns, enabling the identification of complex and large-scale hematological features. Further, the framework incorporates a graph-based feature reconstruction module to reveal the hidden or unobserved hard-to-see biological features of leukocyte cells and employs a Population-based Meta-Heuristic Algorithm for feature selection and optimization. To mitigate data imbalance issues, we employ a synthetic leukocyte generator. In the evaluation phase, we initially assess CoTCoNet on a dataset containing 16,982 annotated cells, and it achieves remarkable accuracy and F1-Score rates of 0.9894 and 0.9893, respectively. To broaden the generalizability of our model, we evaluate it across four publicly available diverse datasets, which include the aforementioned dataset. This evaluation demonstrates that our method outperforms current state-of-the-art approaches. We also incorporate an explainability approach in the form of feature visualization closely aligned with cell annotations to provide a deeper understanding of the framework.
♻ ☆ PUMA: Empowering Unified MLLM with Multi-granular Visual Generation
Recent advancements in multimodal foundation models have yielded significant progress in vision-language understanding. Initial attempts have also explored the potential of multimodal large language models (MLLMs) for visual content generation. However, existing works have insufficiently addressed the varying granularity demands of different image generation tasks within a unified MLLM paradigm - from the diversity required in text-to-image generation to the precise controllability needed in image manipulation. In this work, we propose PUMA, emPowering Unified MLLM with Multi-grAnular visual generation. PUMA unifies multi-granular visual features as both inputs and outputs of MLLMs, elegantly addressing the different granularity requirements of various image generation tasks within a unified MLLM framework. Following multimodal pretraining and task-specific instruction tuning, PUMA demonstrates proficiency in a wide range of multimodal tasks. This work represents a significant step towards a truly unified MLLM capable of adapting to the granularity demands of various visual tasks. The code and model will be released in https://github.com/rongyaofang/PUMA.
comment: Project page: https://rongyaofang.github.io/puma/
♻ ☆ Pre-processing and Compression: Understanding Hidden Representation Refinement Across Imaging Domains via Intrinsic Dimension NeurIPS 2024
In recent years, there has been interest in how geometric properties such as intrinsic dimension (ID) of a neural network's hidden representations change through its layers, and how such properties are predictive of important model behavior such as generalization ability. However, evidence has begun to emerge that such behavior can change significantly depending on the domain of the network's training data, such as natural versus medical images. Here, we further this inquiry by exploring how the ID of a network's learned representations changes through its layers, in essence, characterizing how the network successively refines the information content of input data to be used for predictions. Analyzing eleven natural and medical image datasets across six network architectures, we find that how ID changes through the network differs noticeably between natural and medical image models. Specifically, medical image models peak in representation ID earlier in the network, implying a difference in the image features and their abstractness that are typically used for downstream tasks in these domains. Additionally, we discover a strong correlation of this peak representation ID with the ID of the data in its input space, implying that the intrinsic information content of a model's learned representations is guided by that of the data it was trained on. Overall, our findings emphasize notable discrepancies in network behavior between natural and non-natural imaging domains regarding hidden representation information content, and provide further insights into how a network's learned features are shaped by its training data.
comment: Published in NeurIPS 2024 Workshop on Scientific Methods for Understanding Deep Learning (SciForDL)
♻ ☆ SETA: Semantic-Aware Token Augmentation for Domain Generalization
Domain generalization (DG) aims to enhance the model robustness against domain shifts without accessing target domains. A prevalent category of methods for DG is data augmentation, which focuses on generating virtual samples to simulate domain shifts. However, existing augmentation techniques in DG are mainly tailored for convolutional neural networks (CNNs), with limited exploration in token-based architectures, i.e., vision transformer (ViT) and multi-layer perceptrons (MLP) models. In this paper, we study the impact of prior CNN-based augmentation methods on token-based models, revealing their performance is suboptimal due to the lack of incentivizing the model to learn holistic shape information. To tackle the issue, we propose the SEmantic-aware Token Augmentation (SETA) method. SETA transforms token features by perturbing local edge cues while preserving global shape features, thereby enhancing the model learning of shape information. To further enhance the generalization ability of the model, we introduce two stylized variants of our method combined with two state-of-the-art style augmentation methods in DG. We provide a theoretical insight into our method, demonstrating its effectiveness in reducing the generalization risk bound. Comprehensive experiments on five benchmarks prove that our method achieves SOTA performances across various ViT and MLP architectures. Our code is available at https://github.com/lingeringlight/SETA.
comment: Accepted by IEEE TIP 2024. The code is available at https://github.com/lingeringlight/SETA
♻ ☆ Machine Unlearning in Forgettability Sequence
Machine unlearning (MU) is becoming a promising paradigm to achieve the "right to be forgotten", where the training trace of any chosen data points could be eliminated, while maintaining the model utility on general testing samples after unlearning. With the advancement of forgetting research, many fundamental open questions remain unanswered: do different samples exhibit varying levels of difficulty in being forgotten? Further, does the sequence in which samples are forgotten, determined by their respective difficulty levels, influence the performance of forgetting algorithms? In this paper, we identify key factor affecting unlearning difficulty and the performance of unlearning algorithms. We find that samples with higher privacy risks are more likely to be unlearning, indicating that the unlearning difficulty varies among different samples which motives a more precise unlearning mode. Built upon this insight, we propose a general unlearning framework, dubbed RSU, which consists of Ranking module and SeqUnlearn module.
comment: The senior authors of the draft are not fully convinced that the novelty is significant enough for this submission compared to the latest research progress in this area. Additionally, the senior authors have identified writing issues. Based on these two reasons, we have decided to withdraw the draft from arXiv
♻ ☆ From FDG to PSMA: A Hitchhiker's Guide to Multitracer, Multicenter Lesion Segmentation in PET/CT Imaging
Automated lesion segmentation in PET/CT scans is crucial for improving clinical workflows and advancing cancer diagnostics. However, the task is challenging due to physiological variability, different tracers used in PET imaging, and diverse imaging protocols across medical centers. To address this, the autoPET series was created to challenge researchers to develop algorithms that generalize across diverse PET/CT environments. This paper presents our solution for the autoPET III challenge, targeting multitracer, multicenter generalization using the nnU-Net framework with the ResEncL architecture. Key techniques include misalignment data augmentation and multi-modal pretraining across CT, MR, and PET datasets to provide an initial anatomical understanding. We incorporate organ supervision as a multitask approach, enabling the model to distinguish between physiological uptake and tracer-specific patterns, which is particularly beneficial in cases where no lesions are present. Compared to the default nnU-Net, which achieved a Dice score of 57.61, or the larger ResEncL (65.31) our model significantly improved performance with a Dice score of 68.40, alongside a reduction in false positive (FPvol: 7.82) and false negative (FNvol: 10.35) volumes. These results underscore the effectiveness of combining advanced network design, augmentation, pretraining, and multitask learning for PET/CT lesion segmentation. After evaluation on the test set, our approach was awarded the first place in the model-centric category (Team LesionTracer). Code is publicly available at https://github.com/MIC-DKFZ/autopet-3-submission.
comment: Winning method of the autoPET III challenge (model-centric) - Team LesionTracer
♻ ☆ Deep Correlated Prompting for Visual Recognition with Missing Modalities NeurIPS 2024
Large-scale multimodal models have shown excellent performance over a series of tasks powered by the large corpus of paired multimodal training data. Generally, they are always assumed to receive modality-complete inputs. However, this simple assumption may not always hold in the real world due to privacy constraints or collection difficulty, where models pretrained on modality-complete data easily demonstrate degraded performance on missing-modality cases. To handle this issue, we refer to prompt learning to adapt large pretrained multimodal models to handle missing-modality scenarios by regarding different missing cases as different types of input. Instead of only prepending independent prompts to the intermediate layers, we present to leverage the correlations between prompts and input features and excavate the relationships between different layers of prompts to carefully design the instructions. We also incorporate the complementary semantics of different modalities to guide the prompting design for each modality. Extensive experiments on three commonly-used datasets consistently demonstrate the superiority of our method compared to the previous approaches upon different missing scenarios. Plentiful ablations are further given to show the generalizability and reliability of our method upon different modality-missing ratios and types.
comment: NeurIPS 2024, add some results
♻ ☆ A Comprehensive Survey of Data Augmentation in Visual Reinforcement Learning
Visual reinforcement learning (RL), which makes decisions directly from high-dimensional visual inputs, has demonstrated significant potential in various domains. However, deploying visual RL techniques in the real world remains challenging due to their low sample efficiency and large generalization gaps. To tackle these obstacles, data augmentation (DA) has become a widely used technique in visual RL for acquiring sample-efficient and generalizable policies by diversifying the training data. This survey aims to provide a timely and essential review of DA techniques in visual RL in recognition of the thriving development in this field. In particular, we propose a unified framework for analyzing visual RL and understanding the role of DA in it. We then present a principled taxonomy of the existing augmentation techniques used in visual RL and conduct an in-depth discussion on how to better leverage augmented data in different scenarios. Moreover, we report a systematic empirical evaluation of DA-based techniques in visual RL and conclude by highlighting the directions for future research. As the first comprehensive survey of DA in visual RL, this work is expected to offer valuable guidance to this emerging field.
comment: A well-classified paper list that will be continuously updated can be found at https://github.com/Guozheng-Ma/DA-in-visualRL
♻ ☆ UNetMamba: An Efficient UNet-Like Mamba for Semantic Segmentation of High-Resolution Remote Sensing Images
Semantic segmentation of high-resolution remote sensing images is vital in downstream applications such as land-cover mapping, urban planning and disaster assessment.Existing Transformer-based methods suffer from the constraint between accuracy and efficiency, while the recently proposed Mamba is renowned for being efficient. Therefore, to overcome the dilemma, we propose UNetMamba, a UNet-like semantic segmentation model based on Mamba. It incorporates a mamba segmentation decoder (MSD) that can efficiently decode the complex information within high-resolution images, and a local supervision module (LSM), which is train-only but can significantly enhance the perception of local contents. Extensive experiments demonstrate that UNetMamba outperforms the state-of-the-art methods with mIoU increased by 0.87% on LoveDA and 0.39% on ISPRS Vaihingen, while achieving high efficiency through the lightweight design, less memory footprint and reduced computational cost. The source code is available at https://github.com/EnzeZhu2001/UNetMamba.
comment: 5 pages, 3 figures
♻ ☆ A gradient-based approach to fast and accurate head motion compensation in cone-beam CT
Cone-beam computed tomography (CBCT) systems, with their flexibility, present a promising avenue for direct point-of-care medical imaging, particularly in critical scenarios such as acute stroke assessment. However, the integration of CBCT into clinical workflows faces challenges, primarily linked to long scan duration resulting in patient motion during scanning and leading to image quality degradation in the reconstructed volumes. This paper introduces a novel approach to CBCT motion estimation using a gradient-based optimization algorithm, which leverages generalized derivatives of the backprojection operator for cone-beam CT geometries. Building on that, a fully differentiable target function is formulated which grades the quality of the current motion estimate in reconstruction space. We drastically accelerate motion estimation yielding a 19-fold speed-up compared to existing methods. Additionally, we investigate the architecture of networks used for quality metric regression and propose predicting voxel-wise quality maps, favoring autoencoder-like architectures over contracting ones. This modification improves gradient flow, leading to more accurate motion estimation. The presented method is evaluated through realistic experiments on head anatomy. It achieves a reduction in reprojection error from an initial average of 3mm to 0.61mm after motion compensation and consistently demonstrates superior performance compared to existing approaches. The analytic Jacobian for the backprojection operation, which is at the core of the proposed method, is made publicly available. In summary, this paper contributes to the advancement of CBCT integration into clinical workflows by proposing a robust motion estimation approach that enhances efficiency and accuracy, addressing critical challenges in time-sensitive scenarios.
comment: \copyright 2024 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works
♻ ☆ SLLEN: Semantic-aware Low-light Image Enhancement Network
How to effectively explore semantic feature is vital for low-light image enhancement (LLE). Existing methods usually utilize the semantic feature that is only drawn from the output produced by high-level semantic segmentation (SS) network. However, if the output is not accurately estimated, it would affect the high-level semantic feature (HSF) extraction, which accordingly interferes with LLE. To this end, we develop a simple and effective semantic-aware LLE network (SSLEN) composed of a LLE main-network (LLEmN) and a SS auxiliary-network (SSaN). In SLLEN, LLEmN integrates the random intermediate embedding feature (IEF), i.e., the information extracted from the intermediate layer of SSaN, together with the HSF into a unified framework for better LLE. SSaN is designed to act as a SS role to provide HSF and IEF. Moreover, thanks to a shared encoder between LLEmN and SSaN, we further propose an alternating training mechanism to facilitate the collaboration between them. Unlike currently available approaches, the proposed SLLEN is able to fully lever the semantic information, e.g., IEF, HSF, and SS dataset, to assist LLE, thereby leading to a more promising enhancement performance. Comparisons between the proposed SLLEN and other state-of-the-art techniques demonstrate the superiority of SLLEN with respect to LLE quality over all the comparable alternatives.
♻ ☆ Convex Relaxations for Isometric and Equiareal NRSfM
Extensible objects form a challenging case for NRSfM, owing to the lack of a sufficiently constrained extensible model of the point-cloud. We tackle the challenge by proposing 1) convex relaxations of the isometric model up to quasi-isometry, and 2) convex relaxations involving the equiareal deformation model, which preserves local area and has not been used in NRSfM. The equiareal model is appealing because it is physically plausible and widely applicable. However, it has two main difficulties: first, when used on its own, it is ambiguous, and second, it involves quartic, hence highly nonconvex, constraints. Our approach handles the first difficulty by mixing the equiareal with the isometric model and the second difficulty by new convex relaxations. We validate our methods on multiple real and synthetic data, including well-known benchmarks.
♻ ☆ VeLoRA: Memory Efficient Training using Rank-1 Sub-Token Projections NeurIPS 2024
Large language models (LLMs) have recently emerged as powerful tools for tackling many language-processing tasks. Despite their success, training and fine-tuning these models is still far too computationally and memory intensive. In this paper, we identify and characterise the important components needed for effective model convergence using gradient descent. In doing so we find that the intermediate activations used to implement backpropagation can be excessively compressed without incurring any degradation in performance. This result leads us to a cheap and memory-efficient algorithm for both fine-tuning and pre-training LLMs. The proposed algorithm simply divides the tokens up into smaller sub-tokens before projecting them onto a fixed 1-dimensional subspace during the forward pass. These features are then coarsely reconstructed during the backward pass to implement the update rules. We confirm the effectiveness of our algorithm as being complimentary to many state-of-the-art PEFT methods on the VTAB-1k fine-tuning benchmark. Furthermore, we outperform QLoRA for fine-tuning LLaMA and show competitive performance against other memory-efficient pre-training methods on the large-scale C4 dataset.
comment: NeurIPS 2024. Code available at https://github.com/roymiles/VeLoRA
♻ ☆ Towards Realistic Data Generation for Real-World Super-Resolution
Existing image super-resolution (SR) techniques often fail to generalize effectively in complex real-world settings due to the significant divergence between training data and practical scenarios. To address this challenge, previous efforts have either manually simulated intricate physical-based degradations or utilized learning-based techniques, yet these approaches remain inadequate for producing large-scale, realistic, and diverse data simultaneously. In this paper, we introduce a novel Realistic Decoupled Data Generator (RealDGen), an unsupervised learning data generation framework designed for real-world super-resolution. We meticulously develop content and degradation extraction strategies, which are integrated into a novel content-degradation decoupled diffusion model to create realistic low-resolution images from unpaired real LR and HR images. Extensive experiments demonstrate that RealDGen excels in generating large-scale, high-quality paired data that mirrors real-world degradations, significantly advancing the performance of popular SR models on various real-world benchmarks.
Information Retrieval 15
☆ STAR: A Simple Training-free Approach for Recommendations using Large Language Models
Recent progress in large language models (LLMs) offers promising new approaches for recommendation system (RecSys) tasks. While the current state-of-the-art methods rely on fine-tuning LLMs to achieve optimal results, this process is costly and introduces significant engineering complexities. Conversely, methods that bypass fine-tuning and use LLMs directly are less resource-intensive but often fail to fully capture both semantic and collaborative information, resulting in sub-optimal performance compared to their fine-tuned counterparts. In this paper, we propose a Simple Training-free Approach for Recommendation (STAR), a framework that utilizes LLMs and can be applied to various recommendation tasks without the need for fine-tuning. Our approach involves a retrieval stage that uses semantic embeddings from LLMs combined with collaborative user information to retrieve candidate items. We then apply an LLM for pairwise ranking to enhance next-item prediction. Experimental results on the Amazon Review dataset show competitive performance for next item prediction, even with our retrieval stage alone. Our full method achieves Hits@10 performance of +23.8% on Beauty, +37.5% on Toys and Games, and -1.8% on Sports and Outdoors relative to the best supervised models. This framework offers an effective alternative to traditional supervised models, highlighting the potential of LLMs in recommendation systems without extensive training or custom architectures.
☆ Limpeh ga li gong: Challenges in Singlish Annotations
Singlish, or Colloquial Singapore English, is a language formed from oral and social communication within multicultural Singapore. In this work, we work on a fundamental Natural Language Processing (NLP) task: Parts-Of-Speech (POS) tagging of Singlish sentences. For our analysis, we build a parallel Singlish dataset containing direct English translations and POS tags, with translation and POS annotation done by native Singlish speakers. Our experiments show that automatic transition- and transformer- based taggers perform with only $\sim 80\%$ accuracy when evaluated against human-annotated POS labels, suggesting that there is indeed room for improvement on computation analysis of the language. We provide an exposition of challenges in Singlish annotation: its inconsistencies in form and semantics, the highly context-dependent particles of the language, its structural unique expressions, and the variation of the language on different mediums. Our task definition, resultant labels and results reflects the challenges in analysing colloquial languages formulated from a variety of dialects, and paves the way for future studies beyond POS tagging.
☆ PODTILE: Facilitating Podcast Episode Browsing with Auto-generated Chapters CIKM
Listeners of long-form talk-audio content, such as podcast episodes, often find it challenging to understand the overall structure and locate relevant sections. A practical solution is to divide episodes into chapters--semantically coherent segments labeled with titles and timestamps. Since most episodes on our platform at Spotify currently lack creator-provided chapters, automating the creation of chapters is essential. Scaling the chapterization of podcast episodes presents unique challenges. First, episodes tend to be less structured than written texts, featuring spontaneous discussions with nuanced transitions. Second, the transcripts are usually lengthy, averaging about 16,000 tokens, which necessitates efficient processing that can preserve context. To address these challenges, we introduce PODTILE, a fine-tuned encoder-decoder transformer to segment conversational data. The model simultaneously generates chapter transitions and titles for the input transcript. To preserve context, each input text is augmented with global context, including the episode's title, description, and previous chapter titles. In our intrinsic evaluation, PODTILE achieved an 11% improvement in ROUGE score over the strongest baseline. Additionally, we provide insights into the practical benefits of auto-generated chapters for listeners navigating episode content. Our findings indicate that auto-generated chapters serve as a useful tool for engaging with less popular podcasts. Finally, we present empirical evidence that using chapter titles can enhance effectiveness of sparse retrieval in search tasks.
comment: 9 pages, 4 figures, CIKM industry track 2024
☆ Unleashing the Potential of Multi-Channel Fusion in Retrieval for Personalized Recommendations
Recommender systems (RS) are pivotal in managing information overload in modern digital services. A key challenge in RS is efficiently processing vast item pools to deliver highly personalized recommendations under strict latency constraints. Multi-stage cascade ranking addresses this by employing computationally efficient retrieval methods to cover diverse user interests, followed by more precise ranking models to refine the results. In the retrieval stage, multi-channel retrieval is often used to generate distinct item subsets from different candidate generators, leveraging the complementary strengths of these methods to maximize coverage. However, forwarding all retrieved items overwhelms downstream rankers, necessitating truncation. Despite advancements in individual retrieval methods, multi-channel fusion, the process of efficiently merging multi-channel retrieval results, remains underexplored. We are the first to identify and systematically investigate multi-channel fusion in the retrieval stage. Current industry practices often rely on heuristic approaches and manual designs, which often lead to suboptimal performance. Moreover, traditional gradient-based methods like SGD are unsuitable for this task due to the non-differentiable nature of the selection process. In this paper, we explore advanced channel fusion strategies by assigning systematically optimized weights to each channel. We utilize black-box optimization techniques, including the Cross Entropy Method and Bayesian Optimization for global weight optimization, alongside policy gradient-based approaches for personalized merging. Our methods enhance both personalization and flexibility, achieving significant performance improvements across multiple datasets and yielding substantial gains in real-world deployments, offering a scalable solution for optimizing multi-channel fusion in retrieval.
comment: 12 pages, 8 figures
☆ Surprising Patterns in Musical Influence Networks
Analyzing musical influence networks, such as those formed by artist influence or sampling, has provided valuable insights into contemporary Western music. Here, computational methods like centrality rankings help identify influential artists. However, little attention has been given to how influence changes over time. In this paper, we apply Bayesian Surprise to track the evolution of musical influence networks. Using two networks -- one of artist influence and another of covers, remixes, and samples -- our results reveal significant periods of change in network structure. Additionally, we demonstrate that Bayesian Surprise is a flexible framework for testing various hypotheses on network evolution with real-world data.
comment: To appear in the Latin American Musical Information Retrieval Workshop
☆ Developing Retrieval Augmented Generation (RAG) based LLM Systems from PDFs: An Experience Report
This paper presents an experience report on the development of Retrieval Augmented Generation (RAG) systems using PDF documents as the primary data source. The RAG architecture combines generative capabilities of Large Language Models (LLMs) with the precision of information retrieval. This approach has the potential to redefine how we interact with and augment both structured and unstructured knowledge in generative models to enhance transparency, accuracy, and contextuality of responses. The paper details the end-to-end pipeline, from data collection, preprocessing, to retrieval indexing and response generation, highlighting technical challenges and practical solutions. We aim to offer insights to researchers and practitioners developing similar systems using two distinct approaches: OpenAI's Assistant API with GPT Series and Llama's open-source models. The practical implications of this research lie in enhancing the reliability of generative AI systems in various sectors where domain-specific knowledge and real-time information retrieval is important. The Python code used in this work is also available at: https://github.com/GPT-Laboratory/RAG-LLM-Development-Guidebook-from-PDFs.
comment: 36 pages, 8 figures, 2 tables, and python code snippets
☆ Centrality-aware Product Retrieval and Ranking EMNLP 2024
This paper addresses the challenge of improving user experience on e-commerce platforms by enhancing product ranking relevant to users' search queries. Ambiguity and complexity of user queries often lead to a mismatch between the user's intent and retrieved product titles or documents. Recent approaches have proposed the use of Transformer-based models, which need millions of annotated query-title pairs during the pre-training stage, and this data often does not take user intent into account. To tackle this, we curate samples from existing datasets at eBay, manually annotated with buyer-centric relevance scores and centrality scores, which reflect how well the product title matches the users' intent. We introduce a User-intent Centrality Optimization (UCO) approach for existing models, which optimises for the user intent in semantic product search. To that end, we propose a dual-loss based optimisation to handle hard negatives, i.e., product titles that are semantically relevant but do not reflect the user's intent. Our contributions include curating challenging evaluation sets and implementing UCO, resulting in significant product ranking efficiency improvements observed for different evaluation metrics. Our work aims to ensure that the most buyer-centric titles for a query are ranked higher, thereby, enhancing the user experience on e-commerce platforms.
comment: EMNLP 2024: Industry track
☆ Using GPT Models for Qualitative and Quantitative News Analytics in the 2024 US Presidental Election Process
The paper considers an approach of using Google Search API and GPT-4o model for qualitative and quantitative analyses of news through retrieval-augmented generation (RAG). This approach was applied to analyze news about the 2024 US presidential election process. Different news sources for different time periods have been analyzed. Quantitative scores generated by GPT model have been analyzed using Bayesian regression to derive trend lines. The distributions found for the regression parameters allow for the analysis of uncertainty in the election process. The obtained results demonstrate that using the GPT models for news analysis, one can get informative analytics and provide key insights that can be applied in further analyses of election processes.
☆ Improve Dense Passage Retrieval with Entailment Tuning EMNLP 2024
Retrieval module can be plugged into many downstream NLP tasks to improve their performance, such as open-domain question answering and retrieval-augmented generation. The key to a retrieval system is to calculate relevance scores to query and passage pairs. However, the definition of relevance is often ambiguous. We observed that a major class of relevance aligns with the concept of entailment in NLI tasks. Based on this observation, we designed a method called entailment tuning to improve the embedding of dense retrievers. Specifically, we unify the form of retrieval data and NLI data using existence claim as a bridge. Then, we train retrievers to predict the claims entailed in a passage with a variant task of masked prediction. Our method can be efficiently plugged into current dense retrieval methods, and experiments show the effectiveness of our method.
comment: EMNLP 2024 Main
☆ Who's Who: Large Language Models Meet Knowledge Conflicts in Practice EMNLP 2024
Retrieval-augmented generation (RAG) methods are viable solutions for addressing the static memory limits of pre-trained language models. Nevertheless, encountering conflicting sources of information within the retrieval context is an inevitable practical challenge. In such situations, the language models are recommended to transparently inform users about the conflicts rather than autonomously deciding what to present based on their inherent biases. To analyze how current large language models (LLMs) align with our recommendation, we introduce WhoQA, a public benchmark dataset to examine model's behavior in knowledge conflict situations. We induce conflicts by asking about a common property among entities having the same name, resulting in questions with up to 8 distinctive answers. WhoQA evaluation set includes 5K questions across 13 Wikidata property types and 150K Wikipedia entities. Our experiments show that despite the simplicity of WhoQA questions, knowledge conflicts significantly degrades LLMs' performance in RAG settings.
comment: Accepted to EMNLP 2024 Findings
☆ A Survey of Conversational Search
As a cornerstone of modern information access, search engines have become indispensable in everyday life. With the rapid advancements in AI and natural language processing (NLP) technologies, particularly large language models (LLMs), search engines have evolved to support more intuitive and intelligent interactions between users and systems. Conversational search, an emerging paradigm for next-generation search engines, leverages natural language dialogue to facilitate complex and precise information retrieval, thus attracting significant attention. Unlike traditional keyword-based search engines, conversational search systems enhance user experience by supporting intricate queries, maintaining context over multi-turn interactions, and providing robust information integration and processing capabilities. Key components such as query reformulation, search clarification, conversational retrieval, and response generation work in unison to enable these sophisticated interactions. In this survey, we explore the recent advancements and potential future directions in conversational search, examining the critical modules that constitute a conversational search system. We highlight the integration of LLMs in enhancing these systems and discuss the challenges and opportunities that lie ahead in this dynamic field. Additionally, we provide insights into real-world applications and robust evaluations of current conversational search systems, aiming to guide future research and development in conversational search.
comment: 35 pages, 8 figures, continue to update
♻ ☆ OAEI-LLM: A Benchmark Dataset for Understanding Large Language Model Hallucinations in Ontology Matching
Hallucinations of large language models (LLMs) commonly occur in domain-specific downstream tasks, with no exception in ontology matching (OM). The prevalence of using LLMs for OM raises the need for benchmarks to better understand LLM hallucinations. The OAEI-LLM dataset is an extended version of the Ontology Alignment Evaluation Initiative (OAEI) datasets that evaluate LLM-specific hallucinations in OM tasks. We outline the methodology used in dataset construction and schema extension, and provide examples of potential use cases.
comment: 5 pages, 1 figure, 1 table
♻ ☆ LLM-based SPARQL Query Generation from Natural Language over Federated Knowledge Graphs
We introduce a Retrieval-Augmented Generation (RAG) system for translating user questions into accurate federated SPARQL queries over bioinformatics knowledge graphs (KGs) leveraging Large Language Models (LLMs). To enhance accuracy and reduce hallucinations in query generation, our system utilises metadata from the KGs, including query examples and schema information, and incorporates a validation step to correct generated queries. The system is available online at chat.expasy.org.
♻ ☆ QUIDS: Query Intent Generation via Dual Space Modeling
Query understanding is a crucial component of Information Retrieval (IR), aimed at identifying the underlying search intent of textual queries. However, most existing approaches oversimplify this task into query classification or clustering, which fails to fully capture the nuanced intent behind the query. In this paper, we address the task of query intent generation: to automatically generate detailed and precise intent descriptions for search queries using relevant and irrelevant documents given a query. These intent descriptions can help users understand why the search engine considered the top-ranked documents relevant, and provide more transparency to the retrieval process. We propose a dual-space model that uses semantic relevance and irrelevance information in the returned documents to explain the understanding of the query intent. Specifically, in the encoding process, we project, separate, and distinguish relevant and irrelevant documents in the representation space. Then, we introduce a semantic decoupling model in the novel disentangling space, where the semantics of irrelevant information are removed from the relevant space, ensuring that only the essential and relevant intent is captured. This process refines the understanding of the query and provides more accurate explanations for the search results. Experiments on benchmark data demonstrate that our methods produce high-quality query intent descriptions, outperforming existing methods for this task, as well as state-of-the-art query-based summarization methods. A token-level visualization of attention scores reveals that our model effectively reduces the focus on irrelevant intent topics. Our findings open up promising research and application directions for query intent generation, particularly in exploratory search.
♻ ☆ Lightweight Correlation-Aware Table Compression NeurIPS 2024
The growing adoption of data lakes for managing relational data necessitates efficient, open storage formats that provide high scan performance and competitive compression ratios. While existing formats achieve fast scans through lightweight encoding techniques, they have reached a plateau in terms of minimizing storage footprint. Recently, correlation-aware compression schemes have been shown to reduce file sizes further. Yet, current approaches either incur significant scan overheads or require manual specification of correlations, limiting their practicability. We present $\texttt{Virtual}$, a framework that integrates seamlessly with existing open formats to automatically leverage data correlations, achieving substantial compression gains while having minimal scan performance overhead. Experiments on data-gov datasets show that $\texttt{Virtual}$ reduces file sizes by up to 40% compared to Apache Parquet.
comment: Third Table Representation Learning Workshop (TRL @ NeurIPS 2024)
Multimedia 7
☆ AlignVSR: Audio-Visual Cross-Modal Alignment for Visual Speech Recognition
Visual Speech Recognition (VSR) aims to recognize corresponding text by analyzing visual information from lip movements. Due to the high variability and weak information of lip movements, VSR tasks require effectively utilizing any information from any source and at any level. In this paper, we propose a VSR method based on audio-visual cross-modal alignment, named AlignVSR. The method leverages the audio modality as an auxiliary information source and utilizes the global and local correspondence between the audio and visual modalities to improve visual-to-text inference. Specifically, the method first captures global alignment between video and audio through a cross-modal attention mechanism from video frames to a bank of audio units. Then, based on the temporal correspondence between audio and video, a frame-level local alignment loss is introduced to refine the global alignment, improving the utility of the audio information. Experimental results on the LRS2 and CNVSRC.Single datasets consistently show that AlignVSR outperforms several mainstream VSR methods, demonstrating its superior and robust performance.
☆ Enhancing Multimodal Affective Analysis with Learned Live Comment Features
Live comments, also known as Danmaku, are user-generated messages that are synchronized with video content. These comments overlay directly onto streaming videos, capturing viewer emotions and reactions in real-time. While prior work has leveraged live comments in affective analysis, its use has been limited due to the relative rarity of live comments across different video platforms. To address this, we first construct the Live Comment for Affective Analysis (LCAffect) dataset which contains live comments for English and Chinese videos spanning diverse genres that elicit a wide spectrum of emotions. Then, using this dataset, we use contrastive learning to train a video encoder to produce synthetic live comment features for enhanced multimodal affective content analysis. Through comprehensive experimentation on a wide range of affective analysis tasks (sentiment, emotion recognition, and sarcasm detection) in both English and Chinese, we demonstrate that these synthetic live comment features significantly improve performance over state-of-the-art methods.
☆ Shorter Is Different: Characterizing the Dynamics of Short-Form Video Platforms
The emerging short-form video platforms have been growing tremendously and become one of the leading social media recently. Although the expanded popularity of these platforms has attracted increasing research attention, there has been a lack of understanding of whether and how they deviate from traditional long-form video-sharing platforms such as YouTube and Bilibili. To address this, we conduct a large-scale data-driven analysis of Kuaishou, one of the largest short-form video platforms in China. Based on 248 million videos uploaded to the platform across all categories, we identify their notable differences from long-form video platforms through a comparison study with Bilibili, a leading long-form video platform in China. We find that videos are shortened by multiples on Kuaishou, with distinctive categorical distributions over-represented by life-related rather than interest-based videos. Users interact with videos less per view, but top videos can even more effectively acquire users' collective attention. More importantly, ordinary content creators have higher probabilities of producing hit videos. Our results shed light on the uniqueness of short-form video platforms and pave the way for future research and design for better short-form video ecology.
☆ Modelling Concurrent RTP Flows for End-to-end Predictions of QoS in Real Time Communications
The Real-time Transport Protocol (RTP)-based real-time communications (RTC) applications, exemplified by video conferencing, have experienced an unparalleled surge in popularity and development in recent years. In pursuit of optimizing their performance, the prediction of Quality of Service (QoS) metrics emerges as a pivotal endeavor, bolstering network monitoring and proactive solutions. However, contemporary approaches are confined to individual RTP flows and metrics, falling short in relationship capture and computational efficiency. To this end, we propose Packet-to-Prediction (P2P), a novel deep learning (DL) framework that hinges on raw packets to simultaneously process concurrent RTP flows and perform end-to-end prediction of multiple QoS metrics. Specifically, we implement a streamlined architecture, namely length-free Transformer with cross and neighbourhood attention, capable of handling an unlimited number of RTP flows, and employ a multi-task learning paradigm to forecast four key metrics in a single shot. Our work is based on extensive traffic collected during real video calls, and conclusively, P2P excels comparative models in both prediction performance and temporal efficiency.
☆ OpenMU: Your Swiss Army Knife for Music Understanding
We present OpenMU-Bench, a large-scale benchmark suite for addressing the data scarcity issue in training multimodal language models to understand music. To construct OpenMU-Bench, we leveraged existing datasets and bootstrapped new annotations. OpenMU-Bench also broadens the scope of music understanding by including lyrics understanding and music tool usage. Using OpenMU-Bench, we trained our music understanding model, OpenMU, with extensive ablations, demonstrating that OpenMU outperforms baseline models such as MU-Llama. Both OpenMU and OpenMU-Bench are open-sourced to facilitate future research in music understanding and to enhance creative music production efficiency.
comment: Resources: https://github.com/mzhaojp22/openmu
♻ ☆ Proceedings of The second international workshop on eXplainable AI for the Arts (XAIxArts)
This second international workshop on explainable AI for the Arts (XAIxArts) brought together a community of researchers in HCI, Interaction Design, AI, explainable AI (XAI), and digital arts to explore the role of XAI for the Arts. Workshop held at the 16th ACM Conference on Creativity and Cognition (C&C 2024), Chicago, USA.
comment: Proceedings of The second international workshop on eXplainable AI for the Arts (XAIxArts)
♻ ☆ CinePile: A Long Video Question Answering Dataset and Benchmark
Current datasets for long-form video understanding often fall short of providing genuine long-form comprehension challenges, as many tasks derived from these datasets can be successfully tackled by analyzing just one or a few random frames from a video. To address this issue, we present a novel dataset and benchmark, CinePile, specifically designed for authentic long-form video understanding. This paper details our innovative approach for creating a question-answer dataset, utilizing advanced LLMs with human-in-the-loop and building upon human-generated raw data. Our comprehensive dataset comprises 305,000 multiple-choice questions (MCQs), covering various visual and multimodal aspects, including temporal comprehension, understanding human-object interactions, and reasoning about events or actions within a scene. Additionally, we fine-tuned open-source Video-LLMs on the training split and evaluated both open-source and proprietary video-centric LLMs on the test split of our dataset. The findings indicate that although current models underperform compared to humans, fine-tuning these models can lead to significant improvements in their performance.
comment: Project page with all the artifacts - https://ruchitrawal.github.io/cinepile/. Updated version with adversarial refinement pipeline and more model evaluations
Information Retrieval 9
☆ ConTReGen: Context-driven Tree-structured Retrieval for Open-domain Long-form Text Generation EMNLP'24
Open-domain long-form text generation requires generating coherent, comprehensive responses that address complex queries with both breadth and depth. This task is challenging due to the need to accurately capture diverse facets of input queries. Existing iterative retrieval-augmented generation (RAG) approaches often struggle to delve deeply into each facet of complex queries and integrate knowledge from various sources effectively. This paper introduces ConTReGen, a novel framework that employs a context-driven, tree-structured retrieval approach to enhance the depth and relevance of retrieved content. ConTReGen integrates a hierarchical, top-down in-depth exploration of query facets with a systematic bottom-up synthesis, ensuring comprehensive coverage and coherent integration of multifaceted information. Extensive experiments on multiple datasets, including LFQA and ODSUM, alongside a newly introduced dataset, ODSUM-WikiHow, demonstrate that ConTReGen outperforms existing state-of-the-art RAG models.
comment: Accepted at EMNLP'24 Findings
☆ Deep Class-guided Hashing for Multi-label Cross-modal Retrieval
Deep hashing, due to its low cost and efficient retrieval advantages, is widely valued in cross-modal retrieval. However, existing cross-modal hashing methods either explore the relationships between data points, which inevitably leads to intra-class dispersion, or explore the relationships between data points and categories while ignoring the preservation of inter-class structural relationships, resulting in the generation of suboptimal hash codes. How to maintain both intra-class aggregation and inter-class structural relationships, In response to this issue, this paper proposes a DCGH method. Specifically, we use proxy loss as the mainstay to maintain intra-class aggregation of data, combined with pairwise loss to maintain inter-class structural relationships, and on this basis, further propose a variance constraint to address the semantic bias issue caused by the combination. A large number of comparative experiments on three benchmark datasets show that the DCGH method has comparable or even better performance compared to existing cross-modal retrieval methods. The code for the implementation of our DCGH framework is available at https://github.com/donnotnormal/DCGH.
☆ Performance-Driven QUBO for Recommender Systems on Quantum Annealers
We propose Counterfactual Analysis Quadratic Unconstrained Binary Optimization (CAQUBO) to solve QUBO problems for feature selection in recommender systems. CAQUBO leverages counterfactual analysis to measure the impact of individual features and feature combinations on model performance and employs the measurements to construct the coefficient matrix for a quantum annealer to select the optimal feature combinations for recommender systems, thereby improving their final recommendation performance. By establishing explicit connections between features and the recommendation performance, the proposed approach demonstrates superior performance compared to the state-of-the-art quantum annealing methods. Extensive experiments indicate that integrating quantum computing with counterfactual analysis holds great promise for addressing these challenges.
☆ HyQE: Ranking Contexts with Hypothetical Query Embeddings
In retrieval-augmented systems, context ranking techniques are commonly employed to reorder the retrieved contexts based on their relevance to a user query. A standard approach is to measure this relevance through the similarity between contexts and queries in the embedding space. However, such similarity often fails to capture the relevance. Alternatively, large language models (LLMs) have been used for ranking contexts. However, they can encounter scalability issues when the number of candidate contexts grows and the context window sizes of the LLMs remain constrained. Additionally, these approaches require fine-tuning LLMs with domain-specific data. In this work, we introduce a scalable ranking framework that combines embedding similarity and LLM capabilities without requiring LLM fine-tuning. Our framework uses a pre-trained LLM to hypothesize the user query based on the retrieved contexts and ranks the context based on the similarity between the hypothesized queries and the user query. Our framework is efficient at inference time and is compatible with many other retrieval and ranking techniques. Experimental results show that our method improves the ranking performance across multiple benchmarks. The complete code and data are available at https://github.com/zwc662/hyqe
♻ ☆ STaRK: Benchmarking LLM Retrieval on Textual and Relational Knowledge Bases NeurIPS 2024
Answering real-world complex queries, such as complex product search, often requires accurate retrieval from semi-structured knowledge bases that involve blend of unstructured (e.g., textual descriptions of products) and structured (e.g., entity relations of products) information. However, many previous works studied textual and relational retrieval tasks as separate topics. To address the gap, we develop STARK, a large-scale Semi-structure retrieval benchmark on Textual and Relational Knowledge Bases. Our benchmark covers three domains: product search, academic paper search, and queries in precision medicine. We design a novel pipeline to synthesize realistic user queries that integrate diverse relational information and complex textual properties, together with their ground-truth answers (items). We conduct rigorous human evaluation to validate the quality of our synthesized queries. We further enhance the benchmark with high-quality human-generated queries to provide an authentic reference. STARK serves as a comprehensive testbed for evaluating the performance of retrieval systems driven by large language models (LLMs). Our experiments suggest that STARK presents significant challenges to the current retrieval and LLM systems, highlighting the need for more capable semi-structured retrieval systems. The benchmark data and code are available on https://github.com/snap-stanford/STaRK.
comment: NeurIPS 2024 Track on Datasets and Benchmarks. 26 Pages, 6 Figures. Website: https://stark.stanford.edu/
♻ ☆ Diffusion-based Contrastive Learning for Sequential Recommendation
Contrastive learning has been effectively utilized to enhance the training of sequential recommendation models by leveraging informative self-supervised signals. Most existing approaches generate augmented views of the same user sequence through random augmentation and subsequently maximize their agreement in the representation space. However, these methods often neglect the rationality of the augmented samples. Due to significant uncertainty, random augmentation can disrupt the semantic information and interest evolution patterns inherent in the original user sequences. Moreover, pulling semantically inconsistent sequences closer in the representation space can render the user sequence embeddings insensitive to variations in user preferences, which contradicts the primary objective of sequential recommendation. To address these limitations, we propose the Context-aware Diffusion-based Contrastive Learning for Sequential Recommendation, named CaDiRec. The core idea is to leverage context information to generate more reasonable augmented views. Specifically, CaDiRec employs a context-aware diffusion model to generate alternative items for the given positions within a sequence. These generated items are aligned with their respective context information and can effectively replace the corresponding original items, thereby generating a positive view of the original sequence. By considering two different augmentations of the same user sequence, we can construct a pair of positive samples for contrastive learning. To ensure representation cohesion, we train the entire framework in an end-to-end manner, with shared item embeddings between the diffusion model and the recommendation model. Extensive experiments on five benchmark datasets demonstrate the advantages of our proposed method over existing baselines.
♻ ☆ Parenting: Optimizing Knowledge Selection of Retrieval-Augmented Language Models with Parameter Decoupling and Tailored Tuning
Retrieval-Augmented Generation (RAG) offers an effective solution to the issues faced by Large Language Models (LLMs) in hallucination generation and knowledge obsolescence by incorporating externally retrieved knowledge. However, existing methods lack effective control mechanisms for integrating internal and external knowledge. Inspired by human cognitive processes, we propose Parenting, a novel framework that decouples, identifies, and purposefully optimizes parameter subspaces related to adherence and robustness. Specifically, Parenting utilizes a key parameter mining method that combines forward and backward propagation signals to localize subspaces representing different capabilities. Then, Parenting employs a type-tailored tuning strategy, applying specific and appropriate optimizations to different subspaces, aiming to achieve a balanced enhancement of both adherence and robustness. Extensive experiments on various datasets and models validate the effectiveness and generalizability of our method.
♻ ☆ UniRAG: Universal Retrieval Augmentation for Multi-Modal Large Language Models
Recently, Multi-Modal (MM) Large Language Models (LLMs) have unlocked many complex use-cases that require MM understanding (e.g., image captioning or visual question answering) and MM generation (e.g., text-guided image generation or editing) capabilities. To further improve the output fidelity of MM-LLMs we introduce UniRAG, a plug-and-play technique that adds relevant retrieved information to prompts as few-shot examples during inference. Unlike the common belief that Retrieval Augmentation (RA) mainly improves generation or understanding of uncommon entities, our evaluation results on the MSCOCO dataset with common entities show that both proprietary models like GPT-4o and Gemini-Pro and smaller open-source models like LLaVA, LaVIT, and Emu2 significantly enhance their generation quality when their input prompts are augmented with relevant information retrieved by MM retrievers like UniIR models.
comment: 14 pages, 6 figures
♻ ☆ Multi-modal clothing recommendation model based on large model and VAE enhancement
Accurately recommending products has long been a subject requiring in-depth research. This study proposes a multimodal paradigm for clothing recommendations. Specifically, it designs a multimodal analysis method that integrates clothing description texts and images, utilizing a pre-trained large language model to deeply explore the hidden meanings of users and products. Additionally, a variational encoder is employed to learn the relationship between user information and products to address the cold start problem in recommendation systems. This study also validates the significant performance advantages of this method over various recommendation system methods through extensive ablation experiments, providing crucial practical guidance for the comprehensive optimization of recommendation systems.
Multimedia 4
☆ EVA: An Embodied World Model for Future Video Anticipation
World models integrate raw data from various modalities, such as images and language to simulate comprehensive interactions in the world, thereby displaying crucial roles in fields like mixed reality and robotics. Yet, applying the world model for accurate video prediction is quite challenging due to the complex and dynamic intentions of the various scenes in practice. In this paper, inspired by the human rethinking process, we decompose the complex video prediction into four meta-tasks that enable the world model to handle this issue in a more fine-grained manner. Alongside these tasks, we introduce a new benchmark named Embodied Video Anticipation Benchmark (EVA-Bench) to provide a well-rounded evaluation. EVA-Bench focused on evaluating the video prediction ability of human and robot actions, presenting significant challenges for both the language model and the generation model. Targeting embodied video prediction, we propose the Embodied Video Anticipator (EVA), a unified framework aiming at video understanding and generation. EVA integrates a video generation model with a visual language model, effectively combining reasoning capabilities with high-quality generation. Moreover, to enhance the generalization of our framework, we tailor-designed a multi-stage pretraining paradigm that adaptatively ensembles LoRA to produce high-fidelity results. Extensive experiments on EVA-Bench highlight the potential of EVA to significantly improve performance in embodied scenes, paving the way for large-scale pre-trained models in real-world prediction tasks.
☆ Scene Graph Generation with Role-Playing Large Language Models NeurIPS 2024
Current approaches for open-vocabulary scene graph generation (OVSGG) use vision-language models such as CLIP and follow a standard zero-shot pipeline -- computing similarity between the query image and the text embeddings for each category (i.e., text classifiers). In this work, we argue that the text classifiers adopted by existing OVSGG methods, i.e., category-/part-level prompts, are scene-agnostic as they remain unchanged across contexts. Using such fixed text classifiers not only struggles to model visual relations with high variance, but also falls short in adapting to distinct contexts. To plug these intrinsic shortcomings, we devise SDSGG, a scene-specific description based OVSGG framework where the weights of text classifiers are adaptively adjusted according to the visual content. In particular, to generate comprehensive and diverse descriptions oriented to the scene, an LLM is asked to play different roles (e.g., biologist and engineer) to analyze and discuss the descriptive features of a given scene from different views. Unlike previous efforts simply treating the generated descriptions as mutually equivalent text classifiers, SDSGG is equipped with an advanced renormalization mechanism to adjust the influence of each text classifier based on its relevance to the presented scene (this is what the term "specific" means). Furthermore, to capture the complicated interplay between subjects and objects, we propose a new lightweight module called mutual visual adapter. It refines CLIP's ability to recognize relations by learning an interaction-aware semantic space. Extensive experiments on prevalent benchmarks show that SDSGG outperforms top-leading methods by a clear margin.
comment: NeurIPS 2024. Code: https://github.com/guikunchen/SDSGG
☆ ContextDet: Temporal Action Detection with Adaptive Context Aggregation
Temporal action detection (TAD), which locates and recognizes action segments, remains a challenging task in video understanding due to variable segment lengths and ambiguous boundaries. Existing methods treat neighboring contexts of an action segment indiscriminately, leading to imprecise boundary predictions. We introduce a single-stage ContextDet framework, which makes use of large-kernel convolutions in TAD for the first time. Our model features a pyramid adaptive context aggragation (ACA) architecture, capturing long context and improving action discriminability. Each ACA level consists of two novel modules. The context attention module (CAM) identifies salient contextual information, encourages context diversity, and preserves context integrity through a context gating block (CGB). The long context module (LCM) makes use of a mixture of large- and small-kernel convolutions to adaptively gather long-range context and fine-grained local features. Additionally, by varying the length of these large kernels across the ACA pyramid, our model provides lightweight yet effective context aggregation and action discrimination. We conducted extensive experiments and compared our model with a number of advanced TAD methods on six challenging TAD benchmarks: MultiThumos, Charades, FineAction, EPIC-Kitchens 100, Thumos14, and HACS, demonstrating superior accuracy at reduced inference speed.
☆ GSSF: Generalized Structural Sparse Function for Deep Cross-modal Metric Learning
Cross-modal metric learning is a prominent research topic that bridges the semantic heterogeneity between vision and language. Existing methods frequently utilize simple cosine or complex distance metrics to transform the pairwise features into a similarity score, which suffers from an inadequate or inefficient capability for distance measurements. Consequently, we propose a Generalized Structural Sparse Function to dynamically capture thorough and powerful relationships across modalities for pair-wise similarity learning while remaining concise but efficient. Specifically, the distance metric delicately encapsulates two formats of diagonal and block-diagonal terms, automatically distinguishing and highlighting the cross-channel relevancy and dependency inside a structured and organized topology. Hence, it thereby empowers itself to adapt to the optimal matching patterns between the paired features and reaches a sweet spot between model complexity and capability. Extensive experiments on cross-modal and two extra uni-modal retrieval tasks (image-text retrieval, person re-identification, fine-grained image retrieval) have validated its superiority and flexibility over various popular retrieval frameworks. More importantly, we further discover that it can be seamlessly incorporated into multiple application scenarios, and demonstrates promising prospects from Attention Mechanism to Knowledge Distillation in a plug-and-play manner. Our code is publicly available at: https://github.com/Paranioar/GSSF.
comment: 12 pages, 9 figures, Accepted by TIP2024
Information Retrieval 10
☆ Crafting Tomorrow: The Influence of Design Choices on Fresh Content in Social Media Recommendation
The rise in popularity of social media platforms, has resulted in millions of new, content pieces being created every day. This surge in content creation underscores the need to pay attention to our design choices as they can greatly impact how long content remains relevant. In today's landscape where regularly recommending new content is crucial, particularly in the absence of detailed information, a variety of factors such as UI features, algorithms and system settings contribute to shaping the journey of content across the platform. While previous research has focused on how new content affects users' experiences, this study takes a different approach by analyzing these decisions considering the content itself. Through a series of carefully crafted experiments we explore how seemingly small decisions can influence the longevity of content, measured by metrics like Content Progression (CVP) and Content Survival (CSR). We also emphasize the importance of recognizing the stages that content goes through underscoring the need to tailor strategies for each stage as a one size fits all approach may not be effective. Additionally we argue for a departure from traditional experimental setups in the study of content lifecycles, to avoid potential misunderstandings while proposing advanced techniques, to achieve greater precision and accuracy in the evaluation process.
☆ Mining Asymmetric Intertextuality
This paper introduces a new task in Natural Language Processing (NLP) and Digital Humanities (DH): Mining Asymmetric Intertextuality. Asymmetric intertextuality refers to one-sided relationships between texts, where one text cites, quotes, or borrows from another without reciprocation. These relationships are common in literature and historical texts, where a later work references aclassical or older text that remain static. We propose a scalable and adaptive approach for mining asymmetric intertextuality, leveraging a split-normalize-merge paradigm. In this approach, documents are split into smaller chunks, normalized into structured data using LLM-assisted metadata extraction, and merged during querying to detect both explicit and implicit intertextual relationships. Our system handles intertextuality at various levels, from direct quotations to paraphrasing and cross-document influence, using a combination of metadata filtering, vector similarity search, and LLM-based verification. This method is particularly well-suited for dynamically growing corpora, such as expanding literary archives or historical databases. By enabling the continuous integration of new documents, the system can scale efficiently, making it highly valuable for digital humanities practitioners in literacy studies, historical research and related fields.
☆ Incorporating Group Prior into Variational Inference for Tail-User Behavior Modeling in CTR Prediction
User behavior modeling -- which aims to extract user interests from behavioral data -- has shown great power in Click-through rate (CTR) prediction, a key component in recommendation systems. Recently, attention-based algorithms have become a promising direction, as attention mechanisms emphasize the relevant interactions from rich behaviors. However, the methods struggle to capture the preferences of tail users with sparse interaction histories. To address the problem, we propose a novel variational inference approach, namely Group Prior Sampler Variational Inference (GPSVI), which introduces group preferences as priors to refine latent user interests for tail users. In GPSVI, the extent of adjustments depends on the estimated uncertainty of individual preference modeling. In addition, We further enhance the expressive power of variational inference by a volume-preserving flow. An appealing property of the GPSVI method is its ability to revert to traditional attention for head users with rich behavioral data while consistently enhancing performance for long-tail users with sparse behaviors. Rigorous analysis and extensive experiments demonstrate that GPSVI consistently improves the performance of tail users. Moreover, online A/B testing on a large-scale real-world recommender system further confirms the effectiveness of our proposed approach.
☆ A Recommendation Model Utilizing Separation Embedding and Self-Attention for Feature Mining
With the explosive growth of Internet data, users are facing the problem of information overload, which makes it a challenge to efficiently obtain the required resources. Recommendation systems have emerged in this context. By filtering massive amounts of information, they provide users with content that meets their needs, playing a key role in scenarios such as advertising recommendation and product recommendation. However, traditional click-through rate prediction and TOP-K recommendation mechanisms are gradually unable to meet the recommendations needs in modern life scenarios due to high computational complexity, large memory consumption, long feature selection time, and insufficient feature interaction. This paper proposes a recommendations system model based on a separation embedding cross-network. The model uses an embedding neural network layer to transform sparse feature vectors into dense embedding vectors, and can independently perform feature cross operations on different dimensions, thereby improving the accuracy and depth of feature mining. Experimental results show that the model shows stronger adaptability and higher prediction accuracy in processing complex data sets, effectively solving the problems existing in existing models.
☆ Transit Pulse: Utilizing Social Media as a Source for Customer Feedback and Information Extraction with Large Language Model
Users of the transit system flood social networks daily with messages that contain valuable insights crucial for improving service quality. These posts help transit agencies quickly identify emerging issues. Parsing topics and sentiments is key to gaining comprehensive insights to foster service excellence. However, the volume of messages makes manual analysis impractical, and standard NLP techniques like Term Frequency-Inverse Document Frequency (TF-IDF) fall short in nuanced interpretation. Traditional sentiment analysis separates topics and sentiments before integrating them, often missing the interaction between them. This incremental approach complicates classification and reduces analytical productivity. To address these challenges, we propose a novel approach to extracting and analyzing transit-related information, including sentiment and sarcasm detection, identification of unusual system problems, and location data from social media. Our method employs Large Language Models (LLM), specifically Llama 3, for a streamlined analysis free from pre-established topic labels. To enhance the model's domain-specific knowledge, we utilize Retrieval-Augmented Generation (RAG), integrating external knowledge sources into the information extraction pipeline. We validated our method through extensive experiments comparing its performance with traditional NLP approaches on user tweet data from the real world transit system. Our results demonstrate the potential of LLMs to transform social media data analysis in the public transit domain, providing actionable insights and enhancing transit agencies' responsiveness by extracting a broader range of information.
comment: 17 pages, 21 figures
☆ Visual Navigation of Digital Libraries: Retrieval and Classification of Images in the National Library of Norway's Digitised Book Collection
Digital tools for text analysis have long been essential for the searchability and accessibility of digitised library collections. Recent computer vision advances have introduced similar capabilities for visual materials, with deep learning-based embeddings showing promise for analysing visual heritage. Given that many books feature visuals in addition to text, taking advantage of these breakthroughs is critical to making library collections open and accessible. In this work, we present a proof-of-concept image search application for exploring images in the National Library of Norway's pre-1900 books, comparing Vision Transformer (ViT), Contrastive Language-Image Pre-training (CLIP), and Sigmoid loss for Language-Image Pre-training (SigLIP) embeddings for image retrieval and classification. Our results show that the application performs well for exact image retrieval, with SigLIP embeddings slightly outperforming CLIP and ViT in both retrieval and classification tasks. Additionally, SigLIP-based image classification can aid in cleaning image datasets from a digitisation pipeline.
comment: 13 pages, 2 figures, 4 tables, Accepted to the 2024 Computational Humanities Research Conference (CHR)
♻ ☆ Enhancing Short-Text Topic Modeling with LLM-Driven Context Expansion and Prefix-Tuned VAEs EMNLP
Topic modeling is a powerful technique for uncovering hidden themes within a collection of documents. However, the effectiveness of traditional topic models often relies on sufficient word co-occurrence, which is lacking in short texts. Therefore, existing approaches, whether probabilistic or neural, frequently struggle to extract meaningful patterns from such data, resulting in incoherent topics. To address this challenge, we propose a novel approach that leverages large language models (LLMs) to extend short texts into more detailed sequences before applying topic modeling. To further improve the efficiency and solve the problem of semantic inconsistency from LLM-generated texts, we propose to use prefix tuning to train a smaller language model coupled with a variational autoencoder for short-text topic modeling. Our method significantly improves short-text topic modeling performance, as demonstrated by extensive experiments on real-world datasets with extreme data sparsity, outperforming current state-of-the-art topic models.
comment: EMNLP Findings 2024. arXiv admin note: substantial text overlap with arXiv:2310.15420
♻ ☆ A Survey on Intent-aware Recommender Systems
Many modern online services feature personalized recommendations. A central challenge when providing such recommendations is that the reason why an individual user accesses the service may change from visit to visit or even during an ongoing usage session. To be effective, a recommender system should therefore aim to take the users' probable intent of using the service at a certain point in time into account. In recent years, researchers have thus started to address this challenge by incorporating intent-awareness into recommender systems. Correspondingly, a number of technical approaches were put forward, including diversification techniques, intent prediction models or latent intent modeling approaches. In this paper, we survey and categorize existing approaches to building the next generation of Intent-Aware Recommender Systems (IARS). Based on an analysis of current evaluation practices, we outline open gaps and possible future directions in this area, which in particular include the consideration of additional interaction signals and contextual information to further improve the effectiveness of such systems.
♻ ☆ Dreaming User Multimodal Representation Guided by The Platonic Representation Hypothesis for Micro-Video Recommendation
The proliferation of online micro-video platforms has underscored the necessity for advanced recommender systems to mitigate information overload and deliver tailored content. Despite advancements, accurately and promptly capturing dynamic user interests remains a formidable challenge. Inspired by the Platonic Representation Hypothesis, which posits that different data modalities converge towards a shared statistical model of reality, we introduce DreamUMM (Dreaming User Multi-Modal Representation), a novel approach leveraging user historical behaviors to create real-time user representation in a multimoda space. DreamUMM employs a closed-form solution correlating user video preferences with multimodal similarity, hypothesizing that user interests can be effectively represented in a unified multimodal space. Additionally, we propose Candidate-DreamUMM for scenarios lacking recent user behavior data, inferring interests from candidate videos alone. Extensive online A/B tests demonstrate significant improvements in user engagement metrics, including active days and play count. The successful deployment of DreamUMM in two micro-video platforms with hundreds of millions of daily active users, illustrates its practical efficacy and scalability in personalized micro-video content delivery. Our work contributes to the ongoing exploration of representational convergence by providing empirical evidence supporting the potential for user interest representations to reside in a multimodal space.
comment: 4 Figure; 2 Table
♻ ☆ PromptReps: Prompting Large Language Models to Generate Dense and Sparse Representations for Zero-Shot Document Retrieval EMNLP2024
Utilizing large language models (LLMs) for zero-shot document ranking is done in one of two ways: (1) prompt-based re-ranking methods, which require no further training but are only feasible for re-ranking a handful of candidate documents due to computational costs; and (2) unsupervised contrastive trained dense retrieval methods, which can retrieve relevant documents from the entire corpus but require a large amount of paired text data for contrastive training. In this paper, we propose PromptReps, which combines the advantages of both categories: no need for training and the ability to retrieve from the whole corpus. Our method only requires prompts to guide an LLM to generate query and document representations for effective document retrieval. Specifically, we prompt the LLMs to represent a given text using a single word, and then use the last token's hidden states and the corresponding logits associated with the prediction of the next token to construct a hybrid document retrieval system. The retrieval system harnesses both dense text embedding and sparse bag-of-words representations given by the LLM. Our experimental evaluation on the MSMARCO, TREC deep learning and BEIR zero-shot document retrieval datasets illustrates that this simple prompt-based LLM retrieval method can achieve a similar or higher retrieval effectiveness than state-of-the-art LLM embedding methods that are trained with large amounts of unsupervised data, especially when using a larger LLM.
comment: EMNLP2024 main
Multimedia 3
☆ DiffuseST: Unleashing the Capability of the Diffusion Model for Style Transfer
Style transfer aims to fuse the artistic representation of a style image with the structural information of a content image. Existing methods train specific networks or utilize pre-trained models to learn content and style features. However, they rely solely on textual or spatial representations that are inadequate to achieve the balance between content and style. In this work, we propose a novel and training-free approach for style transfer, combining textual embedding with spatial features and separating the injection of content or style. Specifically, we adopt the BLIP-2 encoder to extract the textual representation of the style image. We utilize the DDIM inversion technique to extract intermediate embeddings in content and style branches as spatial features. Finally, we harness the step-by-step property of diffusion models by separating the injection of content and style in the target branch, which improves the balance between content preservation and style fusion. Various experiments have demonstrated the effectiveness and robustness of our proposed DiffeseST for achieving balanced and controllable style transfer results, as well as the potential to extend to other tasks.
comment: Accepted to ACMMM Asia 2024. Code is available at https://github.com/I2-Multimedia-Lab/DiffuseST
☆ Testing and validation of innovative eXtended Reality technologies for astronaut training in a partial-gravity parabolic flight campaign
The use of eXtended Reality (XR) technologies in the space domain has increased significantly over the past few years as it can offer many advantages when simulating complex and challenging environments. Space agencies are currently using these disruptive tools to train astronauts for Extravehicular Activities (EVAs), to test equipment and procedures, and to assess spacecraft and hardware designs. With the Moon being the current focus of the next generation of space exploration missions, simulating its harsh environment is one of the key areas where XR can be applied, particularly for astronaut training. Peculiar lunar lighting conditions in combination with reduced gravity levels will highly impact human locomotion especially for movements such as walking, jumping, and running. In order to execute operations on the lunar surface and to safely live on the Moon for an extended period of time, innovative training methodologies and tools such as XR are becoming paramount to perform pre-mission validation and certification. This research work presents the findings of the experiments aimed at exploring the integration of XR technology and parabolic flight activities for astronaut training. In addition, the study aims to consolidate these findings into a set of guidelines that can assist future researchers who wish to incorporate XR technology into lunar training and preparation activities, including the use of such XR tools during long duration missions.
comment: 75th International Astronautical Congress (IAC), Milan, Italy, 14-18 October 2024
♻ ☆ Towards Multimodal Emotional Support Conversation Systems
The integration of conversational artificial intelligence (AI) into mental health care promises a new horizon for therapist-client interactions, aiming to closely emulate the depth and nuance of human conversations. Despite the potential, the current landscape of conversational AI is markedly limited by its reliance on single-modal data, constraining the systems' ability to empathize and provide effective emotional support. This limitation stems from a paucity of resources that encapsulate the multimodal nature of human communication essential for therapeutic counseling. To address this gap, we introduce the Multimodal Emotional Support Conversation (MESC) dataset, a first-of-its-kind resource enriched with comprehensive annotations across text, audio, and video modalities. This dataset captures the intricate interplay of user emotions, system strategies, system emotion, and system responses, setting a new precedent in the field. Leveraging the MESC dataset, we propose a general Sequential Multimodal Emotional Support framework (SMES) grounded in Therapeutic Skills Theory. Tailored for multimodal dialogue systems, the SMES framework incorporates an LLM-based reasoning model that sequentially generates user emotion recognition, system strategy prediction, system emotion prediction, and response generation. Our rigorous evaluations demonstrate that this framework significantly enhances the capability of AI systems to mimic therapist behaviors with heightened empathy and strategic responsiveness. By integrating multimodal data in this innovative manner, we bridge the critical gap between emotion recognition and emotional support, marking a significant advancement in conversational AI for mental health support.
comment: This work has been submitted to the IEEE for possible publication
Computation and Language 156
☆ Are AI Detectors Good Enough? A Survey on Quality of Datasets With Machine-Generated Texts
The rapid development of autoregressive Large Language Models (LLMs) has significantly improved the quality of generated texts, necessitating reliable machine-generated text detectors. A huge number of detectors and collections with AI fragments have emerged, and several detection methods even showed recognition quality up to 99.9% according to the target metrics in such collections. However, the quality of such detectors tends to drop dramatically in the wild, posing a question: Are detectors actually highly trustworthy or do their high benchmark scores come from the poor quality of evaluation datasets? In this paper, we emphasise the need for robust and qualitative methods for evaluating generated data to be secure against bias and low generalising ability of future model. We present a systematic review of datasets from competitions dedicated to AI-generated content detection and propose methods for evaluating the quality of datasets containing AI-generated fragments. In addition, we discuss the possibility of using high-quality generated data to achieve two goals: improving the training of detection models and improving the training datasets themselves. Our contribution aims to facilitate a better understanding of the dynamics between human and machine text, which will ultimately support the integrity of information in an increasingly automated world.
☆ SudoLM: Learning Access Control of Parametric Knowledge with Authorization Alignment
Existing preference alignment is a one-size-fits-all alignment mechanism, where the part of the large language model (LLM) parametric knowledge with non-preferred features is uniformly blocked to all the users. However, this part of knowledge can be useful to advanced users whose expertise qualifies them to handle these information. The one-size-fits-all alignment mechanism undermines LLM's utility for these qualified users. To address this problem, we propose SudoLM, a framework that lets LLMs learn access control over specific parametric knowledge for users with different credentials via authorization alignment. SudoLM allows authorized users to unlock their access to all the parametric knowledge with an assigned SUDO key while blocking access to non-qualified users. Experiments on two application scenarios demonstrate that SudoLM effectively controls the user's access to the parametric knowledge and maintains its general utility.
☆ Enhancing Large Language Models' Situated Faithfulness to External Contexts
Large Language Models (LLMs) are often augmented with external information as contexts, but this external information can sometimes be inaccurate or even intentionally misleading. We argue that robust LLMs should demonstrate situated faithfulness, dynamically calibrating their trust in external information based on their confidence in the internal knowledge and the external context. To benchmark this capability, we evaluate LLMs across several QA datasets, including a newly created dataset called RedditQA featuring in-the-wild incorrect contexts sourced from Reddit posts. We show that when provided with both correct and incorrect contexts, both open-source and proprietary models tend to overly rely on external information, regardless of its factual accuracy. To enhance situated faithfulness, we propose two approaches: Self-Guided Confidence Reasoning (SCR) and Rule-Based Confidence Reasoning (RCR). SCR enables models to self-access the confidence of external information relative to their own internal knowledge to produce the most accurate answer. RCR, in contrast, extracts explicit confidence signals from the LLM and determines the final answer using predefined rules. Our results show that for LLMs with strong reasoning capabilities, such as GPT-4o and GPT-4o mini, SCR outperforms RCR, achieving improvements of up to 24.2% over a direct input augmentation baseline. Conversely, for a smaller model like Llama-3-8B, RCR outperforms SCR. Fine-tuning SCR with our proposed Confidence Reasoning Direct Preference Optimization (CR-DPO) method improves performance on both seen and unseen datasets, yielding an average improvement of 8.9% on Llama-3-8B. In addition to quantitative results, we offer insights into the relative strengths of SCR and RCR. Our findings highlight promising avenues for improving situated faithfulness in LLMs. The data and code are released.
☆ NaturalBench: Evaluating Vision-Language Models on Natural Adversarial Samples NeurIPS 24
Vision-language models (VLMs) have made significant progress in recent visual-question-answering (VQA) benchmarks that evaluate complex visio-linguistic reasoning. However, are these models truly effective? In this work, we show that VLMs still struggle with natural images and questions that humans can easily answer, which we term natural adversarial samples. We also find it surprisingly easy to generate these VQA samples from natural image-text corpora using off-the-shelf models like CLIP and ChatGPT. We propose a semi-automated approach to collect a new benchmark, NaturalBench, for reliably evaluating VLMs with 10,000 human-verified VQA samples. Crucially, we adopt a $\textbf{vision-centric}$ design by pairing each question with two images that yield different answers, preventing blind solutions from answering without using the images. This makes NaturalBench more challenging than previous benchmarks that can be solved with commonsense priors. We evaluate 53 state-of-the-art VLMs on NaturalBench, showing that models like LLaVA-OneVision, Cambrian-1, Llama3.2-Vision, Molmo, Qwen2-VL, and even GPT-4o lag 50%-70% behind human performance (over 90%). We analyze why NaturalBench is hard from two angles: (1) Compositionality: Solving NaturalBench requires diverse visio-linguistic skills, including understanding attribute bindings, object relationships, and advanced reasoning like logic and counting. To this end, unlike prior work that uses a single tag per sample, we tag each NaturalBench sample with 1 to 8 skill tags for fine-grained evaluation. (2) Biases: NaturalBench exposes severe biases in VLMs, as models often choose the same answer regardless of the image. Lastly, we apply our benchmark curation method to diverse data sources, including long captions (over 100 words) and non-English languages like Chinese and Hindi, highlighting its potential for dynamic evaluations of VLMs.
comment: Accepted to NeurIPS 24; We open-source our dataset at: https://huggingface.co/datasets/BaiqiL/NaturalBench; Project page at: https://linzhiqiu.github.io/papers/naturalbench/
☆ MiCEval: Unveiling Multimodal Chain of Thought's Quality via Image Description and Reasoning Steps
Multimodal Chain of Thought (MCoT) is a popular prompting strategy for improving the performance of multimodal large language models (MLLMs) across a range of complex reasoning tasks. Despite its popularity, there is a notable absence of automated methods for evaluating the quality of reasoning steps in MCoT. To address this gap, we propose Multimodal Chain-of-Thought Evaluation (MiCEval), a framework designed to assess the correctness of reasoning chains by evaluating the quality of both the description and each reasoning step. The evaluation of the description component focuses on the accuracy of the image descriptions, while the reasoning step evaluates the quality of each step as it is conditionally generated based on the preceding steps. MiCEval is built upon a fine-grained dataset with annotations that rate each step according to correctness, relevance, and informativeness. Extensive experiments on four state-of-the-art MLLMs show that step-wise evaluations using MiCEval align more closely with human judgments compared to existing methods based on cosine similarity or fine-tuning approaches. MiCEval datasets and code can be found in https://github.com/alenai97/MiCEval.
comment: 40 pages
☆ DiscoGraMS: Enhancing Movie Screen-Play Summarization using Movie Character-Aware Discourse Graph
Summarizing movie screenplays presents a unique set of challenges compared to standard document summarization. Screenplays are not only lengthy, but also feature a complex interplay of characters, dialogues, and scenes, with numerous direct and subtle relationships and contextual nuances that are difficult for machine learning models to accurately capture and comprehend. Recent attempts at screenplay summarization focus on fine-tuning transformer-based pre-trained models, but these models often fall short in capturing long-term dependencies and latent relationships, and frequently encounter the "lost in the middle" issue. To address these challenges, we introduce DiscoGraMS, a novel resource that represents movie scripts as a movie character-aware discourse graph (CaD Graph). This approach is well-suited for various downstream tasks, such as summarization, question-answering, and salience detection. The model aims to preserve all salient information, offering a more comprehensive and faithful representation of the screenplay's content. We further explore a baseline method that combines the CaD Graph with the corresponding movie script through a late fusion of graph and text modalities, and we present very initial promising results.
☆ Real-time Fake News from Adversarial Feedback
We show that existing evaluations for fake news detection based on conventional sources, such as claims on fact-checking websites, result in an increasing accuracy over time for LLM-based detectors -- even after their knowledge cutoffs. This suggests that recent popular political claims, which form the majority of fake news on such sources, are easily classified using surface-level shallow patterns. Instead, we argue that a proper fake news detection dataset should test a model's ability to reason factually about the current world by retrieving and reading related evidence. To this end, we develop a novel pipeline that leverages natural language feedback from a RAG-based detector to iteratively modify real-time news into deceptive fake news that challenges LLMs. Our iterative rewrite decreases the binary classification AUC by an absolute 17.5 percent for a strong RAG GPT-4o detector. Our experiments reveal the important role of RAG in both detecting and generating fake news, as retrieval-free LLM detectors are vulnerable to unseen events and adversarial attacks, while feedback from RAG detection helps discover more deceitful patterns in fake news.
☆ Distance between Relevant Information Pieces Causes Bias in Long-Context LLMs
Positional bias in large language models (LLMs) hinders their ability to effectively process long inputs. A prominent example is the "lost in the middle" phenomenon, where LLMs struggle to utilize relevant information situated in the middle of the input. While prior research primarily focuses on single pieces of relevant information, real-world applications often involve multiple relevant information pieces. To bridge this gap, we present LongPiBench, a benchmark designed to assess positional bias involving multiple pieces of relevant information. Thorough experiments are conducted with five commercial and six open-source models. These experiments reveal that while most current models are robust against the "lost in the middle" issue, there exist significant biases related to the spacing of relevant information pieces. These findings highlight the importance of evaluating and reducing positional biases to advance LLM's capabilities.
comment: work in progress
☆ GenEOL: Harnessing the Generative Power of LLMs for Training-Free Sentence Embeddings
Training-free embedding methods directly leverage pretrained large language models (LLMs) to embed text, bypassing the costly and complex procedure of contrastive learning. Previous training-free embedding methods have mainly focused on optimizing embedding prompts and have overlooked the benefits of utilizing the generative abilities of LLMs. We propose a novel method, GenEOL, which uses LLMs to generate diverse transformations of a sentence that preserve its meaning, and aggregates the resulting embeddings of these transformations to enhance the overall sentence embedding. GenEOL significantly outperforms the existing training-free embedding methods by an average of 2.85 points across several LLMs on the sentence semantic text similarity (STS) benchmark. Our analysis shows that GenEOL stabilizes representation quality across LLM layers and is robust to perturbations of embedding prompts. GenEOL also achieves notable gains on multiple clustering, reranking and pair-classification tasks from the MTEB benchmark.
☆ Diverging Preferences: When do Annotators Disagree and do Models Know?
We examine diverging preferences in human-labeled preference datasets. We develop a taxonomy of disagreement sources spanning 10 categories across four high-level classes -- task underspecification, response style, refusals, and annotation errors. We find that the majority of disagreements are in opposition with standard reward modeling approaches, which are designed with the assumption that annotator disagreement is noise. We then explore how these findings impact two areas of LLM development: reward modeling and evaluation. In our experiments, we demonstrate how standard reward modeling methods, like the Bradley-Terry model, fail to differentiate whether a given preference judgment is the result of unanimous agreement among annotators or the majority opinion among diverging user preferences. We also find that these tendencies are also echoed by popular LLM-as-Judge evaluation methods, which consistently identify a winning response in cases of diverging preferences. These findings highlight remaining challenges in LLM evaluations, which are greatly influenced by divisive features like response style, and in developing pluralistically aligned LLMs. To address these issues, we develop methods for identifying diverging preferences to mitigate their influence on evaluation and training.
☆ CELI: Controller-Embedded Language Model Interactions
We introduce Controller-Embedded Language Model Interactions (CELI), a framework that integrates control logic directly within language model (LM) prompts, facilitating complex, multi-stage task execution. CELI addresses limitations of existing prompt engineering and workflow optimization techniques by embedding control logic directly within the operational context of language models, enabling dynamic adaptation to evolving task requirements. Our framework transfers control from the traditional programming execution environment to the LMs, allowing them to autonomously manage computational workflows while maintaining seamless interaction with external systems and functions. CELI supports arbitrary function calls with variable arguments, bridging the gap between LMs' adaptive reasoning capabilities and conventional software paradigms' structured control mechanisms. To evaluate CELI's versatility and effectiveness, we conducted case studies in two distinct domains: code generation (HumanEval benchmark) and multi-stage content generation (Wikipedia-style articles). The results demonstrate notable performance improvements across a range of domains. CELI achieved a 4.9 percentage point improvement over the best reported score of the baseline GPT-4 model on the HumanEval code generation benchmark. In multi-stage content generation, 94.4% of CELI-produced Wikipedia-style articles met or exceeded first draft quality when optimally configured, with 44.4% achieving high quality. These outcomes underscore CELI's potential for optimizing AI-driven workflows across diverse computational domains.
comment: 26 pages, 2 figures
☆ You Shall Know a Tool by the Traces it Leaves: The Predictability of Sentiment Analysis Tools
If sentiment analysis tools were valid classifiers, one would expect them to provide comparable results for sentiment classification on different kinds of corpora and for different languages. In line with results of previous studies we show that sentiment analysis tools disagree on the same dataset. Going beyond previous studies we show that the sentiment tool used for sentiment annotation can even be predicted from its outcome, revealing an algorithmic bias of sentiment analysis. Based on Twitter, Wikipedia and different news corpora from the English, German and French languages, our classifiers separate sentiment tools with an averaged F1-score of 0.89 (for the English corpora). We therefore warn against taking sentiment annotations as face value and argue for the need of more and systematic NLP evaluation studies.
☆ DiSCo Meets LLMs: A Unified Approach for Sparse Retrieval and Contextual Distillation in Conversational Search
Conversational Search (CS) is the task of retrieving relevant documents from a corpus within a conversational context, combining retrieval with conversational context modeling. With the explosion of Large Language Models (LLMs), the CS field has seen major improvements with LLMs rewriting user queries, accounting for conversational context. However, engaging LLMs at inference time harms efficiency. Current methods address this by distilling embeddings from human-rewritten queries to learn the context modeling task. Yet, these approaches predominantly focus on context modeling, and only treat the contrastive component of the retrieval task within a distillation-independent loss term. To address these limitations, we propose a new distillation method, as a relaxation of the previous objective, unifying retrieval and context modeling. We relax the existing training objectives by distilling similarity scores between conversations and documents, rather than relying solely on representation learning. Our proposed distillation objective allows for more freedom in the representation space and leverages the contrastive nature of document relevance. Through experiments on Learned Sparse Retrieval (LSR) across 5 CS datasets, our approach demonstrates substantial improvements in both in-domain and out-of-domain retrieval performance, outperforming state-of-the-art with gains of up to 6 points in recall for out-of-domain datasets. Additionally, through the relaxation of the objective, we propose a multi-teacher distillation, using multiple LLMs as teachers, yielding additional gains, and outperforming the teachers themselves in in-domain experiments. Finally, analysis of the sparsity of the models reveals that our distillation allows for better control over the sparsity of the trained models.
☆ Teaching Models to Balance Resisting and Accepting Persuasion
Large language models (LLMs) are susceptible to persuasion, which can pose risks when models are faced with an adversarial interlocutor. We take a first step towards defending models against persuasion while also arguing that defense against adversarial (i.e. negative) persuasion is only half of the equation: models should also be able to accept beneficial (i.e. positive) persuasion to improve their answers. We show that optimizing models for only one side results in poor performance on the other. In order to balance positive and negative persuasion, we introduce Persuasion-Balanced Training (or PBT), which leverages multi-agent recursive dialogue trees to create data and trains models via preference optimization to accept persuasion when appropriate. PBT consistently improves resistance to misinformation and resilience to being challenged while also resulting in the best overall performance on holistic data containing both positive and negative persuasion. Crucially, we show that PBT models are better teammates in multi-agent debates. We find that without PBT, pairs of stronger and weaker models have unstable performance, with the order in which the models present their answers determining whether the team obtains the stronger or weaker model's performance. PBT leads to better and more stable results and less order dependence, with the stronger model consistently pulling the weaker one up.
comment: Code: https://github.com/esteng/persuasion_balanced_training
☆ Toolshed: Scale Tool-Equipped Agents with Advanced RAG-Tool Fusion and Tool Knowledge Bases
Recent advancements in tool-equipped Agents (LLMs) have enabled complex tasks like secure database interactions and multi-agent code development. However, scaling tool capacity beyond agent reasoning or model limits remains a challenge. In this paper, we address these challenges by introducing Toolshed Knowledge Bases, a tool knowledge base (vector database) designed to store enhanced tool representations and optimize tool selection for large-scale tool-equipped Agents. Additionally, we propose Advanced RAG-Tool Fusion, a novel ensemble of tool-applied advanced retrieval-augmented generation (RAG) techniques across the pre-retrieval, intra-retrieval, and post-retrieval phases, without requiring model fine-tuning. During pre-retrieval, tool documents are enhanced with key information and stored in the Toolshed Knowledge Base. Intra-retrieval focuses on query planning and transformation to increase retrieval accuracy. Post-retrieval refines the retrieved tool documents and enables self-reflection. Furthermore, by varying both the total number of tools (tool-M) an Agent has access to and the tool selection threshold (top-k), we address trade-offs between retrieval accuracy, agent performance, and token cost. Our approach achieves 46%, 56%, and 47% absolute improvements on the ToolE single-tool, ToolE multi-tool and Seal-Tools benchmark datasets, respectively (Recall@5).
☆ Dialetto, ma Quanto Dialetto? Transcribing and Evaluating Dialects on a Continuum
There is increasing interest in looking at dialects in NLP. However, most work to date still treats dialects as discrete categories. For instance, evaluative work in variation-oriented NLP for English often works with Indian English or African-American Venacular English as homogeneous categories (Faisal et al., 2024; Ziems et al., 2023), yet even within one variety there is substantial variation. We examine within-dialect variation and show that performance critically varies within categories. We measure speech-to-text performance on Italian dialects, and empirically observe a geographical performance disparity. This disparity correlates substantially (-0.5) with linguistic similarity to the highest performing dialect variety. We cross-examine our results against dialectometry methods, and interpret the performance disparity to be due to a bias towards dialects that are more similar to the standard variety in the speech-to-text model examined. We additionally leverage geostatistical methods to predict zero-shot performance at unseen sites, and find the incorporation of geographical information to substantially improve prediction performance, indicating there to be geographical structure in the performance distribution.
☆ Do LLMs estimate uncertainty well in instruction-following?
Large language models (LLMs) could be valuable personal AI agents across various domains, provided they can precisely follow user instructions. However, recent studies have shown significant limitations in LLMs' instruction-following capabilities, raising concerns about their reliability in high-stakes applications. Accurately estimating LLMs' uncertainty in adhering to instructions is critical to mitigating deployment risks. We present, to our knowledge, the first systematic evaluation of the uncertainty estimation abilities of LLMs in the context of instruction-following. Our study identifies key challenges with existing instruction-following benchmarks, where multiple factors are entangled with uncertainty stems from instruction-following, complicating the isolation and comparison across methods and models. To address these issues, we introduce a controlled evaluation setup with two benchmark versions of data, enabling a comprehensive comparison of uncertainty estimation methods under various conditions. Our findings show that existing uncertainty methods struggle, particularly when models make subtle errors in instruction following. While internal model states provide some improvement, they remain inadequate in more complex scenarios. The insights from our controlled evaluation setups provide a crucial understanding of LLMs' limitations and potential for uncertainty estimation in instruction-following tasks, paving the way for more trustworthy AI agents.
☆ Optimizing Attention with Mirror Descent: Generalized Max-Margin Token Selection
Attention mechanisms have revolutionized several domains of artificial intelligence, such as natural language processing and computer vision, by enabling models to selectively focus on relevant parts of the input data. While recent work has characterized the optimization dynamics of gradient descent (GD) in attention-based models and the structural properties of its preferred solutions, less is known about more general optimization algorithms such as mirror descent (MD). In this paper, we investigate the convergence properties and implicit biases of a family of MD algorithms tailored for softmax attention mechanisms, with the potential function chosen as the $p$-th power of the $\ell_p$-norm. Specifically, we show that these algorithms converge in direction to a generalized hard-margin SVM with an $\ell_p$-norm objective when applied to a classification problem using a softmax attention model. Notably, our theoretical results reveal that the convergence rate is comparable to that of traditional GD in simpler models, despite the highly nonlinear and nonconvex nature of the present problem. Additionally, we delve into the joint optimization dynamics of the key-query matrix and the decoder, establishing conditions under which this complex joint optimization converges to their respective hard-margin SVM solutions. Lastly, our numerical experiments on real data demonstrate that MD algorithms improve generalization over standard GD and excel in optimal token selection.
☆ Large Language Models Are Overparameterized Text Encoders
Large language models (LLMs) demonstrate strong performance as text embedding models when finetuned with supervised contrastive training. However, their large size balloons inference time and memory requirements. In this paper, we show that by pruning the last $p\%$ layers of an LLM before supervised training for only 1000 steps, we can achieve a proportional reduction in memory and inference time. We evaluate four different state-of-the-art LLMs on text embedding tasks and find that our method can prune up to 30\% of layers with negligible impact on performance and up to 80\% with only a modest drop. With only three lines of code, our method is easily implemented in any pipeline for transforming LLMs to text encoders. We also propose $\text{L}^3 \text{Prune}$, a novel layer-pruning strategy based on the model's initial loss that provides two optimal pruning configurations: a large variant with negligible performance loss and a small variant for resource-constrained settings. On average, the large variant prunes 21\% of the parameters with a $-0.3$ performance drop, and the small variant only suffers from a $-5.1$ decrease while pruning 74\% of the model. We consider these results strong evidence that LLMs are overparameterized for text embedding tasks, and can be easily pruned.
comment: 8 pages of content + 1 for limitations and ethical considerations, 14 pages in total including references and appendix, 5+1 figures
☆ MomentumSMoE: Integrating Momentum into Sparse Mixture of Experts NeurIPS 2024
Sparse Mixture of Experts (SMoE) has become the key to unlocking unparalleled scalability in deep learning. SMoE has the potential to exponentially increase parameter count while maintaining the efficiency of the model by only activating a small subset of these parameters for a given sample. However, it has been observed that SMoE suffers from unstable training and has difficulty adapting to new distributions, leading to the model's lack of robustness to data contamination. To overcome these limitations, we first establish a connection between the dynamics of the expert representations in SMoEs and gradient descent on a multi-objective optimization problem. Leveraging our framework, we then integrate momentum into SMoE and propose a new family of SMoEs named MomentumSMoE. We theoretically prove and numerically demonstrate that MomentumSMoE is more stable and robust than SMoE. In particular, we verify the advantages of MomentumSMoE over SMoE on a variety of practical tasks including ImageNet-1K object recognition and WikiText-103 language modeling. We demonstrate the applicability of MomentumSMoE to many types of SMoE models, including those in the Sparse MoE model for vision (V-MoE) and the Generalist Language Model (GLaM). We also show that other advanced momentum-based optimization methods, such as Adam, can be easily incorporated into the MomentumSMoE framework for designing new SMoE models with even better performance, almost negligible additional computation cost, and simple implementations.
comment: 10 pages in the main text. Published at NeurIPS 2024. The code is available at https://github.com/rachtsy/MomentumSMoE
☆ RAG-ConfusionQA: A Benchmark for Evaluating LLMs on Confusing Questions
Conversational AI agents use Retrieval Augmented Generation (RAG) to provide verifiable document-grounded responses to user inquiries. However, many natural questions do not have good answers: about 25\% contain false assumptions~\cite{Yu2023:CREPE}, and over 50\% are ambiguous~\cite{Min2020:AmbigQA}. RAG agents need high-quality data to improve their responses to confusing questions. This paper presents a novel synthetic data generation method to efficiently create a diverse set of context-grounded confusing questions from a given document corpus. We conduct an empirical comparative evaluation of several large language models as RAG agents to measure the accuracy of confusion detection and appropriate response generation. We contribute a benchmark dataset to the public domain.
comment: under review
☆ Tell me what I need to know: Exploring LLM-based (Personalized) Abstractive Multi-Source Meeting Summarization
Meeting summarization is crucial in digital communication, but existing solutions struggle with salience identification to generate personalized, workable summaries, and context understanding to fully comprehend the meetings' content. Previous attempts to address these issues by considering related supplementary resources (e.g., presentation slides) alongside transcripts are hindered by models' limited context sizes and handling the additional complexities of the multi-source tasks, such as identifying relevant information in additional files and seamlessly aligning it with the meeting content. This work explores multi-source meeting summarization considering supplementary materials through a three-stage large language model approach: identifying transcript passages needing additional context, inferring relevant details from supplementary materials and inserting them into the transcript, and generating a summary from this enriched transcript. Our multi-source approach enhances model understanding, increasing summary relevance by ~9% and producing more content-rich outputs. We introduce a personalization protocol that extracts participant characteristics and tailors summaries accordingly, improving informativeness by ~10%. This work further provides insights on performance-cost trade-offs across four leading model families, including edge-device capable options. Our approach can be extended to similar complex generative tasks benefitting from additional resources and personalization, such as dialogue systems and action planning.
☆ Do LLMs "know" internally when they follow instructions?
Instruction-following is crucial for building AI agents with large language models (LLMs), as these models must adhere strictly to user-provided constraints and guidelines. However, LLMs often fail to follow even simple and clear instructions. To improve instruction-following behavior and prevent undesirable outputs, a deeper understanding of how LLMs' internal states relate to these outcomes is required. Our analysis of LLM internal states reveal a dimension in the input embedding space linked to successful instruction-following. We demonstrate that modifying representations along this dimension improves instruction-following success rates compared to random changes, without compromising response quality. Further investigation reveals that this dimension is more closely related to the phrasing of prompts rather than the inherent difficulty of the task or instructions. This discovery also suggests explanations for why LLMs sometimes fail to follow clear instructions and why prompt engineering is often effective, even when the content remains largely unchanged. This work provides insight into the internal workings of LLMs' instruction-following, paving the way for reliable LLM agents.
☆ SignAttention: On the Interpretability of Transformer Models for Sign Language Translation NeurIPS 2024
This paper presents the first comprehensive interpretability analysis of a Transformer-based Sign Language Translation (SLT) model, focusing on the translation from video-based Greek Sign Language to glosses and text. Leveraging the Greek Sign Language Dataset, we examine the attention mechanisms within the model to understand how it processes and aligns visual input with sequential glosses. Our analysis reveals that the model pays attention to clusters of frames rather than individual ones, with a diagonal alignment pattern emerging between poses and glosses, which becomes less distinct as the number of glosses increases. We also explore the relative contributions of cross-attention and self-attention at each decoding step, finding that the model initially relies on video frames but shifts its focus to previously predicted tokens as the translation progresses. This work contributes to a deeper understanding of SLT models, paving the way for the development of more transparent and reliable translation systems essential for real-world applications.
comment: Accepted at IAI Workshop @ NeurIPS 2024
☆ Combining Entropy and Matrix Nuclear Norm for Enhanced Evaluation of Language Models
As large language models (LLMs) continue to advance, the need for precise and efficient evaluation metrics becomes more pressing. Traditional approaches, while informative, often face limitations in computational demands and interpretability. In this paper, we introduce a novel hybrid evaluation method that integrates two established techniques: entropy derived from covariance matrices and the Matrix Nuclear Norm (MNN). Our method begins by normalizing hidden states from LLMs, then computes the covariance matrix and MNN from these representations. We further calculate the entropy of the covariance matrix to capture uncertainty and redundancy in the model's outputs. By combining these metrics into a composite score, we offer a comprehensive evaluation framework that balances accuracy with computational efficiency. Additionally, our approach allows for flexibility in adjusting the weightings between entropy and MNN, tailoring the evaluation for different objectives. Through a series of experiments on various LLMs, we demonstrate the robustness and efficacy of our method, offering deeper insights into model performance. This work contributes to the ongoing development of LLM evaluation and opens avenues for future innovations in model assessment techniques.
comment: The method is currently under experimentation
☆ A Systematic Study of Cross-Layer KV Sharing for Efficient LLM Inference
Recently, sharing key-value (KV) cache across layers has been found effective in efficient inference of large language models (LLMs). To systematically investigate different techniques of cross-layer KV sharing, we propose a unified framework that covers several recent methods and their novel variants. We conduct comprehensive experiments on all the configurations of the framework, evaluating their generation throughput and performance in language modeling and downstream tasks. We find that when reducing the size of the KV cache by 2x, most configurations can achieve competitive performance to and higher throughput than standard transformers, but when further reducing the size of the KV cache, pairing queries of all layers with KVs of upper layers can better maintain performance, although it also introduces additional training cost and prefilling latency. We hope that this work will help users choose the appropriate approach according to their requirements and facilitate research on the acceleration of LLM inference.
☆ Unlearning Backdoor Attacks for LLMs with Weak-to-Strong Knowledge Distillation
Parameter-efficient fine-tuning (PEFT) can bridge the gap between large language models (LLMs) and downstream tasks. However, PEFT has been proven vulnerable to malicious attacks. Research indicates that poisoned LLMs, even after PEFT, retain the capability to activate internalized backdoors when input samples contain predefined triggers. In this paper, we introduce a novel weak-to-strong unlearning algorithm to defend against backdoor attacks based on feature alignment knowledge distillation, named W2SDefense. Specifically, we first train a small-scale language model through full-parameter fine-tuning to serve as the clean teacher model. Then, this teacher model guides the large-scale poisoned student model in unlearning the backdoor, leveraging PEFT. Theoretical analysis suggests that W2SDefense has the potential to enhance the student model's ability to unlearn backdoor features, preventing the activation of the backdoor. We conduct experiments on text classification tasks involving three state-of-the-art language models and three different backdoor attack algorithms. Our empirical results demonstrate the outstanding performance of W2SDefense in defending against backdoor attacks without compromising model performance.
☆ Fact Recall, Heuristics or Pure Guesswork? Precise Interpretations of Language Models for Fact Completion
Previous interpretations of language models (LMs) miss important distinctions in how these models process factual information. For example, given the query "Astrid Lindgren was born in" with the corresponding completion "Sweden", no difference is made between whether the prediction was based on having the exact knowledge of the birthplace of the Swedish author or assuming that a person with a Swedish-sounding name was born in Sweden. In this paper, we investigate four different prediction scenarios for which the LM can be expected to show distinct behaviors. These scenarios correspond to different levels of model reliability and types of information being processed - some being less desirable for factual predictions. To facilitate precise interpretations of LMs for fact completion, we propose a model-specific recipe called PrISM for constructing datasets with examples of each scenario based on a set of diagnostic criteria. We apply a popular interpretability method, causal tracing (CT), to the four prediction scenarios and find that while CT produces different results for each scenario, aggregations over a set of mixed examples may only represent the results from the scenario with the strongest measured signal. In summary, we contribute tools for a more granular study of fact completion in language models and analyses that provide a more nuanced understanding of how LMs process fact-related queries.
☆ SylloBio-NLI: Evaluating Large Language Models on Biomedical Syllogistic Reasoning
Syllogistic reasoning is crucial for Natural Language Inference (NLI). This capability is particularly significant in specialized domains such as biomedicine, where it can support automatic evidence interpretation and scientific discovery. This paper presents SylloBio-NLI, a novel framework that leverages external ontologies to systematically instantiate diverse syllogistic arguments for biomedical NLI. We employ SylloBio-NLI to evaluate Large Language Models (LLMs) on identifying valid conclusions and extracting supporting evidence across 28 syllogistic schemes instantiated with human genome pathways. Extensive experiments reveal that biomedical syllogistic reasoning is particularly challenging for zero-shot LLMs, which achieve an average accuracy between 70% on generalized modus ponens and 23% on disjunctive syllogism. At the same time, we found that few-shot prompting can boost the performance of different LLMs, including Gemma (+14%) and LLama-3 (+43%). However, a deeper analysis shows that both techniques exhibit high sensitivity to superficial lexical variations, highlighting a dependency between reliability, models' architecture, and pre-training regime. Overall, our results indicate that, while in-context examples have the potential to elicit syllogistic reasoning in LLMs, existing models are still far from achieving the robustness and consistency required for safe biomedical NLI applications.
☆ Generative AI, Pragmatics, and Authenticity in Second Language Learning
There are obvious benefits to integrating generative AI (artificial intelligence) into language learning and teaching. Those include using AI as a language tutor, creating learning materials, or assessing learner output. However, due to how AI systems under-stand human language, based on a mathematical model using statistical probability, they lack the lived experience to be able to use language with the same social aware-ness as humans. Additionally, there are built-in linguistic and cultural biases based on their training data which is mostly in English and predominantly from Western sources. Those facts limit AI suitability for some language learning interactions. Stud-ies have clearly shown that systems such as ChatGPT often do not produce language that is pragmatically appropriate. The lack of linguistic and cultural authenticity has important implications for how AI is integrated into second language acquisition as well as in instruction targeting development of intercultural communication compe-tence.
☆ Analyzing Context Utilization of LLMs in Document-Level Translation
Large language models (LLM) are increasingly strong contenders in machine translation. We study document-level translation, where some words cannot be translated without context from outside the sentence. We investigate the ability of prominent LLMs to utilize context by analyzing models' robustness to perturbed and randomized document context. We find that LLMs' improved document-translation performance is not always reflected in pronoun translation performance. We highlight the need for context-aware finetuning of LLMs with a focus on relevant parts of the context to improve their reliability for document-level translation.
comment: 4 pages, 2 figures, 2 tables
☆ How Do Multilingual Models Remember? Investigating Multilingual Factual Recall Mechanisms
Large Language Models (LLMs) store and retrieve vast amounts of factual knowledge acquired during pre-training. Prior research has localized and identified mechanisms behind knowledge recall; however, it has primarily focused on English monolingual models. The question of how these processes generalize to other languages and multilingual LLMs remains unexplored. In this paper, we address this gap by conducting a comprehensive analysis of two highly multilingual LLMs. We assess the extent to which previously identified components and mechanisms of factual recall in English apply to a multilingual context. Then, we examine when language plays a role in the recall process, uncovering evidence of language-independent and language-dependent mechanisms.
☆ Fine-Tuning Pre-trained Language Models for Robust Causal Representation Learning
The fine-tuning of pre-trained language models (PLMs) has been shown to be effective across various domains. By using domain-specific supervised data, the general-purpose representation derived from PLMs can be transformed into a domain-specific representation. However, these methods often fail to generalize to out-of-domain (OOD) data due to their reliance on non-causal representations, often described as spurious features. Existing methods either make use of adjustments with strong assumptions about lack of hidden common causes, or mitigate the effect of spurious features using multi-domain data. In this work, we investigate how fine-tuned pre-trained language models aid generalizability from single-domain scenarios under mild assumptions, targeting more general and practical real-world scenarios. We show that a robust representation can be derived through a so-called causal front-door adjustment, based on a decomposition assumption, using fine-tuned representations as a source of data augmentation. Comprehensive experiments in both synthetic and real-world settings demonstrate the superior generalizability of the proposed method compared to existing approaches. Our work thus sheds light on the domain generalization problem by introducing links between fine-tuning and causal mechanisms into representation learning.
☆ Efficiently Computing Susceptibility to Context in Language Models
One strength of modern language models is their ability to incorporate information from a user-input context when answering queries. However, they are not equally sensitive to the subtle changes to that context. To quantify this, Du et al. (2024) gives an information-theoretic metric to measure such sensitivity. Their metric, susceptibility, is defined as the degree to which contexts can influence a model's response to a query at a distributional level. However, exactly computing susceptibility is difficult and, thus, Du et al. (2024) falls back on a Monte Carlo approximation. Due to the large number of samples required, the Monte Carlo approximation is inefficient in practice. As a faster alternative, we propose Fisher susceptibility, an efficient method to estimate the susceptibility based on Fisher information. Empirically, we validate that Fisher susceptibility is comparable to Monte Carlo estimated susceptibility across a diverse set of query domains despite its being $70\times$ faster. Exploiting the improved efficiency, we apply Fisher susceptibility to analyze factors affecting the susceptibility of language models. We observe that larger models are as susceptible as smaller ones.
☆ Critical Questions Generation: Motivation and Challenges CoNLL 2024
The development of Large Language Models (LLMs) has brought impressive performances on mitigation strategies against misinformation, such as counterargument generation. However, LLMs are still seriously hindered by outdated knowledge and by their tendency to generate hallucinated content. In order to circumvent these issues, we propose a new task, namely, Critical Questions Generation, consisting of processing an argumentative text to generate the critical questions (CQs) raised by it. In argumentation theory CQs are tools designed to lay bare the blind spots of an argument by pointing at the information it could be missing. Thus, instead of trying to deploy LLMs to produce knowledgeable and relevant counterarguments, we use them to question arguments, without requiring any external knowledge. Research on CQs Generation using LLMs requires a reference dataset for large scale experimentation. Thus, in this work we investigate two complementary methods to create such a resource: (i) instantiating CQs templates as defined by Walton's argumentation theory and (ii), using LLMs as CQs generators. By doing so, we contribute with a procedure to establish what is a valid CQ and conclude that, while LLMs are reasonable CQ generators, they still have a wide margin for improvement in this task.
comment: 14 pages, 3 figures, 7 tables, to be published in the 28th Conference on Computational Natural Language Learning (CoNLL 2024)
☆ LoGU: Long-form Generation with Uncertainty Expressions
While Large Language Models (LLMs) demonstrate impressive capabilities, they still struggle with generating factually incorrect content (i.e., hallucinations). A promising approach to mitigate this issue is enabling models to express uncertainty when unsure. Previous research on uncertainty modeling has primarily focused on short-form QA, but realworld applications often require much longer responses. In this work, we introduce the task of Long-form Generation with Uncertainty(LoGU). We identify two key challenges: Uncertainty Suppression, where models hesitate to express uncertainty, and Uncertainty Misalignment, where models convey uncertainty inaccurately. To tackle these challenges, we propose a refinement-based data collection framework and a two-stage training pipeline. Our framework adopts a divide-and-conquer strategy, refining uncertainty based on atomic claims. The collected data are then used in training through supervised fine-tuning (SFT) and direct preference optimization (DPO) to enhance uncertainty expression. Extensive experiments on three long-form instruction following datasets show that our method significantly improves accuracy, reduces hallucinations, and maintains the comprehensiveness of responses.
☆ SwaQuAD-24: QA Benchmark Dataset in Swahili
This paper proposes the creation of a Swahili Question Answering (QA) benchmark dataset, aimed at addressing the underrepresentation of Swahili in natural language processing (NLP). Drawing from established benchmarks like SQuAD, GLUE, KenSwQuAD, and KLUE, the dataset will focus on providing high-quality, annotated question-answer pairs that capture the linguistic diversity and complexity of Swahili. The dataset is designed to support a variety of applications, including machine translation, information retrieval, and social services like healthcare chatbots. Ethical considerations, such as data privacy, bias mitigation, and inclusivity, are central to the dataset development. Additionally, the paper outlines future expansion plans to include domain-specific content, multimodal integration, and broader crowdsourcing efforts. The Swahili QA dataset aims to foster technological innovation in East Africa and provide an essential resource for NLP research and applications in low-resource languages.
☆ EcomEdit: An Automated E-commerce Knowledge Editing Framework for Enhanced Product and Purchase Intention Understanding
Knowledge Editing (KE) aims to correct and update factual information in Large Language Models (LLMs) to ensure accuracy and relevance without computationally expensive fine-tuning. Though it has been proven effective in several domains, limited work has focused on its application within the e-commerce sector. However, there are naturally occurring scenarios that make KE necessary in this domain, such as the timely updating of product features and trending purchase intentions by customers, which necessitate further exploration. In this paper, we pioneer the application of KE in the e-commerce domain by presenting ECOMEDIT, an automated e-commerce knowledge editing framework tailored for e-commerce-related knowledge and tasks. Our framework leverages more powerful LLMs as judges to enable automatic knowledge conflict detection and incorporates conceptualization to enhance the semantic coverage of the knowledge to be edited. Through extensive experiments, we demonstrate the effectiveness of ECOMEDIT in improving LLMs' understanding of product descriptions and purchase intentions. We also show that LLMs, after our editing, can achieve stronger performance on downstream e-commerce tasks.
☆ REEF: Representation Encoding Fingerprints for Large Language Models
Protecting the intellectual property of open-source Large Language Models (LLMs) is very important, because training LLMs costs extensive computational resources and data. Therefore, model owners and third parties need to identify whether a suspect model is a subsequent development of the victim model. To this end, we propose a training-free REEF to identify the relationship between the suspect and victim models from the perspective of LLMs' feature representations. Specifically, REEF computes and compares the centered kernel alignment similarity between the representations of a suspect model and a victim model on the same samples. This training-free REEF does not impair the model's general capabilities and is robust to sequential fine-tuning, pruning, model merging, and permutations. In this way, REEF provides a simple and effective way for third parties and models' owners to protect LLMs' intellectual property together. The code is available at https://github.com/tmylla/REEF.
☆ MoDification: Mixture of Depths Made Easy
Long-context efficiency has recently become a trending topic in serving large language models (LLMs). And mixture of depths (MoD) is proposed as a perfect fit to bring down both latency and memory. In this paper, however, we discover that MoD can barely transform existing LLMs without costly training over an extensive number of tokens. To enable the transformations from any LLMs to MoD ones, we showcase top-k operator in MoD should be promoted to threshold-p operator, and refinement to architecture and data should also be crafted along. All these designs form our method termed MoDification. Through a comprehensive set of experiments covering model scales from 3B to 70B, we exhibit MoDification strikes an excellent balance between efficiency and effectiveness. MoDification can achieve up to ~1.2x speedup in latency and ~1.8x reduction in memory compared to original LLMs especially in long-context applications.
comment: 12 pages, 9 figures, 5 tables, work in progress
☆ Good Parenting is all you need -- Multi-agentic LLM Hallucination Mitigation
This study explores the ability of Large Language Model (LLM) agents to detect and correct hallucinations in AI-generated content. A primary agent was tasked with creating a blog about a fictional Danish artist named Flipfloppidy, which was then reviewed by another agent for factual inaccuracies. Most LLMs hallucinated the existence of this artist. Across 4,900 test runs involving various combinations of primary and reviewing agents, advanced AI models such as Llama3-70b and GPT-4 variants demonstrated near-perfect accuracy in identifying hallucinations and successfully revised outputs in 85% to 100% of cases following feedback. These findings underscore the potential of advanced AI models to significantly enhance the accuracy and reliability of generated content, providing a promising approach to improving AI workflow orchestration.
☆ Beyond Binary: Towards Fine-Grained LLM-Generated Text Detection via Role Recognition and Involvement Measurement
The rapid development of large language models (LLMs), like ChatGPT, has resulted in the widespread presence of LLM-generated content on social media platforms, raising concerns about misinformation, data biases, and privacy violations, which can undermine trust in online discourse. While detecting LLM-generated content is crucial for mitigating these risks, current methods often focus on binary classification, failing to address the complexities of real-world scenarios like human-AI collaboration. To move beyond binary classification and address these challenges, we propose a new paradigm for detecting LLM-generated content. This approach introduces two novel tasks: LLM Role Recognition (LLM-RR), a multi-class classification task that identifies specific roles of LLM in content generation, and LLM Influence Measurement (LLM-IM), a regression task that quantifies the extent of LLM involvement in content creation. To support these tasks, we propose LLMDetect, a benchmark designed to evaluate detectors' performance on these new tasks. LLMDetect includes the Hybrid News Detection Corpus (HNDC) for training detectors, as well as DetectEval, a comprehensive evaluation suite that considers five distinct cross-context variations and multi-intensity variations within the same LLM role. This allows for a thorough assessment of detectors' generalization and robustness across diverse contexts. Our empirical validation of 10 baseline detection methods demonstrates that fine-tuned PLM-based models consistently outperform others on both tasks, while advanced LLMs face challenges in accurately detecting their own generated content. Our experimental results and analysis offer insights for developing more effective detection models for LLM-generated content. This research enhances the understanding of LLM-generated content and establishes a foundation for more nuanced detection methodologies.
comment: Social Media, Large Language Models, LLM-generated Text Detection, AI-assisted News Detection
☆ Nova: An Iterative Planning and Search Approach to Enhance Novelty and Diversity of LLM Generated Ideas
Scientific innovation is pivotal for humanity, and harnessing large language models (LLMs) to generate research ideas could transform discovery. However, existing LLMs often produce simplistic and repetitive suggestions due to their limited ability in acquiring external knowledge for innovation. To address this problem, we introduce an enhanced planning and search methodology designed to boost the creative potential of LLM-based systems. Our approach involves an iterative process to purposely plan the retrieval of external knowledge, progressively enriching the idea generation with broader and deeper insights. Validation through automated and human assessments indicates that our framework substantially elevates the quality of generated ideas, particularly in novelty and diversity. The number of unique novel ideas produced by our framework is 3.4 times higher than without it. Moreover, our method outperforms the current state-of-the-art, generating at least 2.5 times more top-rated ideas based on 170 seed papers in a Swiss Tournament evaluation.
☆ Synthesizing Post-Training Data for LLMs through Multi-Agent Simulation
Post-training is essential for enabling large language models (LLMs) to follow human instructions. Inspired by the recent success of using LLMs to simulate human society, we leverage multi-agent simulation to automatically generate diverse text-based scenarios, capturing a wide range of real-world human needs. We propose MATRIX, a multi-agent simulator that creates realistic and scalable scenarios. Leveraging these outputs, we introduce a novel scenario-driven instruction generator MATRIX-Gen for controllable and highly realistic data synthesis. Extensive experiments demonstrate that our framework effectively generates both general and domain-specific data. Notably, on AlpacaEval 2 and Arena-Hard benchmarks, Llama-3-8B-Base, post-trained on datasets synthesized by MATRIX-Gen with just 20K instruction-response pairs, outperforms Meta's Llama-3-8B-Instruct model, which was trained on over 10M pairs; see our project at https://github.com/ShuoTang123/MATRIX-Gen.
☆ Addressing Blind Guessing: Calibration of Selection Bias in Multiple-Choice Question Answering by Video Language Models
Evaluating Video Language Models (VLMs) is a challenging task. Due to its transparency, Multiple-Choice Question Answering (MCQA) is widely used to measure the performance of these models through accuracy. However, existing MCQA benchmarks fail to capture the full reasoning capabilities of VLMs due to selection bias, when models disproportionately favor certain answer options based on positional patterns observed during training. In this work, we conduct a comprehensive empirical analysis of several VLM architectures across major datasets designed to assess complex video-focused reasoning. We identify where the bias is most pronounced and demonstrate to what extent model responses reflect genuine understanding of video content and related questions, as opposed to reliance on arbitrary patterns or superficial cues, such as answer position. By decomposing the MCQA task and adapting fairness bias metrics to VLMs, we introduce a post-processing calibration technique BOLD to balance this bias. Our results show that reducing selection bias improves not only debiasing metrics but also overall model performance, including Accuracy and F1 Mean score. Our method, by suppressing "blind guessing", offers a more cost- and time-effective approach to mitigating selection bias compared to existing techniques. This study represents the first focused investigation of selection bias in video-to-text LLM-powered models.
☆ A Novel Method to Metigate Demographic and Expert Bias in ICD Coding with Causal Inference
ICD(International Classification of Diseases) coding involves assigning ICD codes to patients visit based on their medical notes. Considering ICD coding as a multi-label text classification task, researchers have developed sophisticated methods. Despite progress, these models often suffer from label imbalance and may develop spurious correlations with demographic factors. Additionally, while human coders assign ICD codes, the inclusion of irrelevant information from unrelated experts introduces biases. To combat these issues, we propose a novel method to mitigate Demographic and Expert biases in ICD coding through Causal Inference (DECI). We provide a novel causality-based interpretation in ICD Coding that models make predictions by three distinct pathways. And based counterfactual reasoning, DECI mitigate demographic and expert biases. Experimental results show that DECI outperforms state-of-the-art models, offering a significant advancement in accurate and unbiased ICD coding.
☆ Towards Robust Knowledge Representations in Multilingual LLMs for Equivalence and Inheritance based Consistent Reasoning
Reasoning and linguistic skills form the cornerstone of human intelligence, facilitating problem-solving and decision-making. Recent advances in Large Language Models (LLMs) have led to impressive linguistic capabilities and emergent reasoning behaviors, fueling widespread adoption across application domains. However, LLMs still struggle with complex reasoning tasks, highlighting their systemic limitations. In this work, we focus on evaluating whether LLMs have the requisite representations to reason using two foundational relationships: "equivalence" and "inheritance". We introduce novel tasks and benchmarks spanning six languages and observe that current SOTA LLMs often produce conflicting answers to the same questions across languages in 17.3-57.5% of cases and violate inheritance constraints in up to 37.2% cases. To enhance consistency across languages, we propose novel "Compositional Representations" where tokens are represented as composition of equivalent tokens across languages, with resulting conflict reduction (up to -4.7%) indicating benefits of shared LLM representations.
☆ Unveiling Large Language Models Generated Texts: A Multi-Level Fine-Grained Detection Framework
Large language models (LLMs) have transformed human writing by enhancing grammar correction, content expansion, and stylistic refinement. However, their widespread use raises concerns about authorship, originality, and ethics, even potentially threatening scholarly integrity. Existing detection methods, which mainly rely on single-feature analysis and binary classification, often fail to effectively identify LLM-generated text in academic contexts. To address these challenges, we propose a novel Multi-level Fine-grained Detection (MFD) framework that detects LLM-generated text by integrating low-level structural, high-level semantic, and deep-level linguistic features, while conducting sentence-level evaluations of lexicon, grammar, and syntax for comprehensive analysis. To improve detection of subtle differences in LLM-generated text and enhance robustness against paraphrasing, we apply two mainstream evasion techniques to rewrite the text. These variations, along with original texts, are used to train a text encoder via contrastive learning, extracting high-level semantic features of sentence to boost detection generalization. Furthermore, we leverage advanced LLM to analyze the entire text and extract deep-level linguistic features, enhancing the model's ability to capture complex patterns and nuances while effectively incorporating contextual information. Extensive experiments on public datasets show that the MFD model outperforms existing methods, achieving an MAE of 0.1346 and an accuracy of 88.56%. Our research provides institutions and publishers with an effective mechanism to detect LLM-generated text, mitigating risks of compromised authorship. Educators and editors can use the model's predictions to refine verification and plagiarism prevention protocols, ensuring adherence to standards.
☆ Few-Shot Joint Multimodal Entity-Relation Extraction via Knowledge-Enhanced Cross-modal Prompt Model ACM MM 2024
Joint Multimodal Entity-Relation Extraction (JMERE) is a challenging task that aims to extract entities and their relations from text-image pairs in social media posts. Existing methods for JMERE require large amounts of labeled data. However, gathering and annotating fine-grained multimodal data for JMERE poses significant challenges. Initially, we construct diverse and comprehensive multimodal few-shot datasets fitted to the original data distribution. To address the insufficient information in the few-shot setting, we introduce the \textbf{K}nowledge-\textbf{E}nhanced \textbf{C}ross-modal \textbf{P}rompt \textbf{M}odel (KECPM) for JMERE. This method can effectively address the problem of insufficient information in the few-shot setting by guiding a large language model to generate supplementary background knowledge. Our proposed method comprises two stages: (1) a knowledge ingestion stage that dynamically formulates prompts based on semantic similarity guide ChatGPT generating relevant knowledge and employs self-reflection to refine the knowledge; (2) a knowledge-enhanced language model stage that merges the auxiliary knowledge with the original input and utilizes a transformer-based model to align with JMERE's required output format. We extensively evaluate our approach on a few-shot dataset derived from the JMERE dataset, demonstrating its superiority over strong baselines in terms of both micro and macro F$_1$ scores. Additionally, we present qualitative analyses and case studies to elucidate the effectiveness of our model.
comment: accepted by ACM MM 2024
☆ Paths-over-Graph: Knowledge Graph Enpowered Large Language Model Reasoning
Large Language Models (LLMs) have achieved impressive results in various tasks but struggle with hallucination problems and lack of relevant knowledge, especially in deep complex reasoning and knowledge-intensive tasks. Knowledge Graphs (KGs), which capture vast amounts of facts in a structured format, offer a reliable source of knowledge for reasoning. However, existing KG-based LLM reasoning methods face challenges like handling multi-hop reasoning, multi-entity questions, and effectively utilizing graph structures. To address these issues, we propose Paths-over-Graph (PoG), a novel method that enhances LLM reasoning by integrating knowledge reasoning paths from KGs, improving the interpretability and faithfulness of LLM outputs. PoG tackles multi-hop and multi-entity questions through a three-phase dynamic multi-hop path exploration, which combines the inherent knowledge of LLMs with factual knowledge from KGs. In order to improve the efficiency, PoG prunes irrelevant information from the graph exploration first and introduces efficient three-step pruning techniques that incorporate graph structures, LLM prompting, and a pre-trained language model (e.g., SBERT) to effectively narrow down the explored candidate paths. This ensures all reasoning paths contain highly relevant information captured from KGs, making the reasoning faithful and interpretable in problem-solving. PoG innovatively utilizes graph structure to prune the irrelevant noise and represents the first method to implement multi-entity deep path detection on KGs for LLM reasoning tasks. Comprehensive experiments on five benchmark KGQA datasets demonstrate PoG outperforms the state-of-the-art method ToG across GPT-3.5-Turbo and GPT-4, achieving an average accuracy improvement of 18.9%. Notably, PoG with GPT-3.5-Turbo surpasses ToG with GPT-4 by up to 23.9%.
☆ Montessori-Instruct: Generate Influential Training Data Tailored for Student Learning
Synthetic data has been widely used to train large language models, but their generative nature inevitably introduces noisy, non-informative, and misleading learning signals. In this paper, we propose Montessori-Instruct, a novel data synthesis framework that tailors the data synthesis ability of the teacher language model toward the student language model's learning process. Specifically, we utilize local data influence of synthetic training data points on students to characterize students' learning preferences. Then, we train the teacher model with Direct Preference Optimization (DPO) to generate synthetic data tailored toward student learning preferences. Experiments with Llama3-8B-Instruct (teacher) and Llama3-8B (student) on Alpaca Eval and MT-Bench demonstrate that Montessori-Instruct significantly outperforms standard synthesis methods by 18.35\% and 46.24\% relatively. Our method also beats data synthesized by a stronger teacher model, GPT-4o. Further analysis confirms the benefits of teacher's learning to generate more influential training data in the student's improved learning, the advantages of local data influence in accurately measuring student preferences, and the robustness of Montessori-Instruct across different student models. Our code and data are open-sourced at https://github.com/cxcscmu/Montessori-Instruct.
comment: Codes and data are open-sourced at https://github.com/cxcscmu/Montessori-Instruct
☆ MediTOD: An English Dialogue Dataset for Medical History Taking with Comprehensive Annotations EMNLP2024
Medical task-oriented dialogue systems can assist doctors by collecting patient medical history, aiding in diagnosis, or guiding treatment selection, thereby reducing doctor burnout and expanding access to medical services. However, doctor-patient dialogue datasets are not readily available, primarily due to privacy regulations. Moreover, existing datasets lack comprehensive annotations involving medical slots and their different attributes, such as symptoms and their onset, progression, and severity. These comprehensive annotations are crucial for accurate diagnosis. Finally, most existing datasets are non-English, limiting their utility for the larger research community. In response, we introduce MediTOD, a new dataset of doctor-patient dialogues in English for the medical history-taking task. Collaborating with doctors, we devise a questionnaire-based labeling scheme tailored to the medical domain. Then, medical professionals create the dataset with high-quality comprehensive annotations, capturing medical slots and their attributes. We establish benchmarks in supervised and few-shot settings on MediTOD for natural language understanding, policy learning, and natural language generation subtasks, evaluating models from both TOD and biomedical domains. We make MediTOD publicly available for future research.
comment: EMNLP2024 Camera Ready Version
☆ Rationale Behind Essay Scores: Enhancing S-LLM's Multi-Trait Essay Scoring with Rationale Generated by LLMs
Existing automated essay scoring (AES) has solely relied on essay text without using explanatory rationales for the scores, thereby forgoing an opportunity to capture the specific aspects evaluated by rubric indicators in a fine-grained manner. This paper introduces Rationale-based Multiple Trait Scoring (RMTS), a novel approach for multi-trait essay scoring that integrates prompt-engineering-based large language models (LLMs) with a fine-tuning-based essay scoring model using a smaller large language model (S-LLM). RMTS uses an LLM-based trait-wise rationale generation system where a separate LLM agent generates trait-specific rationales based on rubric guidelines, which the scoring model uses to accurately predict multi-trait scores. Extensive experiments on benchmark datasets, including ASAP, ASAP++, and Feedback Prize, show that RMTS significantly outperforms state-of-the-art models and vanilla S-LLMs in trait-specific scoring. By assisting quantitative assessment with fine-grained qualitative rationales, RMTS enhances the trait-wise reliability, providing partial explanations about essays.
☆ E3D-GPT: Enhanced 3D Visual Foundation for Medical Vision-Language Model
The development of 3D medical vision-language models holds significant potential for disease diagnosis and patient treatment. However, compared to 2D medical images, 3D medical images, such as CT scans, face challenges related to limited training data and high dimension, which severely restrict the progress of 3D medical vision-language models. To address these issues, we collect a large amount of unlabeled 3D CT data and utilize self-supervised learning to construct a 3D visual foundation model for extracting 3D visual features. Then, we apply 3D spatial convolutions to aggregate and project high-level image features, reducing computational complexity while preserving spatial information. We also construct two instruction-tuning datasets based on BIMCV-R and CT-RATE to fine-tune the 3D vision-language model. Our model demonstrates superior performance compared to existing methods in report generation, visual question answering, and disease diagnosis. Code and data will be made publicly available soon.
☆ Supervised Chain of Thought
Large Language Models (LLMs) have revolutionized natural language processing and hold immense potential for advancing Artificial Intelligence. However, the core architecture of most mainstream LLMs -- the Transformer -- has inherent limitations in computational depth, rendering them theoretically incapable of solving many reasoning tasks that demand increasingly deep computations. Chain of Thought (CoT) prompting has emerged as a technique to address these architectural limitations, as evidenced by several theoretical studies. It offers a promising approach to solving complex reasoning tasks that were previously beyond the capabilities of these models. Despite its successes, CoT and its variants (such as Tree of Thought, Graph of Thought, etc.) rely on a "one-prompt-for-all" approach, using a single prompt structure (e.g., "think step by step") for a wide range of tasks -- from counting and sorting to solving mathematical and algorithmic problems. This approach poses significant challenges for models to generate the correct reasoning steps, as the model must navigate through a vast prompt template space to find the appropriate template for each task. In this work, we build upon previous theoretical analyses of CoT to demonstrate how the one-prompt-for-all approach can negatively affect the computability of LLMs. We partition the solution search space into two: the prompt space and the answer space. Our findings show that task-specific supervision is essential for navigating the prompt space accurately and achieving optimal performance. Through experiments with state-of-the-art LLMs, we reveal a gap in reasoning performance when supervision is applied versus when it is not.
☆ Speciesism in Natural Language Processing Research
Natural Language Processing (NLP) research on AI Safety and social bias in AI has focused on safety for humans and social bias against human minorities. However, some AI ethicists have argued that the moral significance of nonhuman animals has been ignored in AI research. Therefore, the purpose of this study is to investigate whether there is speciesism, i.e., discrimination against nonhuman animals, in NLP research. First, we explain why nonhuman animals are relevant in NLP research. Next, we survey the findings of existing research on speciesism in NLP researchers, data, and models and further investigate this problem in this study. The findings of this study suggest that speciesism exists within researchers, data, and models, respectively. Specifically, our survey and experiments show that (a) among NLP researchers, even those who study social bias in AI, do not recognize speciesism or speciesist bias; (b) among NLP data, speciesist bias is inherent in the data annotated in the datasets used to evaluate NLP models; (c) OpenAI GPTs, recent NLP models, exhibit speciesist bias by default. Finally, we discuss how we can reduce speciesism in NLP research.
comment: This article is a preprint and has not been peer-reviewed. The postprint has been accepted for publication in AI and Ethics. Please cite the final version of the article once it is published
☆ MetaAlign: Align Large Language Models with Diverse Preferences during Inference Time
Large Language Models (LLMs) acquire extensive knowledge and remarkable abilities from extensive text corpora, making them powerful tools for various applications. To make LLMs more usable, aligning them with human preferences is essential. Existing alignment techniques, such as Reinforcement Learning from Human Feedback (RLHF) and Direct Preference Optimization (DPO), typically embed predefined preferences directly within the model's parameters. These methods, however, often result in a static alignment that can not account for the diversity of human preferences in practical applications. In response to this challenge, we propose an effective method, \textbf{MetaAlign}, which aims to help LLMs dynamically align with various explicit or implicit preferences specified at inference time. Experimental results show that LLMs optimized on our meticulously constructed MetaAlign Dataset can effectively align with any preferences specified at the inference stage, validating the feasibility of MetaAlign. We hope that our work can provide some insights into the alignment of language models.
comment: 19 pages, 6 figures
☆ LabSafety Bench: Benchmarking LLMs on Safety Issues in Scientific Labs
Laboratory accidents pose significant risks to human life and property, underscoring the importance of robust safety protocols. Despite advancements in safety training, laboratory personnel may still unknowingly engage in unsafe practices. With the increasing reliance on large language models (LLMs) for guidance in various fields, including laboratory settings, there is a growing concern about their reliability in critical safety-related decision-making. Unlike trained human researchers, LLMs lack formal lab safety education, raising questions about their ability to provide safe and accurate guidance. Existing research on LLM trustworthiness primarily focuses on issues such as ethical compliance, truthfulness, and fairness but fails to fully cover safety-critical real-world applications, like lab safety. To address this gap, we propose the Laboratory Safety Benchmark (LabSafety Bench), a comprehensive evaluation framework based on a new taxonomy aligned with Occupational Safety and Health Administration (OSHA) protocols. This benchmark includes 765 multiple-choice questions verified by human experts, assessing LLMs and vision language models (VLMs) performance in lab safety contexts. Our evaluations demonstrate that while GPT-4o outperforms human participants, it is still prone to critical errors, highlighting the risks of relying on LLMs in safety-critical environments. Our findings emphasize the need for specialized benchmarks to accurately assess the trustworthiness of LLMs in real-world safety applications.
comment: 50 pages, 19 figures
☆ XForecast: Evaluating Natural Language Explanations for Time Series Forecasting
Time series forecasting aids decision-making, especially for stakeholders who rely on accurate predictions, making it very important to understand and explain these models to ensure informed decisions. Traditional explainable AI (XAI) methods, which underline feature or temporal importance, often require expert knowledge. In contrast, natural language explanations (NLEs) are more accessible to laypeople. However, evaluating forecast NLEs is difficult due to the complex causal relationships in time series data. To address this, we introduce two new performance metrics based on simulatability, assessing how well a human surrogate can predict model forecasts using the explanations. Experiments show these metrics differentiate good from poor explanations and align with human judgments. Utilizing these metrics, we further evaluate the ability of state-of-the-art large language models (LLMs) to generate explanations for time series data, finding that numerical reasoning, rather than model size, is the main factor influencing explanation quality.
☆ MultiChartQA: Benchmarking Vision-Language Models on Multi-Chart Problems
Multimodal Large Language Models (MLLMs) have demonstrated impressive abilities across various tasks, including visual question answering and chart comprehension, yet existing benchmarks for chart-related tasks fall short in capturing the complexity of real-world multi-chart scenarios. Current benchmarks primarily focus on single-chart tasks, neglecting the multi-hop reasoning required to extract and integrate information from multiple charts, which is essential in practical applications. To fill this gap, we introduce MultiChartQA, a benchmark that evaluates MLLMs' capabilities in four key areas: direct question answering, parallel question answering, comparative reasoning, and sequential reasoning. Our evaluation of a wide range of MLLMs reveals significant performance gaps compared to humans. These results highlight the challenges in multi-chart comprehension and the potential of MultiChartQA to drive advancements in this field. Our code and data are available at https://github.com/Zivenzhu/Multi-chart-QA
comment: 18 pages, 9 figures
☆ LLM The Genius Paradox: A Linguistic and Math Expert's Struggle with Simple Word-based Counting Problems
Interestingly, LLMs yet struggle with some basic tasks that humans find trivial to handle, e.g., counting the number of character r's in the word "strawberry". There are several popular conjectures (e.g., tokenization, architecture and training data) regarding the reason for deficiency of LLMs in simple word-based counting problems, sharing the similar belief that such failure stems from model pretraining hence probably inevitable during deployment. In this paper, we carefully design multiple evaluation settings to investigate validity of prevalent conjectures. Meanwhile, we measure transferability of advanced mathematical and coding reasoning capabilities from specialized LLMs to simple counting tasks. Although specialized LLMs suffer from counting problems as well, we find conjectures about inherent deficiency of LLMs invalid and further seek opportunities to elicit knowledge and capabilities from LLMs that are beneficial to counting tasks. Compared with strategies such as finetuning and in-context learning that are commonly adopted to enhance performance on new or challenging tasks, we show that engaging reasoning is the most robust and efficient way to help LLMs better perceive tasks with more accurate responses. We hope our conjecture validation design could provide insights into the study of future critical failure modes of LLMs. Based on challenges in transferring advanced capabilities to much simpler tasks, we call for more attention to model capability acquisition and evaluation. We also highlight the importance of cultivating consciousness of "reasoning before responding" during model pretraining.
☆ Automated Genre-Aware Article Scoring and Feedback Using Large Language Models
This paper focuses on the development of an advanced intelligent article scoring system that not only assesses the overall quality of written work but also offers detailed feature-based scoring tailored to various article genres. By integrating the pre-trained BERT model with the large language model Chat-GPT, the system gains a deep understanding of both the content and structure of the text, enabling it to provide a thorough evaluation along with targeted suggestions for improvement. Experimental results demonstrate that this system outperforms traditional scoring methods across multiple public datasets, particularly in feature-based assessments, offering a more accurate reflection of the quality of different article types. Moreover, the system generates personalized feedback to assist users in enhancing their writing skills, underscoring the potential and practical value of automated scoring technologies in educational contexts.
☆ Beyond Autoregression: Discrete Diffusion for Complex Reasoning and Planning
Autoregressive language models, despite their impressive capabilities, struggle with complex reasoning and long-term planning tasks. We introduce discrete diffusion models as a novel solution to these challenges. Through the lens of subgoal imbalance, we demonstrate how diffusion models effectively learn difficult subgoals that elude autoregressive approaches. We propose Multi-granularity Diffusion Modeling (MDM), which prioritizes subgoals based on difficulty during learning. On complex tasks like Countdown, Sudoku, and Boolean Satisfiability Problems, MDM significantly outperforms autoregressive models without using search techniques. For instance, MDM achieves 91.5\% and 100\% accuracy on Countdown and Sudoku, respectively, compared to 45.8\% and 20.7\% for autoregressive models. Our work highlights the potential of diffusion-based approaches in advancing AI capabilities for sophisticated language understanding and problem-solving tasks.
☆ Towards Faithful Natural Language Explanations: A Study Using Activation Patching in Large Language Models
Large Language Models (LLMs) are capable of generating persuasive Natural Language Explanations (NLEs) to justify their answers. However, the faithfulness of these explanations should not be readily trusted at face value. Recent studies have proposed various methods to measure the faithfulness of NLEs, typically by inserting perturbations at the explanation or feature level. We argue that these approaches are neither comprehensive nor correctly designed according to the established definition of faithfulness. Moreover, we highlight the risks of grounding faithfulness findings on out-of-distribution samples. In this work, we leverage a causal mediation technique called activation patching, to measure the faithfulness of an explanation towards supporting the explained answer. Our proposed metric, Causal Faithfulness quantifies the consistency of causal attributions between explanations and the corresponding model outputs as the indicator of faithfulness. We experimented across models varying from 2B to 27B parameters and found that models that underwent alignment tuning tend to produce more faithful and plausible explanations. We find that Causal Faithfulness is a promising improvement over existing faithfulness tests by taking into account the model's internal computations and avoiding out of distribution concerns that could otherwise undermine the validity of faithfulness assessments. We release the code in \url{https://github.com/wj210/Causal-Faithfulness}
comment: Under review
☆ SRAP-Agent: Simulating and Optimizing Scarce Resource Allocation Policy with LLM-based Agent
Public scarce resource allocation plays a crucial role in economics as it directly influences the efficiency and equity in society. Traditional studies including theoretical model-based, empirical study-based and simulation-based methods encounter limitations due to the idealized assumption of complete information and individual rationality, as well as constraints posed by limited available data. In this work, we propose an innovative framework, SRAP-Agent (Simulating and Optimizing Scarce Resource Allocation Policy with LLM-based Agent), which integrates Large Language Models (LLMs) into economic simulations, aiming to bridge the gap between theoretical models and real-world dynamics. Using public housing allocation scenarios as a case study, we conduct extensive policy simulation experiments to verify the feasibility and effectiveness of the SRAP-Agent and employ the Policy Optimization Algorithm with certain optimization objectives. The source code can be found in https://github.com/jijiarui-cather/SRAPAgent_Framework
☆ Utilizing Large Language Models for Event Deconstruction to Enhance Multimodal Aspect-Based Sentiment Analysis
With the rapid development of the internet, the richness of User-Generated Contentcontinues to increase, making Multimodal Aspect-Based Sentiment Analysis (MABSA) a research hotspot. Existing studies have achieved certain results in MABSA, but they have not effectively addressed the analytical challenges in scenarios where multiple entities and sentiments coexist. This paper innovatively introduces Large Language Models (LLMs) for event decomposition and proposes a reinforcement learning framework for Multimodal Aspect-based Sentiment Analysis (MABSA-RL) framework. This framework decomposes the original text into a set of events using LLMs, reducing the complexity of analysis, introducing reinforcement learning to optimize model parameters. Experimental results show that MABSA-RL outperforms existing advanced methods on two benchmark datasets. This paper provides a new research perspective and method for multimodal aspect-level sentiment analysis.
☆ Fine-Grained Verifiers: Preference Modeling as Next-token Prediction in Vision-Language Alignment
The recent advancements in large language models (LLMs) and pre-trained vision models have accelerated the development of vision-language large models (VLLMs), enhancing the interaction between visual and linguistic modalities. Despite their notable success across various domains, VLLMs face challenges in modality alignment, which can lead to issues like hallucinations and unsafe content generation. Current alignment techniques often rely on coarse feedback and external datasets, limiting scalability and performance. In this paper, we propose FiSAO (Fine-Grained Self-Alignment Optimization), a novel self-alignment method that utilizes the model's own visual encoder as a fine-grained verifier to improve vision-language alignment without the need for additional data. By leveraging token-level feedback from the vision encoder, FiSAO significantly improves vision-language alignment, even surpassing traditional preference tuning methods that require additional data. Through both theoretical analysis and experimental validation, we demonstrate that FiSAO effectively addresses the misalignment problem in VLLMs, marking the first instance of token-level rewards being applied to such models.
comment: 23 pages
☆ CAPE: A Chinese Dataset for Appraisal-based Emotional Generation using Large Language Models
Generating emotionally appropriate responses in conversations with large language models presents a significant challenge due to the complexities of human emotions and cognitive processes, which remain largely underexplored in their critical role in social interactions. In this study, we introduce a two-stage automatic data generation framework to create CAPE, a Chinese dataset named Cognitive Appraisal theory-based Emotional corpus. This corpus facilitates the generation of dialogues with contextually appropriate emotional responses by accounting for diverse personal and situational factors. We propose two tasks utilizing this dataset: emotion prediction and next utterance prediction. Both automated and human evaluations demonstrate that agents trained on our dataset can deliver responses that are more aligned with human emotional expressions. Our study shows the potential for advancing emotional expression in conversational agents, paving the way for more nuanced and meaningful human-computer interactions.
☆ A Lightweight Multi Aspect Controlled Text Generation Solution For Large Language Models
Large language models (LLMs) show remarkable abilities with instruction tuning. However, they fail to achieve ideal tasks when lacking high-quality instruction tuning data on target tasks. Multi-Aspect Controllable Text Generation (MCTG) is a representative task for this dilemma, where aspect datasets are usually biased and correlated. Existing work exploits additional model structures and strategies for solutions, limiting adaptability to LLMs. To activate MCTG ability of LLMs, we propose a lightweight MCTG pipeline based on data augmentation. We analyze bias and correlations in traditional datasets, and address these concerns with augmented control attributes and sentences. Augmented datasets are feasible for instruction tuning. In our experiments, LLMs perform better in MCTG after data augmentation, with a 20% accuracy rise and less aspect correlations.
☆ Coherence-Driven Multimodal Safety Dialogue with Active Learning for Embodied Agents
When assisting people in daily tasks, robots need to accurately interpret visual cues and respond effectively in diverse safety-critical situations, such as sharp objects on the floor. In this context, we present M-CoDAL, a multimodal-dialogue system specifically designed for embodied agents to better understand and communicate in safety-critical situations. The system leverages discourse coherence relations to enhance its contextual understanding and communication abilities. To train this system, we introduce a novel clustering-based active learning mechanism that utilizes an external Large Language Model (LLM) to identify informative instances. Our approach is evaluated using a newly created multimodal dataset comprising 1K safety violations extracted from 2K Reddit images. These violations are annotated using a Large Multimodal Model (LMM) and verified by human annotators. Results with this dataset demonstrate that our approach improves resolution of safety situations, user sentiment, as well as safety of the conversation. Next, we deploy our dialogue system on a Hello Robot Stretch robot and conduct a within-subject user study with real-world participants. In the study, participants role-play two safety scenarios with different levels of severity with the robot and receive interventions from our model and a baseline system powered by OpenAI's ChatGPT. The study results corroborate and extend the findings from automated evaluation, showing that our proposed system is more persuasive and competent in a real-world embodied agent setting.
☆ ViConsFormer: Constituting Meaningful Phrases of Scene Texts using Transformer-based Method in Vietnamese Text-based Visual Question Answering
Text-based VQA is a challenging task that requires machines to use scene texts in given images to yield the most appropriate answer for the given question. The main challenge of text-based VQA is exploiting the meaning and information from scene texts. Recent studies tackled this challenge by considering the spatial information of scene texts in images via embedding 2D coordinates of their bounding boxes. In this study, we follow the definition of meaning from linguistics to introduce a novel method that effectively exploits the information from scene texts written in Vietnamese. Experimental results show that our proposed method obtains state-of-the-art results on two large-scale Vietnamese Text-based VQA datasets. The implementation can be found at this link.
♻ ☆ Locate-then-edit for Multi-hop Factual Recall under Knowledge Editing
The locate-then-edit paradigm has shown significant promise for knowledge editing (KE) in Large Language Models (LLMs). While previous methods perform well on single-hop fact recall tasks, they consistently struggle with multi-hop factual recall tasks involving newly edited knowledge. In this paper, leveraging tools in mechanistic interpretability, we first identify that in multi-hop tasks, LLMs tend to retrieve implicit subject knowledge from deeper MLP layers, unlike single-hop tasks, which rely on earlier layers. This distinction explains the poor performance of current methods in multi-hop queries, as they primarily focus on editing shallow layers, leaving deeper layers unchanged. To address this, we propose IFMET, a novel locate-then-edit KE approach designed to edit both shallow and deep MLP layers. IFMET employs multi-hop editing prompts and supplementary sets to locate and modify knowledge across different reasoning stages. Experimental results demonstrate that IFMET significantly improves performance on multi-hop factual recall tasks, effectively overcoming the limitations of previous locate-then-edit methods.
comment: 21 pages
♻ ☆ System 2 thinking in OpenAI's o1-preview model: Near-perfect performance on a mathematics exam
The processes underlying human cognition are often divided into System 1, which involves fast, intuitive thinking, and System 2, which involves slow, deliberate reasoning. Previously, large language models were criticized for lacking the deeper, more analytical capabilities of System 2. In September 2024, OpenAI introduced the o1 model series, designed to handle System 2-like reasoning. While OpenAI's benchmarks are promising, independent validation is still needed. In this study, we tested the o1-preview model twice on the Dutch 'Mathematics B' final exam. It scored a near-perfect 76 and 74 out of 76 points. For context, only 24 out of 16,414 students in the Netherlands achieved a perfect score. By comparison, the GPT-4o model scored 66 and 62 out of 76, well above the Dutch average of 40.63 points. Neither model had access to the exam figures. Since there was a risk of model contamination (i.e., the knowledge cutoff of o1-preview and GPT-4o was after the exam was published online), we repeated the procedure with a new Mathematics B exam that was published after the cutoff date. The results again indicated that o1-preview performed strongly (97.8th percentile), which suggests that contamination was not a factor. We also show that there is some variability in the output of o1-preview, which means that sometimes there is 'luck' (the answer is correct) or 'bad luck' (the output has diverged into something that is incorrect). We demonstrate that a self-consistency approach, where repeated prompts are given and the most common answer is selected, is a useful strategy for identifying the correct answer. It is concluded that while OpenAI's new model series holds great potential, certain risks must be considered.
♻ ☆ Liger Kernel: Efficient Triton Kernels for LLM Training
Training Large Language Models (LLMs) efficiently at scale presents a formidable challenge, driven by their ever-increasing computational demands and the need for enhanced performance. In this work, we introduce Liger-Kernel, an open-sourced set of Triton kernels developed specifically for LLM training. With kernel optimization techniques like kernel operation fusing and input chunking, our kernels achieve on average a 20% increase in training throughput and a 60% reduction in GPU memory usage for popular LLMs compared to HuggingFace implementations. In addition, Liger-Kernel is designed with modularity, accessibility, and adaptability in mind, catering to both casual and expert users. Comprehensive benchmarks and integration tests are built in to ensure compatibility, performance, correctness, and convergence across diverse computing environments and model architectures. The source code is available under a permissive license at: github.com/linkedin/Liger-Kernel.
comment: 17 pages, 12 figures
♻ ☆ Contextual Document Embeddings
Dense document embeddings are central to neural retrieval. The dominant paradigm is to train and construct embeddings by running encoders directly on individual documents. In this work, we argue that these embeddings, while effective, are implicitly out-of-context for targeted use cases of retrieval, and that a contextualized document embedding should take into account both the document and neighboring documents in context - analogous to contextualized word embeddings. We propose two complementary methods for contextualized document embeddings: first, an alternative contrastive learning objective that explicitly incorporates the document neighbors into the intra-batch contextual loss; second, a new contextual architecture that explicitly encodes neighbor document information into the encoded representation. Results show that both methods achieve better performance than biencoders in several settings, with differences especially pronounced out-of-domain. We achieve state-of-the-art results on the MTEB benchmark with no hard negative mining, score distillation, dataset-specific instructions, intra-GPU example-sharing, or extremely large batch sizes. Our method can be applied to improve performance on any contrastive learning dataset and any biencoder.
♻ ☆ Learning Linear Attention in Polynomial Time
Previous research has explored the computational expressivity of Transformer models in simulating Boolean circuits or Turing machines. However, the learnability of these simulators from observational data has remained an open question. Our study addresses this gap by providing the first polynomial-time learnability results (specifically strong, agnostic PAC learning) for single-layer Transformers with linear attention. We show that linear attention may be viewed as a linear predictor in a suitably defined RKHS. As a consequence, the problem of learning any linear transformer may be converted into the problem of learning an ordinary linear predictor in an expanded feature space, and any such predictor may be converted back into a multiheaded linear transformer. Moving to generalization, we show how to efficiently identify training datasets for which every empirical risk minimizer is equivalent (up to trivial symmetries) to the linear Transformer that generated the data, thereby guaranteeing the learned model will correctly generalize across all inputs. Finally, we provide examples of computations expressible via linear attention and therefore polynomial-time learnable, including associative memories, finite automata, and a class of Universal Turing Machine (UTMs) with polynomially bounded computation histories. We empirically validate our theoretical findings on three tasks: learning random linear attention networks, key--value associations, and learning to execute finite automata. Our findings bridge a critical gap between theoretical expressivity and learnability of Transformers, and show that flexible and general models of computation are efficiently learnable.
♻ ☆ One size doesn't fit all: Predicting the Number of Examples for In-Context Learning
In-context learning (ICL) refers to the process of adding a small number of localized examples (ones that are semantically similar to the input) from a training set of labelled data to an LLM's prompt with an objective to effectively control the generative process seeking to improve the downstream task performance. Existing ICL approaches use an identical number of examples (a pre-configured hyper-parameter) for each data instance. Our work alleviates the limitations of this 'one fits all' approach by dynamically predicting the number of examples for each data instance to be used in few-shot inference with LLMs. In particular, we employ a multi-label classifier, the parameters of which are fitted using a training set, where the label for each instance in the training set indicates if using a specific value of k (number of most similar examples from 0 up to a maximum value) leads to correct k-shot downstream predictions. Our experiments on a number of text classification benchmarks show that AICL substantially outperforms standard ICL by up to 17%.
♻ ☆ Movie101v2: Improved Movie Narration Benchmark
Automatic movie narration aims to generate video-aligned plot descriptions to assist visually impaired audiences. Unlike standard video captioning, it involves not only describing key visual details but also inferring plots that unfold across multiple movie shots, presenting distinct and complex challenges. To advance this field, we introduce Movie101v2, a large-scale, bilingual dataset with enhanced data quality specifically designed for movie narration. Revisiting the task, we propose breaking down the ultimate goal of automatic movie narration into three progressive stages, offering a clear roadmap with corresponding evaluation metrics. Based on our new benchmark, we baseline a range of large vision-language models, including GPT-4V, and conduct an in-depth analysis of the challenges in narration generation. Our findings highlight that achieving applicable movie narration generation is a fascinating goal that requires significant research.
♻ ☆ MCQG-SRefine: Multiple Choice Question Generation and Evaluation with Iterative Self-Critique, Correction, and Comparison Feedback
Automatic question generation (QG) is essential for AI and NLP, particularly in intelligent tutoring, dialogue systems, and fact verification. Generating multiple-choice questions (MCQG) for professional exams, like the United States Medical Licensing Examination (USMLE), is particularly challenging, requiring domain expertise and complex multi-hop reasoning for high-quality questions. However, current large language models (LLMs) like GPT-4 struggle with professional MCQG due to outdated knowledge, hallucination issues, and prompt sensitivity, resulting in unsatisfactory quality and difficulty. To address these challenges, we propose MCQG-SRefine, an LLM self-refine-based (Critique and Correction) framework for converting medical cases into high-quality USMLE-style questions. By integrating expert-driven prompt engineering with iterative self-critique and self-correction feedback, MCQG-SRefine significantly enhances human expert satisfaction regarding both the quality and difficulty of the questions. Furthermore, we introduce an LLM-as-Judge-based automatic metric to replace the complex and costly expert evaluation process, ensuring reliable and expert-aligned assessments.
comment: Equal contribution for the first two authors
♻ ☆ Advocating Character Error Rate for Multilingual ASR Evaluation
Automatic speech recognition (ASR) systems have traditionally been evaluated using English datasets, with the word error rate (WER) serving as the predominant metric. WER's simplicity and ease of interpretation have contributed to its widespread adoption, particularly for English. However, as ASR systems expand to multilingual contexts, WER fails in various ways, particularly with morphologically complex languages or those without clear word boundaries. Our work documents the limitations of WER as an evaluation metric and advocates for the character error rate (CER) as the primary metric in multilingual ASR evaluation. We show that CER avoids many of the challenges WER faces and exhibits greater consistency across writing systems. We support our proposition by conducting human evaluations of ASR transcriptions in three languages: Malayalam, English, and Arabic, which exhibit distinct morphological characteristics. We show that CER correlates more closely with human judgments than WER, even for English. To facilitate further research, we release our human evaluation dataset for future benchmarking of ASR metrics. Our findings suggest that CER should be prioritized, or at least supplemented, in multilingual ASR evaluations to account for the varying linguistic characteristics of different languages.
comment: 4 pages
♻ ☆ English offensive text detection using CNN based Bi-GRU model
Over the years, the number of users of social media has increased drastically. People frequently share their thoughts through social platforms, and this leads to an increase in hate content. In this virtual community, individuals share their views, express their feelings, and post photos, videos, blogs, and more. Social networking sites like Facebook and Twitter provide platforms to share vast amounts of content with a single click. However, these platforms do not impose restrictions on the uploaded content, which may include abusive language and explicit images unsuitable for social media. To resolve this issue, a new idea must be implemented to divide the inappropriate content. Numerous studies have been done to automate the process. In this paper, we propose a new Bi-GRU-CNN model to classify whether the text is offensive or not. The combination of the Bi-GRU and CNN models outperforms the existing model.
comment: 5 pages and 6 figures
♻ ☆ Improving Reward Models with Synthetic Critiques
Reward models (RMs) play a critical role in aligning language models through the process of reinforcement learning from human feedback. RMs are trained to predict a score reflecting human preference, which requires significant time and cost for human annotation. Additionally, RMs tend to quickly overfit on superficial features in the training set, hindering their generalization performance on unseen distributions. We propose a novel approach using synthetic natural language critiques generated by large language models to provide additional feedback, evaluating aspects such as instruction following, correctness, and style. This offers richer signals and more robust features for RMs to assess and score on. We demonstrate that high-quality critiques improve the performance and data efficiency of RMs initialized from different pretrained models, reducing the reliance on costly human annotations. Furthermore, incorporating critiques improves both the interpretability and robustness of RM training.
♻ ☆ With Ears to See and Eyes to Hear: Sound Symbolism Experiments with Multimodal Large Language Models EMNLP 2024
Recently, Large Language Models (LLMs) and Vision Language Models (VLMs) have demonstrated aptitude as potential substitutes for human participants in experiments testing psycholinguistic phenomena. However, an understudied question is to what extent models that only have access to vision and text modalities are able to implicitly understand sound-based phenomena via abstract reasoning from orthography and imagery alone. To investigate this, we analyse the ability of VLMs and LLMs to demonstrate sound symbolism (i.e., to recognise a non-arbitrary link between sounds and concepts) as well as their ability to "hear" via the interplay of the language and vision modules of open and closed-source multimodal models. We perform multiple experiments, including replicating the classic Kiki-Bouba and Mil-Mal shape and magnitude symbolism tasks, and comparing human judgements of linguistic iconicity with that of LLMs. Our results show that VLMs demonstrate varying levels of agreement with human labels, and more task information may be required for VLMs versus their human counterparts for in silico experimentation. We additionally see through higher maximum agreement levels that Magnitude Symbolism is an easier pattern for VLMs to identify than Shape Symbolism, and that an understanding of linguistic iconicity is highly dependent on model size.
comment: Accepted to EMNLP 2024 (Camera Ready)
♻ ☆ Crossroads of Continents: Automated Artifact Extraction for Cultural Adaptation with Large Multimodal Models
We present a comprehensive three-phase study to examine (1) the cultural understanding of Large Multimodal Models (LMMs) by introducing DalleStreet, a large-scale dataset generated by DALL-E 3 and validated by humans, containing 9,935 images of 67 countries and 10 concept classes; (2) the underlying implicit and potentially stereotypical cultural associations with a cultural artifact extraction task; and (3) an approach to adapt cultural representation in an image based on extracted associations using a modular pipeline, CultureAdapt. We find disparities in cultural understanding at geographic sub-region levels with both open-source (LLaVA) and closed-source (GPT-4V) models on DalleStreet and other existing benchmarks, which we try to understand using over 18,000 artifacts that we identify in association to different countries. Our findings reveal a nuanced picture of the cultural competence of LMMs, highlighting the need to develop culture-aware systems. Dataset and code are available at https://github.com/iamshnoo/crossroads
comment: under review
♻ ☆ What's under the hood: Investigating Automatic Metrics on Meeting Summarization
Meeting summarization has become a critical task considering the increase in online interactions. While new techniques are introduced regularly, their evaluation uses metrics not designed to capture meeting-specific errors, undermining effective evaluation. This paper investigates what the frequently used automatic metrics capture and which errors they mask by correlating automatic metric scores with human evaluations across a broad error taxonomy. We commence with a comprehensive literature review on English meeting summarization to define key challenges like speaker dynamics and contextual turn-taking and error types such as missing information and linguistic inaccuracy, concepts previously loosely defined in the field. We examine the relationship between characteristic challenges and errors by using annotated transcripts and summaries from Transformer-based sequence-to-sequence and autoregressive models from the general summary QMSum dataset. Through experimental validation, we find that different model architectures respond variably to challenges in meeting transcripts, resulting in different pronounced links between challenges and errors. Current default-used metrics struggle to capture observable errors, showing weak to mid-correlations, while a third of the correlations show trends of error masking. Only a subset reacts accurately to specific errors, while most correlations show either unresponsiveness or failure to reflect the error's impact on summary quality.
♻ ☆ On Debiasing Text Embeddings Through Context Injection
Current advances in Natural Language Processing (NLP) have made it increasingly feasible to build applications leveraging textual data. Generally, the core of these applications rely on having a good semantic representation of text into vectors, via embedding models. However, it has been shown that these embeddings capture and perpetuate biases already present in text. While a few techniques have been proposed to debias embeddings, they do not take advantage of the recent advances in context understanding of modern embedding models. In this paper, we fill this gap by conducting a review of 19 embedding models by quantifying their biases and how well they respond to context injection as a mean of debiasing. We show that higher performing models are more prone to capturing biases, but are also better at incorporating context. Surprisingly, we find that while models can easily embed affirmative semantics, they fail at embedding neutral semantics. Finally, in a retrieval task, we show that biases in embeddings can lead to non-desirable outcomes. We use our new-found insights to design a simple algorithm for top $k$ retrieval, where $k$ is dynamically selected. We show that our algorithm is able to retrieve all relevant gendered and neutral chunks.
♻ ☆ Train & Constrain: Phonologically Informed Tongue-Twister Generation from Topics and Paraphrases
Previous work in phonologically and phonetically grounded language generation has mainly focused on domains such as puns and poetry. In this article, we present new work on the generation of English tongue twisters - a form of language that is required to be conditioned on a phoneme level to maximize sound overlap, while maintaining semantic consistency with an input topic or phrase and still being grammatically correct. We present TwisterLister, a pipeline for generating phonologically informed tongue twisters from large language models (LLMs) that we use to generate TwistList 2.0, the largest annotated dataset of tongue twisters to date, consisting of 17K+ examples from a combination of human and LLM authors. Our generation pipeline involves the use of a phonologically constrained vocabulary alongside LLM prompting to generate novel, non-derivative tongue twister examples. We additionally present the results of automatic and human evaluation of smaller models trained on our generated dataset to demonstrate the extent to which phonologically motivated language types can be generated without explicit injection of phonological knowledge. Additionally, we introduce a phoneme-aware constrained decoding module (PACD) that can be integrated into an autoregressive language model and demonstrate that this method generates good quality tongue twisters both with and without fine-tuning the underlying language model. We also design and implement a range of automatic metrics for the task of tongue twister generation that is phonologically motivated and captures the unique essence of tongue twisters, primarily based on phonemic edit distance (PED)
comment: Accepted Final Version to Computational Linguistics
♻ ☆ Error Span Annotation: A Balanced Approach for Human Evaluation of Machine Translation
High-quality Machine Translation (MT) evaluation relies heavily on human judgments. Comprehensive error classification methods, such as Multidimensional Quality Metrics (MQM), are expensive as they are time-consuming and can only be done by experts, whose availability may be limited especially for low-resource languages. On the other hand, just assigning overall scores, like Direct Assessment (DA), is simpler and faster and can be done by translators of any level, but is less reliable. In this paper, we introduce Error Span Annotation (ESA), a human evaluation protocol which combines the continuous rating of DA with the high-level error severity span marking of MQM. We validate ESA by comparing it to MQM and DA for 12 MT systems and one human reference translation (English to German) from WMT23. The results show that ESA offers faster and cheaper annotations than MQM at the same quality level, without the requirement of expensive MQM experts.
♻ ☆ BlackDAN: A Black-Box Multi-Objective Approach for Effective and Contextual Jailbreaking of Large Language Models
While large language models (LLMs) exhibit remarkable capabilities across various tasks, they encounter potential security risks such as jailbreak attacks, which exploit vulnerabilities to bypass security measures and generate harmful outputs. Existing jailbreak strategies mainly focus on maximizing attack success rate (ASR), frequently neglecting other critical factors, including the relevance of the jailbreak response to the query and the level of stealthiness. This narrow focus on single objectives can result in ineffective attacks that either lack contextual relevance or are easily recognizable. In this work, we introduce BlackDAN, an innovative black-box attack framework with multi-objective optimization, aiming to generate high-quality prompts that effectively facilitate jailbreaking while maintaining contextual relevance and minimizing detectability. BlackDAN leverages Multiobjective Evolutionary Algorithms (MOEAs), specifically the NSGA-II algorithm, to optimize jailbreaks across multiple objectives including ASR, stealthiness, and semantic relevance. By integrating mechanisms like mutation, crossover, and Pareto-dominance, BlackDAN provides a transparent and interpretable process for generating jailbreaks. Furthermore, the framework allows customization based on user preferences, enabling the selection of prompts that balance harmfulness, relevance, and other factors. Experimental results demonstrate that BlackDAN outperforms traditional single-objective methods, yielding higher success rates and improved robustness across various LLMs and multimodal LLMs, while ensuring jailbreak responses are both relevant and less detectable.
♻ ☆ Improving Retrieval in Sponsored Search by Leveraging Query Context Signals EMNLP 2024
Accurately retrieving relevant bid keywords for user queries is critical in Sponsored Search but remains challenging, particularly for short, ambiguous queries. Existing dense and generative retrieval models often fail to capture nuanced user intent in these cases. To address this, we propose an approach to enhance query understanding by augmenting queries with rich contextual signals derived from web search results and large language models, stored in an online cache. Specifically, we use web search titles and snippets to ground queries in real-world information and utilize GPT-4 to generate query rewrites and explanations that clarify user intent. These signals are efficiently integrated through a Fusion-in-Decoder based Unity architecture, enabling both dense and generative retrieval with serving costs on par with traditional context-free models. To address scenarios where context is unavailable in the cache, we introduce context glancing, a curriculum learning strategy that improves model robustness and performance even without contextual signals during inference. Extensive offline experiments demonstrate that our context-aware approach substantially outperforms context-free models. Furthermore, online A/B testing on a prominent search engine across 160+ countries shows significant improvements in user engagement and revenue.
comment: Accepted to EMNLP 2024 Industry Track. 10 pages, 10 tables, 1 figure
♻ ☆ Model Internals-based Answer Attribution for Trustworthy Retrieval-Augmented Generation EMNLP 2024
Ensuring the verifiability of model answers is a fundamental challenge for retrieval-augmented generation (RAG) in the question answering (QA) domain. Recently, self-citation prompting was proposed to make large language models (LLMs) generate citations to supporting documents along with their answers. However, self-citing LLMs often struggle to match the required format, refer to non-existent sources, and fail to faithfully reflect LLMs' context usage throughout the generation. In this work, we present MIRAGE --Model Internals-based RAG Explanations -- a plug-and-play approach using model internals for faithful answer attribution in RAG applications. MIRAGE detects context-sensitive answer tokens and pairs them with retrieved documents contributing to their prediction via saliency methods. We evaluate our proposed approach on a multilingual extractive QA dataset, finding high agreement with human answer attribution. On open-ended QA, MIRAGE achieves citation quality and efficiency comparable to self-citation while also allowing for a finer-grained control of attribution parameters. Our qualitative evaluation highlights the faithfulness of MIRAGE's attributions and underscores the promising application of model internals for RAG answer attribution.
comment: Accepted by EMNLP 2024 Main Conference. Code and data released at https://github.com/Betswish/MIRAGE
♻ ☆ A Novel Cartography-Based Curriculum Learning Method Applied on RoNLI: The First Romanian Natural Language Inference Corpus ACL 2024
Natural language inference (NLI), the task of recognizing the entailment relationship in sentence pairs, is an actively studied topic serving as a proxy for natural language understanding. Despite the relevance of the task in building conversational agents and improving text classification, machine translation and other NLP tasks, to the best of our knowledge, there is no publicly available NLI corpus for the Romanian language. To this end, we introduce the first Romanian NLI corpus (RoNLI) comprising 58K training sentence pairs, which are obtained via distant supervision, and 6K validation and test sentence pairs, which are manually annotated with the correct labels. We conduct experiments with multiple machine learning methods based on distant learning, ranging from shallow models based on word embeddings to transformer-based neural networks, to establish a set of competitive baselines. Furthermore, we improve on the best model by employing a new curriculum learning strategy based on data cartography. Our dataset and code to reproduce the baselines are available at https://github.com/Eduard6421/RONLI.
comment: Accepted at ACL 2024 (Main)
♻ ☆ Towards Lifelong Dialogue Agents via Relation-aware Memory Construction and Timeline-augmented Response Generation
To achieve lifelong human-agent interaction, dialogue agents need to constantly memorize perceived information and properly retrieve it for response generation (RG). While prior work focuses on getting rid of outdated memories to improve retrieval quality, we argue that such memories provide rich, important contextual cues for RG (e.g., changes in user behaviors) in long-term conversations. We present Theanine, a framework for LLM-based lifelong dialogue agents. Theanine discards memory removal and manages large-scale memories by linking them based on their temporal and cause-effect relation. Enabled by this linking structure, Theanine augments RG with memory timelines - series of memories representing the evolution or causality of relevant past events. Along with Theanine, we introduce TeaFarm, a counterfactual-driven evaluation scheme, addressing the limitation of G-Eval and human efforts in measuring memory-augmented dialogue agents. A supplementary video for Theanine and data for TeaFarm are at https://huggingface.co/spaces/ResearcherScholar/Theanine.
comment: Work in Progress
♻ ☆ Large Language Models, scientific knowledge and factuality: A framework to streamline human expert evaluation
The paper introduces a framework for the evaluation of the encoding of factual scientific knowledge, designed to streamline the manual evaluation process typically conducted by domain experts. Inferring over and extracting information from Large Language Models (LLMs) trained on a large corpus of scientific literature can potentially define a step change in biomedical discovery, reducing the barriers for accessing and integrating existing medical evidence. This work explores the potential of LLMs for dialoguing with biomedical background knowledge, using the context of antibiotic discovery. The framework involves of three evaluation steps, each assessing different aspects sequentially: fluency, prompt alignment, semantic coherence, factual knowledge, and specificity of the generated responses. By splitting these tasks between non-experts and experts, the framework reduces the effort required from the latter. The work provides a systematic assessment on the ability of eleven state-of-the-art models LLMs, including ChatGPT, GPT-4 and Llama 2, in two prompting-based tasks: chemical compound definition generation and chemical compound-fungus relation determination. Although recent models have improved in fluency, factual accuracy is still low and models are biased towards over-represented entities. The ability of LLMs to serve as biomedical knowledge bases is questioned, and the need for additional systematic evaluation frameworks is highlighted. While LLMs are currently not fit for purpose to be used as biomedical factual knowledge bases in a zero-shot setting, there is a promising emerging property in the direction of factuality as the models become domain specialised, scale-up in size and level of human feedback.
comment: Accepted at the Journal of Biomedical Informatics, Volume 158, October 2024, 104724
♻ ☆ Multi-LLM QA with Embodied Exploration
Large language models (LLMs) have grown in popularity due to their natural language interface and pre trained knowledge, leading to rapidly increasing success in question-answering (QA) tasks. More recently, multi-agent systems with LLM-based agents (Multi-LLM) have been utilized increasingly more for QA. In these scenarios, the models may each answer the question and reach a consensus or each model is specialized to answer different domain questions. However, most prior work dealing with Multi-LLM QA has focused on scenarios where the models are asked in a zero-shot manner or are given information sources to extract the answer. For question answering of an unknown environment, embodied exploration of the environment is first needed to answer the question. This skill is necessary for personalizing embodied AI to environments such as households. There is a lack of insight into whether a Multi-LLM system can handle question-answering based on observations from embodied exploration. In this work, we address this gap by investigating the use of Multi-Embodied LLM Explorers (MELE) for QA in an unknown environment. Multiple LLM-based agents independently explore and then answer queries about a household environment. We analyze different aggregation methods to generate a single, final answer for each query: debating, majority voting, and training a central answer module (CAM). Using CAM, we observe a $46\%$ higher accuracy compared against the other non-learning-based aggregation methods. We provide code and the query dataset for further research.
comment: 16 pages, 9 Figures, 5 Tables
♻ ☆ MolecularGPT: Open Large Language Model (LLM) for Few-Shot Molecular Property Prediction
Molecular property prediction (MPP) is a fundamental and crucial task in drug discovery. However, prior methods are limited by the requirement for a large number of labeled molecules and their restricted ability to generalize for unseen and new tasks, both of which are essential for real-world applications. To address these challenges, we present MolecularGPT for few-shot MPP. From a perspective on instruction tuning, we fine-tune large language models (LLMs) based on curated molecular instructions spanning over 1000 property prediction tasks. This enables building a versatile and specialized LLM that can be adapted to novel MPP tasks without any fine-tuning through zero- and few-shot in-context learning (ICL). MolecularGPT exhibits competitive in-context reasoning capabilities across 10 downstream evaluation datasets, setting new benchmarks for few-shot molecular prediction tasks. More importantly, with just two-shot examples, MolecularGPT can outperform standard supervised graph neural network methods on 4 out of 7 datasets. It also excels state-of-the-art LLM baselines by up to 15.7% increase on classification accuracy and decrease of 17.9 on regression metrics (e.g., RMSE) under zero-shot. This study demonstrates the potential of LLMs as effective few-shot molecular property predictors. The code is available at https://github.com/NYUSHCS/MolecularGPT.
♻ ☆ A Fundamental Trade-off in Aligned Language Models and its Relation to Sampling Adaptors EMNLP 2024
The relationship between the quality of a string, as judged by a human reader, and its probability, $p(\boldsymbol{y})$ under a language model undergirds the development of better language models. For example, many popular algorithms for sampling from a language model have been conceived with the goal of manipulating $p(\boldsymbol{y})$ to place higher probability on strings that humans deem of high quality. In this article, we examine the probability--quality relationship in language models explicitly aligned to human preferences, e.g., through reinforcement learning through human feedback. We show that, when sampling corpora from an aligned language model, there exists a trade-off between the strings' average reward and average log-likelihood under the prior language model, i.e., the same model before alignment with human preferences. We provide a formal treatment of this phenomenon and demonstrate how a choice of sampling adaptor allows for a selection of how much likelihood we exchange for the reward.
comment: EMNLP 2024
♻ ☆ Entity Matching using Large Language Models EDBT
Entity matching is the task of deciding whether two entity descriptions refer to the same real-world entity. Entity matching is a central step in most data integration pipelines. Many state-of-the-art entity matching methods rely on pre-trained language models (PLMs) such as BERT or RoBERTa. Two major drawbacks of these models for entity matching are that (i) the models require significant amounts of task-specific training data and (ii) the fine-tuned models are not robust concerning out-of-distribution entities. This paper investigates using generative large language models (LLMs) as a less task-specific training data-dependent and more robust alternative to PLM-based matchers. The study covers hosted and open-source LLMs which can be run locally. We evaluate these models in a zero-shot scenario and a scenario where task-specific training data is available. We compare different prompt designs and the prompt sensitivity of the models. We show that there is no single best prompt but that the prompt needs to be tuned for each model/dataset combination. We further investigate (i) the selection of in-context demonstrations, (ii) the generation of matching rules, as well as (iii) fine-tuning LLMs using the same pool of training data. Our experiments show that the best LLMs require no or only a few training examples to perform comparably to PLMs that were fine-tuned using thousands of examples. LLM-based matchers further exhibit higher robustness to unseen entities. We show that GPT4 can generate structured explanations for matching decisions and can automatically identify potential causes of matching errors by analyzing explanations of wrong decisions. We demonstrate that the model can generate meaningful textual descriptions of the identified error classes, which can help data engineers to improve entity matching pipelines.
comment: Published in Proceedings of the 28th International Conference on Extending Database Technology (EDBT), 25th March-28th March, 2025, ISBN 978-3-89318-098-1 on OpenProceedings.org
♻ ☆ 2D-TPE: Two-Dimensional Positional Encoding Enhances Table Understanding for Large Language Models
Tables are ubiquitous across various domains for concisely representing structured information. Empowering large language models (LLMs) to reason over tabular data represents an actively explored direction. However, since typical LLMs only support one-dimensional~(1D) inputs, existing methods often flatten the two-dimensional~(2D) table structure into a sequence of tokens, which can severely disrupt the spatial relationships and result in an inevitable loss of vital contextual information. In this paper, we first empirically demonstrate the detrimental impact of such flattening operations on the performance of LLMs in capturing the spatial information of tables through two elaborate proxy tasks. Subsequently, we introduce a simple yet effective positional encoding method, termed ``2D-TPE'' (Two-Dimensional Table Positional Encoding), to address this challenge. 2D-TPE enables each attention head to dynamically select a permutation order of tokens within the context for attending to them, where each permutation represents a distinct traversal mode for the table, such as column-wise or row-wise traversal. 2D-TPE effectively mitigates the risk of losing essential spatial information while preserving computational efficiency, thus better preserving the table structure. Extensive experiments across five benchmarks demonstrate that 2D-TPE outperforms strong baselines, underscoring the importance of preserving the table structure for accurate table comprehension. Comprehensive analysis further reveals the substantially better scalability of 2D-TPE to large tables than baselines.
♻ ☆ FAME: Towards Factual Multi-Task Model Editing EMNLP 2024
Large language models (LLMs) embed extensive knowledge and utilize it to perform exceptionally well across various tasks. Nevertheless, outdated knowledge or factual errors within LLMs can lead to misleading or incorrect responses, causing significant issues in practical applications. To rectify the fatal flaw without the necessity for costly model retraining, various model editing approaches have been proposed to correct inaccurate knowledge within LLMs in a cost-efficient way. To evaluate these model editing methods, previous work introduced a series of datasets. However, most of the previous datasets only contain fabricated data in a single format, which diverges from real-world model editing scenarios, raising doubts about their usability in practice. To facilitate the application of model editing in real-world scenarios, we propose the challenge of practicality. To resolve such challenges and effectively enhance the capabilities of LLMs, we present FAME, an factual, comprehensive, and multi-task dataset, which is designed to enhance the practicality of model editing. We then propose SKEME, a model editing method that uses a novel caching mechanism to ensure synchronization with the real world. The experiments demonstrate that SKEME performs excellently across various tasks and scenarios, confirming its practicality.
comment: 9 pages, 3 figures. This paper has been accepted by EMNLP 2024
♻ ☆ Prompt Tuning of Deep Neural Networks for Speaker-adaptive Visual Speech Recognition
Visual Speech Recognition (VSR) aims to infer speech into text depending on lip movements alone. As it focuses on visual information to model the speech, its performance is inherently sensitive to personal lip appearances and movements, and this makes the VSR models show degraded performance when they are applied to unseen speakers. In this paper, to remedy the performance degradation of the VSR model on unseen speakers, we propose prompt tuning methods of Deep Neural Networks (DNNs) for speaker-adaptive VSR. Specifically, motivated by recent advances in Natural Language Processing (NLP), we finetune prompts on adaptation data of target speakers instead of modifying the pre-trained model parameters. Different from the previous prompt tuning methods mainly limited to Transformer variant architecture, we explore different types of prompts, the addition, the padding, and the concatenation form prompts that can be applied to the VSR model which is composed of CNN and Transformer in general. With the proposed prompt tuning, we show that the performance of the pre-trained VSR model on unseen speakers can be largely improved by using a small amount of adaptation data (e.g., less than 5 minutes), even if the pre-trained model is already developed with large speaker variations. Moreover, by analyzing the performance and parameters of different types of prompts, we investigate when the prompt tuning is preferred over the finetuning methods. The effectiveness of the proposed method is evaluated on both word- and sentence-level VSR databases, LRW-ID and GRID.
comment: IEEE TPAMI
♻ ☆ BANTH: A Multi-label Hate Speech Detection Dataset for Transliterated Bangla
The proliferation of transliterated texts in digital spaces has emphasized the need for detecting and classifying hate speech in languages beyond English, particularly in low-resource languages. As online discourse can perpetuate discrimination based on target groups, e.g. gender, religion, and origin, multi-label classification of hateful content can help in comprehending hate motivation and enhance content moderation. While previous efforts have focused on monolingual or binary hate classification tasks, no work has yet addressed the challenge of multi-label hate speech classification in transliterated Bangla. We introduce BanTH, the first multi-label transliterated Bangla hate speech dataset comprising 37.3k samples. The samples are sourced from YouTube comments, where each instance is labeled with one or more target groups, reflecting the regional demographic. We establish novel transformer encoder-based baselines by further pre-training on transliterated Bangla corpus. We also propose a novel translation-based LLM prompting strategy for transliterated text. Experiments reveal that our further pre-trained encoders are achieving state-of-the-art performance on the BanTH dataset, while our translation-based prompting outperforms other strategies in the zero-shot setting. The introduction of BanTH not only fills a critical gap in hate speech research for Bangla but also sets the stage for future exploration into code-mixed and multi-label classification challenges in underrepresented languages.
♻ ☆ Understanding Likelihood Over-optimisation in Direct Alignment Algorithms
Direct Alignment Algorithms (DAAs), such as Direct Preference Optimisation (DPO) and Identity Preference Optimisation (IPO), have emerged as alternatives to online Reinforcement Learning from Human Feedback (RLHF) algorithms such as Proximal Policy Optimisation (PPO) for aligning language models to human preferences, without the need for explicit reward modelling. These methods generally aim to increase the likelihood of generating better (preferred) completions while discouraging worse (non-preferred) ones, while staying close to the original model's behaviour. In this work, we explore the relationship between completion likelihood and model performance in state-of-the-art DAAs, and identify a critical issue of likelihood over-optimisation. Contrary to expectations, we find that higher likelihood of better completions and larger margins between better and worse completion likelihoods do not necessarily lead to better performance, and may even degrade it. Our analysis reveals that while higher likelihood correlates with better memorisation of factual knowledge patterns, a slightly lower completion likelihood tends to improve output diversity, thus leading to better generalisation to unseen scenarios. Moreover, we identify two key indicators that signal when over-optimised output diversity begins to harm performance: Decreasing Entropy over Top-k Tokens and Diminishing Top-k Probability Mass. Our experimental results validate that these indicators are reliable signs of declining performance under different regularisations, helping prevent over-optimisation and improve alignment with human preferences.
comment: Preprint Version
♻ ☆ MaiBaam Annotation Guidelines
This document provides the annotation guidelines for MaiBaam, a Bavarian corpus manually annotated with part-of-speech (POS) tags, syntactic dependencies, and German lemmas. MaiBaam belongs to the Universal Dependencies (UD) project, and our annotations elaborate on the general and German UD version 2 guidelines. In this document, we detail how to preprocess and tokenize Bavarian data, provide an overview of the POS tags and dependencies we use, explain annotation decisions that would also apply to closely related languages like German, and lastly we introduce and motivate decisions that are specific to Bavarian grammar.
comment: Updated for UD v2.15 (German lemmas added)
♻ ☆ Evaluating Semantic Variation in Text-to-Image Synthesis: A Causal Perspective
Accurate interpretation and visualization of human instructions are crucial for text-to-image (T2I) synthesis. However, current models struggle to capture semantic variations from word order changes, and existing evaluations, relying on indirect metrics like text-image similarity, fail to reliably assess these challenges. This often obscures poor performance on complex or uncommon linguistic patterns by the focus on frequent word combinations. To address these deficiencies, we propose a novel metric called SemVarEffect and a benchmark named SemVarBench, designed to evaluate the causality between semantic variations in inputs and outputs in T2I synthesis. Semantic variations are achieved through two types of linguistic permutations, while avoiding easily predictable literal variations. Experiments reveal that the CogView-3-Plus and Ideogram 2 performed the best, achieving a score of 0.2/1. Semantic variations in object relations are less understood than attributes, scoring 0.07/1 compared to 0.17-0.19/1. We found that cross-modal alignment in UNet or Transformers plays a crucial role in handling semantic variations, a factor previously overlooked by a focus on textual encoders. Our work establishes an effective evaluation framework that advances the T2I synthesis community's exploration of human instruction understanding. Our benchmark and code are available at https://github.com/zhuxiangru/SemVarBench .
comment: The only change in the current version update is the replacement of the template with a more precise one
♻ ☆ I run as fast as a rabbit, can you? A Multilingual Simile Dialogue Dataset ACL 2023
A simile is a figure of speech that compares two different things (called the tenor and the vehicle) via shared properties. The tenor and the vehicle are usually connected with comparator words such as "like" or "as". The simile phenomena are unique and complex in a real-life dialogue scene where the tenor and the vehicle can be verbal phrases or sentences, mentioned by different speakers, exist in different sentences, or occur in reversed order. However, the current simile research usually focuses on similes in a triplet tuple (tenor, property, vehicle) or a single sentence where the tenor and vehicle are usually entities or noun phrases, which could not reflect complex simile phenomena in real scenarios. In this paper, we propose a novel and high-quality multilingual simile dialogue (MSD) dataset to facilitate the study of complex simile phenomena. The MSD is the largest manually annotated simile data ($\sim$20K) and it contains both English and Chinese data. Meanwhile, the MSD data can also be used on dialogue tasks to test the ability of dialogue systems when using similes. We design 3 simile tasks (recognition, interpretation, and generation) and 2 dialogue tasks (retrieval and generation) with MSD. For each task, we provide experimental results from strong pre-trained or state-of-the-art models. The experiments demonstrate the challenge of MSD and we have released the data/code on GitHub.
comment: 13 Pages, 1 Figure, 12 Tables, ACL 2023 findings
♻ ☆ Can Few-shot Work in Long-Context? Recycling the Context to Generate Demonstrations
Despite recent advancements in Large Language Models (LLMs), their performance on tasks involving long contexts remains sub-optimal. In-Context Learning (ICL) with few-shot examples may be an appealing solution to enhance LLM performance in this scenario; However, na\"ively adding ICL examples with long context introduces challenges, including substantial token overhead added for each few-shot example and context mismatch between the demonstrations and the target query. In this work, we propose to automatically generate few-shot examples for long context QA tasks by recycling contexts. Specifically, given a long input context (1-3k tokens) and a query, we generate additional query-output pairs from the given context as few-shot examples, while introducing the context only once. This ensures that the demonstrations are leveraging the same context as the target query while only adding a small number of tokens to the prompt. We further enhance each demonstration by instructing the model to explicitly identify the relevant paragraphs before the answer, which improves performance while providing fine-grained attribution to the answer source. We apply our method on multiple LLMs and obtain substantial improvements (+16 absolute points on average across models) on various QA datasets with long context, especially when the answer lies within the middle of the context. Surprisingly, despite introducing only single-hop ICL examples, LLMs also successfully generalize to multi-hop long-context QA using our approach.
♻ ☆ Towards Verifiable Text Generation with Evolving Memory and Self-Reflection EMNLP 2024
Despite the remarkable ability of large language models (LLMs) in language comprehension and generation, they often suffer from producing factually incorrect information, also known as hallucination. A promising solution to this issue is verifiable text generation, which prompts LLMs to generate content with citations for accuracy verification. However, verifiable text generation is non-trivial due to the focus-shifting phenomenon, the intricate reasoning needed to align the claim with correct citations, and the dilemma between the precision and breadth of retrieved documents. In this paper, we present VTG, an innovative framework for Verifiable Text Generation with evolving memory and self-reflection. VTG introduces evolving long short-term memory to retain both valuable documents and recent documents. A two-tier verifier equipped with an evidence finder is proposed to rethink and reflect on the relationship between the claim and citations. Furthermore, active retrieval and diverse query generation are utilized to enhance both the precision and breadth of the retrieved documents. We conduct extensive experiments on five datasets across three knowledge-intensive tasks and the results reveal that VTG significantly outperforms baselines.
comment: EMNLP 2024 Main Conference
♻ ☆ Harnessing Webpage UIs for Text-Rich Visual Understanding
Text-rich visual understanding-the ability to process environments where dense textual content is integrated with visuals-is crucial for multimodal large language models (MLLMs) to interact effectively with structured environments. To enhance this capability, we propose synthesizing general multimodal instructions from webpage UIs using text-based large language models (LLMs). Despite lacking direct visual input, text-based LLMs are able to process structured text representations from webpage accessibility trees. These instructions are then paired with UI screenshots to train multimodal models. We introduce MultiUI, a dataset containing 7.3 million samples from 1 million websites, covering diverse multimodal tasks and UI layouts. Models trained on MultiUI not only excel in web UI tasks-achieving up to a 48% improvement on VisualWebBench and a 19.1% boost in element accuracy on a web agent dataset Mind2Web-but also generalize surprisingly well to non-web UI tasks and even to non-UI domains, such as document understanding, OCR, and chart interpretation. These results highlight the broad applicability of web UI data for advancing text-rich visual understanding across various scenarios.
♻ ☆ Dating ancient manuscripts using radiocarbon and AI-based writing style analysis
Determining the chronology of ancient handwritten manuscripts is essential for reconstructing the evolution of ideas. For the Dead Sea Scrolls, this is particularly important. However, there is an almost complete lack of date-bearing manuscripts evenly distributed across the timeline and written in similar scripts available for palaeographic comparison. Here, we present Enoch, a state-of-the-art AI-based date-prediction model, trained on the basis of new radiocarbon-dated samples of the scrolls. Enoch uses established handwriting-style descriptors and applies Bayesian ridge regression. The challenge of this study is that the number of radiocarbon-dated manuscripts is small, while current machine learning requires an abundance of training data. We show that by using combined angular and allographic writing style feature vectors and applying Bayesian ridge regression, Enoch could predict the radiocarbon-based dates from style, supported by leave-one-out validation, with varied MAEs of 27.9 to 30.7 years relative to the radiocarbon dating. Enoch was then used to estimate the dates of 135 unseen manuscripts, revealing that 79 per cent of the samples were considered 'realistic' upon palaeographic post-hoc evaluation. We present a new chronology of the scrolls. The radiocarbon ranges and Enoch's style-based predictions are often older than the traditionally assumed palaeographic estimates. In the range of 300-50 BCE, Enoch's date prediction provides an improved granularity. The study is in line with current developments in multimodal machine-learning techniques, and the methods can be used for date prediction in other partially-dated manuscript collections. This research shows how Enoch's quantitative, probability-based approach can be a tool for palaeographers and historians, re-dating ancient Jewish key texts and contributing to current debates on Jewish and Christian origins.
comment: 16 pages of main article, 103 pages of supplementary materials; the first version of this article is originally prepared in July 2023 after the completion of all the experiments
♻ ☆ Conversational Recommender System and Large Language Model Are Made for Each Other in E-commerce Pre-sales Dialogue EMNLP 2023
E-commerce pre-sales dialogue aims to understand and elicit user needs and preferences for the items they are seeking so as to provide appropriate recommendations. Conversational recommender systems (CRSs) learn user representation and provide accurate recommendations based on dialogue context, but rely on external knowledge. Large language models (LLMs) generate responses that mimic pre-sales dialogues after fine-tuning, but lack domain-specific knowledge for accurate recommendations. Intuitively, the strengths of LLM and CRS in E-commerce pre-sales dialogues are complementary, yet no previous work has explored this. This paper investigates the effectiveness of combining LLM and CRS in E-commerce pre-sales dialogues, proposing two collaboration methods: CRS assisting LLM and LLM assisting CRS. We conduct extensive experiments on a real-world dataset of Ecommerce pre-sales dialogues. We analyze the impact of two collaborative approaches with two CRSs and two LLMs on four tasks of Ecommerce pre-sales dialogue. We find that collaborations between CRS and LLM can be very effective in some cases.
comment: EMNLP 2023 Findings
♻ ☆ Unraveling and Mitigating Retriever Inconsistencies in Retrieval-Augmented Large Language Models ACL 2024
Although Retrieval-Augmented Large Language Models (RALMs) demonstrate their superiority in terms of factuality, they do not consistently outperform the original retrieval-free Language Models (LMs). Our experiments reveal that this example-level performance inconsistency exists not only between retrieval-augmented and retrieval-free LM but also among different retrievers. To understand this phenomenon, we investigate the degeneration behavior of RALMs and theoretically decompose it into four categories. Further analysis based on our decomposition reveals that the innate difference in knowledge sources and the unpredictable degeneration of the reader model contribute most to the inconsistency. Drawing from our analysis, we introduce Ensemble of Retrievers (EoR), a trainable framework that can adaptively retrieve from different knowledge sources and effectively decrease unpredictable reader errors. Our experiments on Open Domain Question Answering show that EoR substantially improves performance over the RALM with a single retriever by considerably reducing inconsistent behaviors.
comment: ACL 2024 (findings)
♻ ☆ PARIKSHA: A Large-Scale Investigation of Human-LLM Evaluator Agreement on Multilingual and Multi-Cultural Data EMNLP 2024
Evaluation of multilingual Large Language Models (LLMs) is challenging due to a variety of factors -- the lack of benchmarks with sufficient linguistic diversity, contamination of popular benchmarks into LLM pre-training data and the lack of local, cultural nuances in translated benchmarks. In this work, we study human and LLM-based evaluation in a multilingual, multi-cultural setting. We evaluate 30 models across 10 Indic languages by conducting 90K human evaluations and 30K LLM-based evaluations and find that models such as GPT-4o and Llama-3 70B consistently perform best for most Indic languages. We build leaderboards for two evaluation settings - pairwise comparison and direct assessment and analyze the agreement between humans and LLMs. We find that humans and LLMs agree fairly well in the pairwise setting but the agreement drops for direct assessment evaluation especially for languages such as Bengali and Odia. We also check for various biases in human and LLM-based evaluation and find evidence of self-bias in the GPT-based evaluator. Our work presents a significant step towards scaling up multilingual evaluation of LLMs.
comment: Accepted to EMNLP 2024
♻ ☆ Graph Neural Network Enhanced Retrieval for Question Answering of LLMs
Retrieval augmented generation has revolutionized large language model (LLM) outputs by providing factual supports. Nevertheless, it struggles to capture all the necessary knowledge for complex reasoning questions. Existing retrieval methods typically divide reference documents into passages, treating them in isolation. These passages, however, are often interrelated, such as passages that are contiguous or share the same keywords. Therefore, it is crucial to recognize such relatedness for enhancing the retrieval process. In this paper, we propose a novel retrieval method, called GNN-Ret, which leverages graph neural networks (GNNs) to enhance retrieval by exploiting the relatedness between passages. Specifically, we first construct a graph of passages by connecting passages that are structure-related or keyword-related. A graph neural network (GNN) is then leveraged to exploit the relationships between passages and improve the retrieval of supporting passages. Furthermore, we extend our method to handle multi-hop reasoning questions using a recurrent graph neural network (RGNN), named RGNN-Ret. At each step, RGNN-Ret integrates the graphs of passages from previous steps, thereby enhancing the retrieval of supporting passages. Extensive experiments on benchmark datasets demonstrate that GNN-Ret achieves higher accuracy for question answering with a single query of LLMs than strong baselines that require multiple queries, and RGNN-Ret further improves accuracy and achieves state-of-the-art performance, with up to 10.4% accuracy improvement on the 2WikiMQA dataset.
comment: Under review
♻ ☆ On the Use of Large Language Models to Generate Capability Ontologies
Capability ontologies are increasingly used to model functionalities of systems or machines. The creation of such ontological models with all properties and constraints of capabilities is very complex and can only be done by ontology experts. However, Large Language Models (LLMs) have shown that they can generate machine-interpretable models from natural language text input and thus support engineers / ontology experts. Therefore, this paper investigates how LLMs can be used to create capability ontologies. We present a study with a series of experiments in which capabilities with varying complexities are generated using different prompting techniques and with different LLMs. Errors in the generated ontologies are recorded and compared. To analyze the quality of the generated ontologies, a semi-automated approach based on RDF syntax checking, OWL reasoning, and SHACL constraints is used. The results of this study are very promising because even for complex capabilities, the generated ontologies are almost free of errors.
comment: \c{opyright} 2024 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works
♻ ☆ Toward a Method to Generate Capability Ontologies from Natural Language Descriptions
To achieve a flexible and adaptable system, capability ontologies are increasingly leveraged to describe functions in a machine-interpretable way. However, modeling such complex ontological descriptions is still a manual and error-prone task that requires a significant amount of effort and ontology expertise. This contribution presents an innovative method to automate capability ontology modeling using Large Language Models (LLMs), which have proven to be well suited for such tasks. Our approach requires only a natural language description of a capability, which is then automatically inserted into a predefined prompt using a few-shot prompting technique. After prompting an LLM, the resulting capability ontology is automatically verified through various steps in a loop with the LLM to check the overall correctness of the capability ontology. First, a syntax check is performed, then a check for contradictions, and finally a check for hallucinations and missing ontology elements. Our method greatly reduces manual effort, as only the initial natural language description and a final human review and possible correction are necessary, thereby streamlining the capability ontology generation process.
comment: \c{opyright} 2024 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works
♻ ☆ Large Language Model Based Generative Error Correction: A Challenge and Baselines for Speech Recognition, Speaker Tagging, and Emotion Recognition
Given recent advances in generative AI technology, a key question is how large language models (LLMs) can enhance acoustic modeling tasks using text decoding results from a frozen, pretrained automatic speech recognition (ASR) model. To explore new capabilities in language modeling for speech processing, we introduce the generative speech transcription error correction (GenSEC) challenge. This challenge comprises three post-ASR language modeling tasks: (i) post-ASR transcription correction, (ii) speaker tagging, and (iii) emotion recognition. These tasks aim to emulate future LLM-based agents handling voice-based interfaces while remaining accessible to a broad audience by utilizing open pretrained language models or agent-based APIs. We also discuss insights from baseline evaluations, as well as lessons learned for designing future evaluations.
comment: IEEE SLT 2024. The initial draft version has been done in December 2023. Post-ASR Text Processing and Understanding Community and LlaMA-7B pre-training correction model: https://huggingface.co/GenSEC-LLM/SLT-Task1-Llama2-7b-HyPo-baseline
♻ ☆ VLFeedback: A Large-Scale AI Feedback Dataset for Large Vision-Language Models Alignment EMNLP 2024
As large vision-language models (LVLMs) evolve rapidly, the demand for high-quality and diverse data to align these models becomes increasingly crucial. However, the creation of such data with human supervision proves costly and time-intensive. In this paper, we investigate the efficacy of AI feedback to scale supervision for aligning LVLMs. We introduce VLFeedback, the first large-scale vision-language feedback dataset, comprising over 82K multi-modal instructions and comprehensive rationales generated by off-the-shelf models without human annotations. To evaluate the effectiveness of AI feedback for vision-language alignment, we train Silkie, an LVLM fine-tuned via direct preference optimization on VLFeedback. Silkie showcases exceptional performance regarding helpfulness, visual faithfulness, and safety metrics. It outperforms its base model by 6.9\% and 9.5\% in perception and cognition tasks, reduces hallucination issues on MMHal-Bench, and exhibits enhanced resilience against red-teaming attacks. Furthermore, our analysis underscores the advantage of AI feedback, particularly in fostering preference diversity to deliver more comprehensive improvements. Our dataset, training code and models are available at https://vlf-silkie.github.io.
comment: EMNLP 2024 Main Conference camera-ready version (fixed small typos). This article supersedes arXiv:2312.10665
♻ ☆ WaterMax: breaking the LLM watermark detectability-robustness-quality trade-off
Watermarking is a technical means to dissuade malfeasant usage of Large Language Models. This paper proposes a novel watermarking scheme, so-called WaterMax, that enjoys high detectability while sustaining the quality of the generated text of the original LLM. Its new design leaves the LLM untouched (no modification of the weights, logits, temperature, or sampling technique). WaterMax balances robustness and complexity contrary to the watermarking techniques of the literature inherently provoking a trade-off between quality and robustness. Its performance is both theoretically proven and experimentally validated. It outperforms all the SotA techniques under the most complete benchmark suite. Code available at https://github.com/eva-giboulot/WaterMax.
♻ ☆ LLM Critics Help Catch Bugs in Mathematics: Towards a Better Mathematical Verifier with Natural Language Feedback
In recent progress, mathematical verifiers have achieved success in mathematical reasoning tasks by validating the correctness of solutions generated by policy models. However, existing verifiers are trained with binary classification labels, which are not informative enough for the model to accurately assess the solutions. To mitigate the aforementioned insufficiency of binary labels, we introduce step-wise natural language feedback as rationale labels, that is, the correctness of each step and the detailed explanations. In this paper, we propose Math-Minos, a natural language feedback-enhanced verifier by constructing automatically generated training data and a two-stage training paradigm for effective training and efficient inference. Our experiments reveal that a small set of natural language feedback can significantly boost the performance of the verifier in both verification and reinforcement learning. We have released the code and data for further exploration.
comment: 15 pages
♻ ☆ PertEval: Unveiling Real Knowledge Capacity of LLMs with Knowledge-Invariant Perturbations NeurIPS '24
Expert-designed close-ended benchmarks are indispensable in assessing the knowledge capacity of large language models (LLMs). Despite their widespread use, concerns have mounted regarding their reliability due to limited test scenarios and an unavoidable risk of data contamination. To rectify this, we present PertEval, a toolkit devised for in-depth probing of LLMs' knowledge capacity through \textbf{knowledge-invariant perturbations}. These perturbations employ human-like restatement techniques to generate on-the-fly test samples from static benchmarks, meticulously retaining knowledge-critical content while altering irrelevant details. Our toolkit further includes a suite of \textbf{response consistency analyses} that compare performance on raw vs. perturbed test sets to precisely assess LLMs' genuine knowledge capacity. Six representative LLMs are re-evaluated using PertEval. Results reveal significantly inflated performance of the LLMs on raw benchmarks, including an absolute 25.8% overestimation for GPT-4. Additionally, through a nuanced response pattern analysis, we discover that PertEval retains LLMs' uncertainty to specious knowledge, and reveals their potential rote memorization to correct options which leads to overestimated performance. We also find that the detailed response consistency analyses by PertEval could illuminate various weaknesses in existing LLMs' knowledge mastery and guide the development of refinement. Our findings provide insights for advancing more robust and genuinely knowledgeable LLMs. Our code is available at \url{https://github.com/aigc-apps/PertEval}.
comment: Accepted by NeurIPS '24 D&B Spotlight; 28 pages, 15 figures, 14 tables
♻ ☆ SciAssess: Benchmarking LLM Proficiency in Scientific Literature Analysis
Recent breakthroughs in Large Language Models (LLMs) have revolutionized scientific literature analysis. However, existing benchmarks fail to adequately evaluate the proficiency of LLMs in this domain, particularly in scenarios requiring higher-level abilities beyond mere memorization and the handling of multimodal data. In response to this gap, we introduce SciAssess, a benchmark specifically designed for the comprehensive evaluation of LLMs in scientific literature analysis. It aims to thoroughly assess the efficacy of LLMs by evaluating their capabilities in Memorization (L1), Comprehension (L2), and Analysis \& Reasoning (L3). It encompasses a variety of tasks drawn from diverse scientific fields, including biology, chemistry, material, and medicine. To ensure the reliability of SciAssess, rigorous quality control measures have been implemented, ensuring accuracy, anonymization, and compliance with copyright standards. SciAssess evaluates 11 LLMs, highlighting their strengths and areas for improvement. We hope this evaluation supports the ongoing development of LLM applications in scientific literature analysis. SciAssess and its resources are available at \url{https://github.com/sci-assess/SciAssess}.
♻ ☆ Does Mapo Tofu Contain Coffee? Probing LLMs for Food-related Cultural Knowledge
Recent studies have highlighted the presence of cultural biases in Large Language Models (LLMs), yet often lack a robust methodology to dissect these phenomena comprehensively. Our work aims to bridge this gap by delving into the Food domain, a universally relevant yet culturally diverse aspect of human life. We introduce FmLAMA, a multilingual dataset centered on food-related cultural facts and variations in food practices. We analyze LLMs across various architectures and configurations, evaluating their performance in both monolingual and multilingual settings. By leveraging templates in six different languages, we investigate how LLMs interact with language-specific and cultural knowledge. Our findings reveal that (1) LLMs demonstrate a pronounced bias towards food knowledge prevalent in the United States; (2) Incorporating relevant cultural context significantly improves LLMs' ability to access cultural knowledge; (3) The efficacy of LLMs in capturing cultural nuances is highly dependent on the interplay between the probing language, the specific model architecture, and the cultural context in question. This research underscores the complexity of integrating cultural understanding into LLMs and emphasizes the importance of culturally diverse datasets to mitigate biases and enhance model performance across different cultural domains.
comment: cultural bias analysis, cultural knowledge probing, large language models, cultural NLP
♻ ☆ Synergizing In-context Learning with Hints for End-to-end Task-oriented Dialog Systems EMNLP2024
End-to-end Task-Oriented Dialog (TOD) systems typically require extensive training datasets to perform well. In contrast, large language model (LLM) based TOD systems can excel even with limited data due to their ability to learn tasks through in-context exemplars. However, these models lack alignment with the style of responses in training data and often generate comprehensive responses, making it difficult for users to grasp the information quickly. In response, we propose SyncTOD that synergizes LLMs with task-specific hints to improve alignment in low-data settings. SyncTOD employs small auxiliary models to provide hints and select exemplars for in-context prompts. With ChatGPT, SyncTOD achieves superior performance compared to LLM-based baselines and SoTA models in low-data settings, while retaining competitive performance in full-data settings.
comment: EMNLP2024 Camera-Ready Version
♻ ☆ Hyper-multi-step: The Truth Behind Difficult Long-context Tasks
Long-context language models (LCLM), characterized by their extensive context window, is becoming increasingly popular. Meanwhile, many long-context benchmarks present challenging tasks that even the most advanced LCLMs struggle to complete. However, the underlying sources of various challenging long-context tasks have seldom been studied. To bridge this gap, we conduct experiments to indicate their difficulty stems primarily from two basic issues: "multi-matching retrieval," which requires the simultaneous retrieval of multiple items, and "logic-based retrieval," which necessitates logical judgment within retrieval criteria. These two problems, while seemingly straightforward, actually exceed the capabilities of LCLMs because they are proven to be hyper-multi-step (demanding numerous steps to solve) in nature. This finding could explain why LLMs struggle with more advanced long-context tasks, providing a more accurate perspective for rethinking solutions for them.
comment: Our code is publicly available at https://github.com/yuyijiong/hard_retrieval_for_llm and the datasets is at https://huggingface.co/datasets/yuyijiong/difficult_retrieval
♻ ☆ Fisher Information-based Efficient Curriculum Federated Learning with Large Language Models EMNLP 2024
As a promising paradigm to collaboratively train models with decentralized data, Federated Learning (FL) can be exploited to fine-tune Large Language Models (LLMs). While LLMs correspond to huge size, the scale of the training data significantly increases, which leads to tremendous amounts of computation and communication costs. The training data is generally non-Independent and Identically Distributed (non-IID), which requires adaptive data processing within each device. Although Low Rank Adaptation (LoRA) can significantly reduce the scale of parameters to update in the fine-tuning process, it still takes unaffordable time to transfer the low-rank parameters of all the layers in LLMs. In this paper, we propose a Fisher Information-based Efficient Curriculum Federated Learning framework (FibecFed) with two novel methods, i.e., adaptive federated curriculum learning and efficient sparse parameter update. First, we propose a fisher information-based method to adaptively sample data within each device to improve the effectiveness of the FL fine-tuning process. Second, we dynamically select the proper layers for global aggregation and sparse parameters for local update with LoRA so as to improve the efficiency of the FL fine-tuning process. Extensive experimental results based on 10 datasets demonstrate that FibecFed yields excellent performance (up to 45.35% in terms of accuracy) and superb fine-tuning speed (up to 98.61% faster) compared with 17 baseline approaches).
comment: 27 pages, 8 figures, 14 tables, to appear in EMNLP 2024
♻ ☆ QUIS: Question-guided Insights Generation for Automated Exploratory Data Analysis
Discovering meaningful insights from a large dataset, known as Exploratory Data Analysis (EDA), is a challenging task that requires thorough exploration and analysis of the data. Automated Data Exploration (ADE) systems use goal-oriented methods with Large Language Models and Reinforcement Learning towards full automation. However, these methods require human involvement to anticipate goals that may limit insight extraction, while fully automated systems demand significant computational resources and retraining for new datasets. We introduce QUIS, a fully automated EDA system that operates in two stages: insight generation (ISGen) driven by question generation (QUGen). The QUGen module generates questions in iterations, refining them from previous iterations to enhance coverage without human intervention or manually curated examples. The ISGen module analyzes data to produce multiple relevant insights in response to each question, requiring no prior training and enabling QUIS to adapt to new datasets.
comment: Accepted for ENLP 2024 Industry Track
♻ ☆ SciLitLLM: How to Adapt LLMs for Scientific Literature Understanding
Scientific literature understanding is crucial for extracting targeted information and garnering insights, thereby significantly advancing scientific discovery. Despite the remarkable success of Large Language Models (LLMs), they face challenges in scientific literature understanding, primarily due to (1) a lack of scientific knowledge and (2) unfamiliarity with specialized scientific tasks. To develop an LLM specialized in scientific literature understanding, we propose a hybrid strategy that integrates continual pre-training (CPT) and supervised fine-tuning (SFT), to simultaneously infuse scientific domain knowledge and enhance instruction-following capabilities for domain-specific tasks.cIn this process, we identify two key challenges: (1) constructing high-quality CPT corpora, and (2) generating diverse SFT instructions. We address these challenges through a meticulous pipeline, including PDF text extraction, parsing content error correction, quality filtering, and synthetic instruction creation. Applying this strategy, we present a suite of LLMs: SciLitLLM, specialized in scientific literature understanding. These models demonstrate promising performance on scientific literature understanding benchmarks. Our contributions are threefold: (1) We present an effective framework that integrates CPT and SFT to adapt LLMs to scientific literature understanding, which can also be easily adapted to other domains. (2) We propose an LLM-based synthesis method to generate diverse and high-quality scientific instructions, resulting in a new instruction set -- SciLitIns -- for supervised fine-tuning in less-represented scientific domains. (3) SciLitLLM achieves promising performance improvements on scientific literature understanding benchmarks.
♻ ☆ SeerAttention: Learning Intrinsic Sparse Attention in Your LLMs
Attention is the cornerstone of modern Large Language Models (LLMs). Yet its quadratic complexity limits the efficiency and scalability of LLMs, especially for those with a long-context window. A promising approach addressing this limitation is to leverage the sparsity in attention. However, existing sparsity-based solutions predominantly rely on predefined patterns or heuristics to approximate sparsity. This practice falls short to fully capture the dynamic nature of attention sparsity in language-based tasks. This paper argues that attention sparsity should be learned rather than predefined. To this end, we design SeerAttention, a new Attention mechanism that augments the conventional attention with a learnable gate that adaptively selects significant blocks in an attention map and deems the rest blocks sparse. Such block-level sparsity effectively balances accuracy and speedup. To enable efficient learning of the gating network, we develop a customized FlashAttention implementation that extracts the block-level ground truth of attention map with minimum overhead. SeerAttention not only applies to post-training, but also excels in long-context fine-tuning. Our results show that at post-training stages, SeerAttention significantly outperforms state-of-the-art static or heuristic-based sparse attention methods, while also being more versatile and flexible to adapt to varying context lengths and sparsity ratios. When applied to long-context fine-tuning with YaRN, SeerAttention can achieve a remarkable 90% sparsity ratio at a 32k context length with minimal perplexity loss, offering a 5.67x speedup over FlashAttention-2.
♻ ☆ P3: A Policy-Driven, Pace-Adaptive, and Diversity-Promoted Framework for data pruning in LLM Training
In the rapidly advancing field of Large Language Models (LLMs), effectively leveraging existing datasets during fine-tuning to maximize the model's potential is of paramount importance. This paper introduces P3, an adaptive framework aimed at optimizing the task-specific fine-tuning process through iterative data pruning. P3 consists of three key components: (1) Policy-driven Difficulty Measurement, which dynamically assesses data difficulty based on the model's real-time performance, replacing static metrics with adaptable evaluations; (2) Pace-Adaptive Selection, leveraging self-paced learning to progressively introduce more challenging data, thereby enhancing model capability; (3) Diversity Promotion, incorporating Determinantal Point Process (DPP) to ensure data diversity across epochs, enriching the learning process. We validate P3 on the reasoning scenarios, APPS and MATH, demonstrating significant improvements over traditional data pruning methods. By advancing dynamic data selection and utilization strategies, P3 contributes both a theoretical framework and concrete approach to fully exploit existing data for LLMs' performance improvement, offering utility across diverse tasks.
♻ ☆ LatentExplainer: Explaining Latent Representations in Deep Generative Models with Multi-modal Foundation Models
Deep generative models like VAEs and diffusion models have advanced various generation tasks by leveraging latent variables to learn data distributions and generate high-quality samples. Despite the field of explainable AI making strides in interpreting machine learning models, understanding latent variables in generative models remains challenging. This paper introduces \textit{LatentExplainer}, a framework for automatically generating semantically meaningful explanations of latent variables in deep generative models. \textit{LatentExplainer} tackles three main challenges: inferring the meaning of latent variables, aligning explanations with inductive biases, and handling varying degrees of explainability. Our approach perturbs latent variables, interpreting changes in generated data, and uses multi-modal large language models (MLLMs) to produce human-understandable explanations. We evaluate our proposed method on several real-world and synthetic datasets, and the results demonstrate superior performance in generating high-quality explanations for latent variables. The results highlight the effectiveness of incorporating inductive biases and uncertainty quantification, significantly enhancing model interpretability.
♻ ☆ Supervised Fine-Tuning Achieve Rapid Task Adaption Via Alternating Attention Head Activation Patterns
LLMs' performance on complex tasks is still unsatisfactory. A key issue is that presently LLMs learn in a data-driven schema, while the instructions about these complex tasks are both scarce and hard to collect or construct. On the contrary, a prominent phenomenon is that LLMs can learn rather fast on simpler tasks with adequate prior knowledge captured during pretraining stage. Thus, if the prerequisite and mechanism of such rapid generalization could be elucidated, it could enhance the efficiency and effectiveness of the LLM's ability to learn complex tasks. Thus, in this paper, we employ a gradient-based method, to dissect the process that the SFT process adapts LLMs to downstream tasks via the perspective of attention patterns. We find that: (1) LLMs selectively activate task-specific attention heads during SFT; (2) activation patterns for complex tasks are combinations of basic task patterns; and (3) changes in a few parameters can significantly impact activation patterns after SFT on a small number of samples.Based on these insights, experiments are conducted to actually enhance the efficiency and effectiveness of SFT.
comment: in review
♻ ☆ From Introspection to Best Practices: Principled Analysis of Demonstrations in Multimodal In-Context Learning
Motivated by in-context learning (ICL) capabilities of Large Language models (LLMs), multimodal LLMs with additional visual modality are also exhibited with similar ICL abilities when multiple image-text pairs are provided as demonstrations. However, relatively less work has been done to investigate the principles behind how and why multimodal ICL works. We conduct a systematic and principled evaluation of multimodal ICL for models of different scales on a broad spectrum of new yet critical tasks. Through perturbations over different modality information, we show that modalities matter differently across tasks in multimodal ICL. Guided by task-specific modality impact, we recommend modality-driven demonstration strategies to boost ICL performance. We also find that models may follow inductive biases from multimodal ICL even if they are rarely seen in or contradict semantic priors from pretraining data. Our principled analysis provides a comprehensive way of understanding the role of demonstrations in multimodal in-context learning, and sheds light on effectively improving multimodal ICL on a wide range of tasks.
♻ ☆ Everything is Editable: Extend Knowledge Editing to Unstructured Data in Large Language Models
Recent knowledge editing methods have primarily focused on modifying structured knowledge in large language models. However, this task setting overlooks the fact that a significant portion of real-world knowledge is stored in an unstructured format, characterized by long-form content, noise, and a complex yet comprehensive nature. Techniques like local layer key-value storage and term-driven optimization, as used in previous methods like MEMIT, are not effective for handling unstructured knowledge. To address these challenges, we propose a novel Unstructured Knowledge Editing method, namely UnKE, which extends previous assumptions in the layer dimension and token dimension. Firstly, in the layer dimension, we propose non-local block key-value storage to replace local layer key-value storage, increasing the representation ability of key-value pairs and incorporating attention layer knowledge. Secondly, in the token dimension, we replace term-driven optimization with cause-driven optimization, which edits the last token directly while preserving context, avoiding the need to locate terms and preventing the loss of context information. Results on newly proposed unstructured knowledge editing dataset (UnKEBench) and traditional structured datasets demonstrate that UnKE achieves remarkable performance, surpassing strong baselines. In addition, UnKE has robust batch editing and sequential editing capabilities.
♻ ☆ Amphista: Bi-directional Multi-head Decoding for Accelerating LLM Inference
Large Language Models (LLMs) inherently use autoregressive decoding, which lacks parallelism in inference and results in significantly slow inference speed. While methods such as Medusa constructs parallelized heads, they lack adequate information interaction across different prediction positions. To overcome this limitation, we introduce Amphista, an enhanced speculative decoding framework that builds upon Medusa. Specifically, Amphista models an Auto-embedding Block capable of parallel inference, incorporating bi-directional attention to enable interaction between different drafting heads. Additionally, Amphista integrates Staged Adaptation Layers, which ensure a seamless transition of semantic information from the target model's autoregressive inference to the drafting heads' non-autoregressive inference, effectively achieving paradigm shift and feature fusion. Experimental results on Vicuna models using MT-Bench and Spec-Bench demonstrate that Amphista achieves substantial acceleration while maintaining generation quality. On MT-Bench, Amphista delivers up to 2.75$\times$ speedup over vanilla autoregressive decoding and 1.40$\times$ over Medusa on Vicuna 33B in wall-clock time.
♻ ☆ LayerSkip: Enabling Early Exit Inference and Self-Speculative Decoding ACL 2024
We present LayerSkip, an end-to-end solution to speed-up inference of large language models (LLMs). First, during training we apply layer dropout, with low dropout rates for earlier layers and higher dropout rates for later layers, and an early exit loss where all transformer layers share the same exit. Second, during inference, we show that this training recipe increases the accuracy of early exit at earlier layers, without adding any auxiliary layers or modules to the model. Third, we present a novel self-speculative decoding solution where we exit at early layers and verify and correct with remaining layers of the model. Our proposed self-speculative decoding approach has less memory footprint than other speculative decoding approaches and benefits from shared compute and activations of the draft and verification stages. We run experiments on different Llama model sizes on different types of training: pretraining from scratch, continual pretraining, finetuning on specific data domain, and finetuning on specific task. We implement our inference solution and show speedups of up to 2.16x on summarization for CNN/DM documents, 1.82x on coding, and 2.0x on TOPv2 semantic parsing task. We open source our code and checkpoints at https://github.com/facebookresearch/LayerSkip.
comment: ACL 2024
♻ ☆ A Tighter Complexity Analysis of SparseGPT
In this work, we improved the analysis of the running time of SparseGPT [Frantar, Alistarh ICML 2023] from $O(d^{3})$ to $O(d^{\omega} + d^{2+a+o(1)} + d^{1+\omega(1,1,a)-a})$ for any $a \in [0, 1]$, where $\omega$ is the exponent of matrix multiplication. In particular, for the current $\omega \approx 2.371$ [Alman, Duan, Williams, Xu, Xu, Zhou 2024], our running time boils down to $O(d^{2.53})$. This running time is due to the analysis of the lazy update behavior in iterative maintenance problems such as [Deng, Song, Weinstein 2022; Brand, Song, Zhou ICML 2024].
♻ ☆ A Comprehensive Study of Multilingual Confidence Estimation on Large Language Models
The tendency of Large Language Models (LLMs) to generate hallucinations raises concerns regarding their reliability. Therefore, confidence estimations indicating the extent of trustworthiness of the generations become essential. However, current LLM confidence estimations in languages other than English remain underexplored. This paper addresses this gap by introducing a comprehensive investigation of Multilingual Confidence estimation (MlingConf) on LLMs, focusing on both language-agnostic (LA) and language-specific (LS) tasks to explore the performance and language dominance effects of multilingual confidence estimations on different tasks. The benchmark comprises four meticulously checked and human-evaluate high-quality multilingual datasets for LA tasks and one for the LS task tailored to specific social, cultural, and geographical contexts of a language. Our experiments reveal that on LA tasks English exhibits notable linguistic dominance in confidence estimations than other languages, while on LS tasks, using question-related language to prompt LLMs demonstrates better linguistic dominance in multilingual confidence estimations. The phenomena inspire a simple yet effective native-tone prompting strategy by employing language-specific prompts for LS tasks, effectively improving LLMs' reliability and accuracy on LS tasks.
comment: Comments: n pages; Previously this version appeared as arXiv:2410.12478 which was submitted as a new work by accident
♻ ☆ ExACT: Teaching AI Agents to Explore with Reflective-MCTS and Exploratory Learning
Autonomous agents have demonstrated significant potential in automating complex multistep decision-making tasks. However, even state-of-the-art vision-language models (VLMs), such as GPT-4o, still fall short of human-level performance, particularly in intricate web environments and long-horizon tasks. To address these limitations, we present ExACT, an approach to combine test-time search and self-learning to build o1-like models for agentic applications. We first introduce Reflective Monte Carlo Tree Search (R-MCTS), a novel test time algorithm designed to enhance AI agents' ability to explore decision space on the fly. R-MCTS extends traditional MCTS by 1) incorporating contrastive reflection, allowing agents to learn from past interactions and dynamically improve their search efficiency; and 2) using multi-agent debate for reliable state evaluation. Next, we introduce Exploratory Learning, a novel learning strategy to teach agents to search at inference time without relying on any external search algorithms. On the challenging VisualWebArena benchmark, our GPT-4o based R-MCTS agent achieves a 6% to 30% relative improvement across various tasks compared to the previous state-of-the-art. Additionally, we show that the knowledge and experience gained from test-time search can be effectively transferred back to GPT-4o via fine-tuning. After Exploratory Learning, GPT-4o 1) demonstrates the ability to explore the environment, evaluate a state, and backtrack to viable ones when it detects that the current state cannot lead to success, and 2) matches 87% of R-MCTS's performance while using significantly less compute. Notably, our work demonstrates the compute scaling properties in both training - data collection with R-MCTS - and testing time. These results suggest a promising research direction to enhance VLMs' capabilities for agentic applications via test-time search and self-learning.
♻ ☆ MlingConf: A Comprehensive Study of Multilingual Confidence Estimation on Large Language Models
The tendency of Large Language Models (LLMs) to generate hallucinations raises concerns regarding their reliability. Therefore, confidence estimations indicating the extent of trustworthiness of the generations become essential. However, current LLM confidence estimations in languages other than English remain underexplored. This paper addresses this gap by introducing a comprehensive investigation of Multilingual Confidence estimation (MlingConf) on LLMs, focusing on both language-agnostic (LA) and language-specific (LS) tasks to explore the performance and language dominance effects of multilingual confidence estimations on different tasks. The benchmark comprises four meticulously checked and human-evaluate high-quality multilingual datasets for LA tasks and one for the LS task tailored to specific social, cultural, and geographical contexts of a language. Our experiments reveal that on LA tasks English exhibits notable linguistic dominance in confidence estimations than other languages, while on LS tasks, using question-related language to prompt LLMs demonstrates better linguistic dominance in multilingual confidence estimations. The phenomena inspire a simple yet effective native-tone prompting strategy by employing language-specific prompts for LS tasks, effectively improving LLMs' reliability and accuracy on LS tasks.
comment: Comments: This work was intended as a replacement of arXiv:2402.13606 and any subsequent updates will appear there
♻ ☆ $\textbf{Only-IF}$:Revealing the Decisive Effect of Instruction Diversity on Generalization
Understanding and accurately following instructions is critical for large language models (LLMs) to be effective across diverse tasks. In this work, we rigorously examine the key factors that enable models to generalize to unseen instructions, providing insights to guide the collection of data for instruction-tuning. Through controlled experiments, inspired by the Turing-complete Markov algorithm, we demonstrate that such generalization $\textbf{only emerges}$ when training data is diversified enough across semantic domains. Our findings also reveal that merely diversifying within limited domains fails to ensure robust generalization. In contrast, cross-domain data diversification, even under constrained data budgets, significantly enhances a model's adaptability. We further extend our analysis to real-world scenarios, including fine-tuning of $\textit{$\textbf{specialist}$}$ and $\textit{$\textbf{generalist}$}$ models. In both cases, we demonstrate that 1) better performance can be achieved by increasing the diversity of an established dataset while keeping the data size constant, and 2) when scaling up the data, diversifying the semantics of instructions is more effective than simply increasing the quantity of similar data. Our research provides important insights for dataset collation, particularly when optimizing model performance by expanding training data for both specialist and generalist scenarios. We show that careful consideration of data diversification is key: training specialist models with data extending beyond their core domain leads to significant performance improvements, while generalist models benefit from diverse data mixtures that enhance their overall instruction-following capabilities across a wide range of applications. Our results highlight the critical role of strategic diversification and offer clear guidelines for improving data quality.
comment: Fix formatting issues
♻ ☆ BenTo: Benchmark Task Reduction with In-Context Transferability
Evaluating large language models (LLMs) is costly: it requires the generation and examination of LLM outputs on a large-scale benchmark of various tasks. This paper investigates how to efficiently reduce the tasks used to benchmark LLMs without affecting the evaluation quality. Our study reveals that task transferability and relevance provide critical information to identify the most representative subset of tasks via optimizing a facility location function. We propose a practically efficient metric for estimating the transferability between two tasks via in-context learning (ICL). By analyzing the pairwise transferability, we can reduce tasks in a modern LLM benchmark (e.g., MMLU or FLAN) to 5% while inducing only a <4% difference to the evaluation on the original benchmark. Compared to prior works, our method is training-free, gradient-free, and highly efficient requiring ICL only.
comment: https://github.com/tianyi-lab/bento
♻ ☆ GraphInsight: Unlocking Insights in Large Language Models for Graph Structure Understanding
Although Large Language Models (LLMs) have demonstrated potential in processing graphs, they struggle with comprehending graphical structure information through prompts of graph description sequences, especially as the graph size increases. We attribute this challenge to the uneven memory performance of LLMs across different positions in graph description sequences, known as ''positional biases''. To address this, we propose GraphInsight, a novel framework aimed at improving LLMs' comprehension of both macro- and micro-level graphical information. GraphInsight is grounded in two key strategies: 1) placing critical graphical information in positions where LLMs exhibit stronger memory performance, and 2) investigating a lightweight external knowledge base for regions with weaker memory performance, inspired by retrieval-augmented generation (RAG). Moreover, GraphInsight explores integrating these two strategies into LLM agent processes for composite graph tasks that require multi-step reasoning. Extensive empirical studies on benchmarks with a wide range of evaluation tasks show that GraphInsight significantly outperforms all other graph description methods (e.g., prompting techniques and reordering strategies) in understanding graph structures of varying sizes.
♻ ☆ AutoPal: Autonomous Adaptation to Users for Personal AI Companionship
Previous research has demonstrated the potential of AI agents to act as companions that can provide constant emotional support for humans. In this paper, we emphasize the necessity of autonomous adaptation in personal AI companionship, an underexplored yet promising direction. Such adaptability is crucial as it can facilitate more tailored interactions with users and allow the agent to evolve in response to users' changing needs. However, imbuing agents with autonomous adaptability presents unique challenges, including identifying optimal adaptations to meet users' expectations and ensuring a smooth transition during the adaptation process. To address them, we devise a hierarchical framework, AutoPal, that enables controllable and authentic adjustments to the agent's persona based on user interactions. A personamatching dataset is constructed to facilitate the learning of optimal persona adaptations. Extensive experiments demonstrate the effectiveness of AutoPal and highlight the importance of autonomous adaptability in AI companionship.
♻ ☆ Efficiently Quantifying and Mitigating Ripple Effects in Model Editing
Large Language Models have revolutionized numerous tasks with their remarkable efficacy. However, editing these models, crucial for rectifying outdated or erroneous information, often leads to a complex issue known as the ripple effect in the hidden space. While difficult to detect, this effect can significantly impede the efficacy of model editing tasks and deteriorate model performance. This paper addresses this scientific challenge by proposing a novel evaluation methodology, Graphical Impact Evaluation(GIE), which quantitatively evaluates the adaptations of the model and the subsequent impact of editing. Furthermore, we introduce the Selective Impact Revision(SIR), a model editing method designed to mitigate this ripple effect. Our comprehensive evaluations reveal that the ripple effect in the hidden space is a significant issue in all current model editing methods. However, our proposed methods, GIE and SIR, effectively identify and alleviate this issue, contributing to the advancement of LLM editing techniques.
♻ ☆ MoR: Mixture of Ranks for Low-Rank Adaptation Tuning
Low-Rank Adaptation (LoRA) drives research to align its performance with full fine-tuning. However, significant challenges remain: (1) Simply increasing the rank size of LoRA does not effectively capture high-rank information, which leads to a performance bottleneck.(2) MoE-style LoRA methods substantially increase parameters and inference latency, contradicting the goals of efficient fine-tuning and ease of application. To address these challenges, we introduce Mixture of Ranks (MoR), which learns rank-specific information for different tasks based on input and efficiently integrates multi-rank information. We firstly propose a new framework that equates the integration of multiple LoRAs to expanding the rank of LoRA. Moreover, we hypothesize that low-rank LoRA already captures sufficient intrinsic information, and MoR can derive high-rank information through mathematical transformations of the low-rank components. Thus, MoR can reduces the learning difficulty of LoRA and enhances its multi-task capabilities. MoR achieves impressive results, with MoR delivering a 1.31\% performance improvement while using only 93.93\% of the parameters compared to baseline methods.
comment: 11 pages, 7 figures
♻ ☆ UniAutoML: A Human-Centered Framework for Unified Discriminative and Generative AutoML with Large Language Models
Automated Machine Learning (AutoML) has simplified complex ML processes such as data pre-processing, model selection, and hyper-parameter searching. However, traditional AutoML frameworks focus solely on discriminative tasks, often falling short in tackling AutoML for generative models. Additionally, these frameworks lack interpretability and user engagement during the training process, primarily due to the absence of human-centered design. It leads to a lack of transparency in final decision-making and limited user control, potentially reducing trust and adoption of AutoML methods. To address these limitations, we introduce UniAutoML, a human-centered AutoML framework that leverages Large Language Models (LLMs) to unify AutoML for both discriminative (e.g., Transformers and CNNs for classification or regression tasks) and generative tasks (e.g., fine-tuning diffusion models or LLMs). The human-centered design of UniAutoML innovatively features a conversational user interface (CUI) that facilitates natural language interactions, providing users with real-time guidance, feedback, and progress updates for better interpretability. This design enhances transparency and user control throughout the AutoML training process, allowing users to seamlessly break down or modify the model being trained. To mitigate potential risks associated with LLM generated content, UniAutoML incorporates a safety guardline that filters inputs and censors outputs. We evaluated UniAutoML's performance and usability through experiments on eight diverse datasets and user studies involving 25 participants, demonstrating that UniAutoML not only enhances performance but also improves user control and trust. Our human-centered design bridges the gap between AutoML capabilities and user understanding, making ML more accessible to a broader audience.
♻ ☆ ACCEPT: Adaptive Codebook for Composite and Efficient Prompt Tuning EMNLP
Prompt Tuning has been a popular Parameter-Efficient Fine-Tuning method attributed to its remarkable performance with few updated parameters on various large-scale pretrained Language Models (PLMs). Traditionally, each prompt has been considered indivisible and updated independently, leading the parameters increase proportionally as prompt length grows. To address this issue, we propose Adaptive Codebook for Composite and Efficient Prompt Tuning (ACCEPT). In our method, we refer to the concept of product quantization (PQ), allowing all soft prompts to share a set of learnable codebook vectors in each subspace, with each prompt differentiated by a set of adaptive weights. We achieve the superior performance on 17 diverse natural language tasks including natural language understanding (NLU) and question answering (QA) tasks by tuning only 0.3% of parameters of the PLMs. Our approach also excels in few-shot and large model settings, highlighting its significant potential.
comment: EMNLP Findings 2024
♻ ☆ On Subjective Uncertainty Quantification and Calibration in Natural Language Generation
Applications of large language models often involve the generation of free-form responses, in which case uncertainty quantification becomes challenging. This is due to the need to identify task-specific uncertainties (e.g., about the semantics) which appears difficult to define in general cases. This work addresses these challenges from a perspective of Bayesian decision theory, starting from the assumption that our utility is characterized by a similarity measure that compares a generated response with a hypothetical true response. We discuss how this assumption enables principled quantification of the model's subjective uncertainty and its calibration. We further derive a measure for epistemic uncertainty, based on a missing data perspective and its characterization as an excess risk. The proposed methods can be applied to black-box language models. We illustrate the methods on question answering and machine translation tasks. Our experiments provide a principled evaluation of task-specific calibration, and demonstrate that epistemic uncertainty offers a promising deferral strategy for efficient data acquisition in in-context learning.
♻ ☆ An Evolved Universal Transformer Memory
Prior methods propose to offset the escalating costs of modern foundation models by dropping specific parts of their contexts with hand-designed rules, while attempting to preserve their original performance. We overcome this trade-off with Neural Attention Memory Models (NAMMs), introducing a learned network for memory management that improves both the performance and efficiency of transformers. We evolve NAMMs atop pre-trained transformers to provide different latent contexts focusing on the most relevant information for individual layers and attention heads. NAMMs are universally applicable to any model using self-attention as they condition exclusively on the values in the produced attention matrices. Learning NAMMs on a small set of problems, we achieve substantial performance improvements across multiple long-context benchmarks while cutting the model's input contexts up to a fraction of the original sizes. We show the generality of our conditioning enables zero-shot transfer of NAMMs trained only on language to entirely new transformer architectures even across input modalities, with their benefits carrying over to vision and reinforcement learning.
comment: 29 pages, 14 figures. Preprint, under submission. Source code is available at https://github.com/SakanaAI/evo-memory
♻ ☆ JAILJUDGE: A Comprehensive Jailbreak Judge Benchmark with Multi-Agent Enhanced Explanation Evaluation Framework
Despite advancements in enhancing LLM safety against jailbreak attacks, evaluating LLM defenses remains a challenge, with current methods often lacking explainability and generalization to complex scenarios, leading to incomplete assessments (e.g., direct judgment without reasoning, low F1 score of GPT-4 in complex cases, bias in multilingual scenarios). To address this, we present JAILJUDGE, a comprehensive benchmark featuring diverse risk scenarios, including synthetic, adversarial, in-the-wild, and multilingual prompts, along with high-quality human-annotated datasets. The JAILJUDGE dataset includes over 35k+ instruction-tune data with reasoning explainability and JAILJUDGETEST, a 4.5k+ labeled set for risk scenarios, and a 6k+ multilingual set across ten languages. To enhance evaluation with explicit reasoning, we propose the JailJudge MultiAgent framework, which enables explainable, fine-grained scoring (1 to 10). This framework supports the construction of instruction-tuning ground truth and facilitates the development of JAILJUDGE Guard, an end-to-end judge model that provides reasoning and eliminates API costs. Additionally, we introduce JailBoost, an attacker-agnostic attack enhancer, and GuardShield, a moderation defense, both leveraging JAILJUDGE Guard. Our experiments demonstrate the state-of-the-art performance of JailJudge methods (JailJudge MultiAgent, JAILJUDGE Guard) across diverse models (e.g., GPT-4, Llama-Guard) and zero-shot scenarios. JailBoost and GuardShield significantly improve jailbreak attack and defense tasks under zero-shot settings, with JailBoost enhancing performance by 29.24% and GuardShield reducing defense ASR from 40.46% to 0.15%.
♻ ☆ A Modular-based Strategy for Mitigating Gradient Conflicts in Simultaneous Speech Translation ICASSP 2025
Simultaneous Speech Translation (SimulST) involves generating target language text while continuously processing streaming speech input, presenting significant real-time challenges. Multi-task learning is often employed to enhance SimulST performance but introduces optimization conflicts between primary and auxiliary tasks, potentially compromising overall efficiency. The existing model-level conflict resolution methods are not well-suited for this task which exacerbates inefficiencies and leads to high GPU memory consumption. To address these challenges, we propose a Modular Gradient Conflict Mitigation (MGCM) strategy that detects conflicts at a finer-grained modular level and resolves them utilizing gradient projection. Experimental results demonstrate that MGCM significantly improves SimulST performance, particularly under medium and high latency conditions, achieving a 0.68 BLEU score gain in offline tasks. Additionally, MGCM reduces GPU memory consumption by over 95\% compared to other conflict mitigation methods, establishing it as a robust solution for SimulST tasks.
comment: Submitted to ICASSP 2025
♻ ☆ Do Not Design, Learn: A Trainable Scoring Function for Uncertainty Estimation in Generative LLMs
Uncertainty estimation (UE) of generative large language models (LLMs) is crucial for evaluating the reliability of generated sequences. A significant subset of UE methods utilize token probabilities to assess uncertainty, aggregating multiple token probabilities into a single UE score using a scoring function. Existing scoring functions for probability-based UE, such as length-normalized scoring and semantic contribution-based weighting, are designed to solve certain aspects of the problem but exhibit limitations, including the inability to handle biased probabilities and complex semantic dependencies between tokens. To address these issues, in this work, we propose Learnable Response Scoring (LARS) function, a novel scoring function that leverages supervised data to capture complex dependencies between tokens and probabilities, thereby producing more reliable and calibrated response scores in computing the uncertainty of LLM generations. Our comprehensive experiments across question-answering and arithmetical reasoning tasks with various datasets demonstrate that LARS significantly outperforms existing scoring functions, achieving improvements of up to 16\% AUROC score.
♻ ☆ Unveiling Entity-Level Unlearning for Large Language Models: A Comprehensive Analysis
Large language model unlearning has garnered increasing attention due to its potential to address security and privacy concerns, leading to extensive research in the field. However, much of this research has concentrated on instance-level unlearning, specifically targeting the removal of predefined instances containing sensitive content. This focus has left a significant gap in the exploration of full entity-level unlearning, which is critical in real-world scenarios such as copyright protection. To this end, we propose a novel task of Entity-level unlearning, which aims to erase entity-related knowledge from the target model completely. To thoroughly investigate this task, we systematically evaluate trending unlearning algorithms, revealing that current methods struggle to achieve effective entity-level unlearning. Then, we further explore the factors that influence the performance of the unlearning algorithms, identifying that knowledge coverage and the size of the forget set play pivotal roles. Notably, our analysis also uncovers that entities introduced through fine-tuning are more vulnerable to unlearning than pre-trained entities. These findings collectively offer valuable insights for advancing entity-level unlearning for LLMs.
comment: Work in progress
♻ ☆ "We Demand Justice!": Towards Social Context Grounding of Political Texts EMNLP 2024
Social media discourse frequently consists of 'seemingly similar language used by opposing sides of the political spectrum', often translating to starkly contrasting perspectives. E.g., 'thoughts and prayers', could express sympathy for mass-shooting victims, or criticize the lack of legislative action on the issue. This paper defines the context required to fully understand such ambiguous statements in a computational setting and ground them in real-world entities, actions, and attitudes. We propose two challenging datasets that require an understanding of the real-world context of the text. We benchmark these datasets against models built upon large pre-trained models, such as RoBERTa and GPT-3. Additionally, we develop and benchmark more structured models building upon existing Discourse Contextualization Framework and Political Actor Representation models. We analyze the datasets and the predictions to obtain further insights into the pragmatic language understanding challenges posed by the proposed social grounding tasks.
comment: Accepted as an oral at EMNLP 2024 main conference
♻ ☆ Open Domain Question Answering with Conflicting Contexts
Open domain question answering systems frequently rely on information retrieved from large collections of text (such as the Web) to answer questions. However, such collections of text often contain conflicting information, and indiscriminately depending on this information may result in untruthful and inaccurate answers. To understand the gravity of this problem, we collect a human-annotated dataset, Question Answering with Conflicting Contexts (QACC), and find that as much as 25% of unambiguous, open domain questions can lead to conflicting contexts when retrieved using Google Search. We evaluate and benchmark three powerful Large Language Models (LLMs) with our dataset QACC and demonstrate their limitations in effectively addressing questions with conflicting information. To explore how humans reason through conflicting contexts, we request our annotators to provide explanations for their selections of correct answers. We demonstrate that by finetuning LLMs to explain their answers, we can introduce richer information into their training that guide them through the process of reasoning with conflicting contexts.
Computer Vision and Pattern Recognition 108
☆ BiGR: Harnessing Binary Latent Codes for Image Generation and Improved Visual Representation Capabilities
We introduce BiGR, a novel conditional image generation model using compact binary latent codes for generative training, focusing on enhancing both generation and representation capabilities. BiGR is the first conditional generative model that unifies generation and discrimination within the same framework. BiGR features a binary tokenizer, a masked modeling mechanism, and a binary transcoder for binary code prediction. Additionally, we introduce a novel entropy-ordered sampling method to enable efficient image generation. Extensive experiments validate BiGR's superior performance in generation quality, as measured by FID-50k, and representation capabilities, as evidenced by linear-probe accuracy. Moreover, BiGR showcases zero-shot generalization across various vision tasks, enabling applications such as image inpainting, outpainting, editing, interpolation, and enrichment, without the need for structural modifications. Our findings suggest that BiGR unifies generative and discriminative tasks effectively, paving the way for further advancements in the field.
comment: Project page: https://haoosz.github.io/BiGR
☆ NaturalBench: Evaluating Vision-Language Models on Natural Adversarial Samples NeurIPS 24
Vision-language models (VLMs) have made significant progress in recent visual-question-answering (VQA) benchmarks that evaluate complex visio-linguistic reasoning. However, are these models truly effective? In this work, we show that VLMs still struggle with natural images and questions that humans can easily answer, which we term natural adversarial samples. We also find it surprisingly easy to generate these VQA samples from natural image-text corpora using off-the-shelf models like CLIP and ChatGPT. We propose a semi-automated approach to collect a new benchmark, NaturalBench, for reliably evaluating VLMs with 10,000 human-verified VQA samples. Crucially, we adopt a $\textbf{vision-centric}$ design by pairing each question with two images that yield different answers, preventing blind solutions from answering without using the images. This makes NaturalBench more challenging than previous benchmarks that can be solved with commonsense priors. We evaluate 53 state-of-the-art VLMs on NaturalBench, showing that models like LLaVA-OneVision, Cambrian-1, Llama3.2-Vision, Molmo, Qwen2-VL, and even GPT-4o lag 50%-70% behind human performance (over 90%). We analyze why NaturalBench is hard from two angles: (1) Compositionality: Solving NaturalBench requires diverse visio-linguistic skills, including understanding attribute bindings, object relationships, and advanced reasoning like logic and counting. To this end, unlike prior work that uses a single tag per sample, we tag each NaturalBench sample with 1 to 8 skill tags for fine-grained evaluation. (2) Biases: NaturalBench exposes severe biases in VLMs, as models often choose the same answer regardless of the image. Lastly, we apply our benchmark curation method to diverse data sources, including long captions (over 100 words) and non-English languages like Chinese and Hindi, highlighting its potential for dynamic evaluations of VLMs.
comment: Accepted to NeurIPS 24; We open-source our dataset at: https://huggingface.co/datasets/BaiqiL/NaturalBench; Project page at: https://linzhiqiu.github.io/papers/naturalbench/
☆ Parallel Backpropagation for Inverse of a Convolution with Application to Normalizing Flows
Inverse of an invertible convolution is an important operation that comes up in Normalizing Flows, Image Deblurring, etc. The naive algorithm for backpropagation of this operation using Gaussian elimination has running time $O(n^3)$ where $n$ is the number of pixels in the image. We give a fast parallel backpropagation algorithm with running time $O(\sqrt{n})$ for a square image and provide a GPU implementation of the same. Inverse Convolutions are usually used in Normalizing Flows in the sampling pass, making them slow. We propose to use Inverse Convolutions in the forward (image to latent vector) pass of the Normalizing flow. Since the sampling pass is the inverse of the forward pass, it will use convolutions only, resulting in efficient sampling times. We use our parallel backpropagation algorithm for optimizing the inverse convolution layer resulting in fast training times also. We implement this approach in various Normalizing Flow backbones, resulting in our Inverse-Flow models. We benchmark Inverse-Flow on standard datasets and show significantly improved sampling times with similar bits per dimension compared to previous models.
comment: Preprint
☆ Swiss Army Knife: Synergizing Biases in Knowledge from Vision Foundation Models for Multi-Task Learning
Vision Foundation Models (VFMs) have demonstrated outstanding performance on numerous downstream tasks. However, due to their inherent representation biases originating from different training paradigms, VFMs exhibit advantages and disadvantages across distinct vision tasks. Although amalgamating the strengths of multiple VFMs for downstream tasks is an intuitive strategy, effectively exploiting these biases remains a significant challenge. In this paper, we propose a novel and versatile "Swiss Army Knife" (SAK) solution, which adaptively distills knowledge from a committee of VFMs to enhance multi-task learning. Unlike existing methods that use a single backbone for knowledge transfer, our approach preserves the unique representation bias of each teacher by collaborating the lightweight Teacher-Specific Adapter Path modules with the Teacher-Agnostic Stem. Through dynamic selection and combination of representations with Mixture-of-Representations Routers, our SAK is capable of synergizing the complementary strengths of multiple VFMs. Extensive experiments show that our SAK remarkably outperforms prior state of the arts in multi-task learning by 10% on the NYUD-v2 benchmark, while also providing a flexible and robust framework that can readily accommodate more advanced model designs.
☆ MultiOrg: A Multi-rater Organoid-detection Dataset
High-throughput image analysis in the biomedical domain has gained significant attention in recent years, driving advancements in drug discovery, disease prediction, and personalized medicine. Organoids, specifically, are an active area of research, providing excellent models for human organs and their functions. Automating the quantification of organoids in microscopy images would provide an effective solution to overcome substantial manual quantification bottlenecks, particularly in high-throughput image analysis. However, there is a notable lack of open biomedical datasets, in contrast to other domains, such as autonomous driving, and, notably, only few of them have attempted to quantify annotation uncertainty. In this work, we present MultiOrg a comprehensive organoid dataset tailored for object detection tasks with uncertainty quantification. This dataset comprises over 400 high-resolution 2d microscopy images and curated annotations of more than 60,000 organoids. Most importantly, it includes three label sets for the test data, independently annotated by two experts at distinct time points. We additionally provide a benchmark for organoid detection, and make the best model available through an easily installable, interactive plugin for the popular image visualization tool Napari, to perform organoid quantification.
☆ DRACO-DehazeNet: An Efficient Image Dehazing Network Combining Detail Recovery and a Novel Contrastive Learning Paradigm
Image dehazing is crucial for clarifying images obscured by haze or fog, but current learning-based approaches is dependent on large volumes of training data and hence consumed significant computational power. Additionally, their performance is often inadequate under non-uniform or heavy haze. To address these challenges, we developed the Detail Recovery And Contrastive DehazeNet, which facilitates efficient and effective dehazing via a dense dilated inverted residual block and an attention-based detail recovery network that tailors enhancements to specific dehazed scene contexts. A major innovation is its ability to train effectively with limited data, achieved through a novel quadruplet loss-based contrastive dehazing paradigm. This approach distinctly separates hazy and clear image features while also distinguish lower-quality and higher-quality dehazed images obtained from each sub-modules of our network, thereby refining the dehazing process to a larger extent. Extensive tests on a variety of benchmarked haze datasets demonstrated the superiority of our approach. The code repository for this work will be available soon.
comment: Submitted to a journal and currently under review. Once the paper is accepted and published, the copyright will be transferred to the corresponding journal
☆ MomentumSMoE: Integrating Momentum into Sparse Mixture of Experts NeurIPS 2024
Sparse Mixture of Experts (SMoE) has become the key to unlocking unparalleled scalability in deep learning. SMoE has the potential to exponentially increase parameter count while maintaining the efficiency of the model by only activating a small subset of these parameters for a given sample. However, it has been observed that SMoE suffers from unstable training and has difficulty adapting to new distributions, leading to the model's lack of robustness to data contamination. To overcome these limitations, we first establish a connection between the dynamics of the expert representations in SMoEs and gradient descent on a multi-objective optimization problem. Leveraging our framework, we then integrate momentum into SMoE and propose a new family of SMoEs named MomentumSMoE. We theoretically prove and numerically demonstrate that MomentumSMoE is more stable and robust than SMoE. In particular, we verify the advantages of MomentumSMoE over SMoE on a variety of practical tasks including ImageNet-1K object recognition and WikiText-103 language modeling. We demonstrate the applicability of MomentumSMoE to many types of SMoE models, including those in the Sparse MoE model for vision (V-MoE) and the Generalist Language Model (GLaM). We also show that other advanced momentum-based optimization methods, such as Adam, can be easily incorporated into the MomentumSMoE framework for designing new SMoE models with even better performance, almost negligible additional computation cost, and simple implementations.
comment: 10 pages in the main text. Published at NeurIPS 2024. The code is available at https://github.com/rachtsy/MomentumSMoE
☆ Multi-modal Pose Diffuser: A Multimodal Generative Conditional Pose Prior
The Skinned Multi-Person Linear (SMPL) model plays a crucial role in 3D human pose estimation, providing a streamlined yet effective representation of the human body. However, ensuring the validity of SMPL configurations during tasks such as human mesh regression remains a significant challenge , highlighting the necessity for a robust human pose prior capable of discerning realistic human poses. To address this, we introduce MOPED: \underline{M}ulti-m\underline{O}dal \underline{P}os\underline{E} \underline{D}iffuser. MOPED is the first method to leverage a novel multi-modal conditional diffusion model as a prior for SMPL pose parameters. Our method offers powerful unconditional pose generation with the ability to condition on multi-modal inputs such as images and text. This capability enhances the applicability of our approach by incorporating additional context often overlooked in traditional pose priors. Extensive experiments across three distinct tasks-pose estimation, pose denoising, and pose completion-demonstrate that our multi-modal diffusion model-based prior significantly outperforms existing methods. These results indicate that our model captures a broader spectrum of plausible human poses.
☆ A Hybrid Feature Fusion Deep Learning Framework for Leukemia Cancer Detection in Microscopic Blood Sample Using Gated Recurrent Unit and Uncertainty Quantification
Acute lymphoblastic leukemia (ALL) is the most malignant form of leukemia and the most common cancer in adults and children. Traditionally, leukemia is diagnosed by analyzing blood and bone marrow smears under a microscope, with additional cytochemical tests for confirmation. However, these methods are expensive, time consuming, and highly dependent on expert knowledge. In recent years, deep learning, particularly Convolutional Neural Networks (CNNs), has provided advanced methods for classifying microscopic smear images, aiding in the detection of leukemic cells. These approaches are quick, cost effective, and not subject to human bias. However, most methods lack the ability to quantify uncertainty, which could lead to critical misdiagnoses. In this research, hybrid deep learning models (InceptionV3-GRU, EfficientNetB3-GRU, MobileNetV2-GRU) were implemented to classify ALL. Bayesian optimization was used to fine tune the model's hyperparameters and improve its performance. Additionally, Deep Ensemble uncertainty quantification was applied to address uncertainty during leukemia image classification. The proposed models were trained on the publicly available datasets ALL-IDB1 and ALL-IDB2. Their results were then aggregated at the score level using the sum rule. The parallel architecture used in these models offers a high level of confidence in differentiating between ALL and non-ALL cases. The proposed method achieved a remarkable detection accuracy rate of 100% on the ALL-IDB1 dataset, 98.07% on the ALL-IDB2 dataset, and 98.64% on the combined dataset, demonstrating its potential for accurate and reliable leukemia diagnosis.
☆ Less is More: Selective Reduction of CT Data for Self-Supervised Pre-Training of Deep Learning Models with Contrastive Learning Improves Downstream Classification Performance
Self-supervised pre-training of deep learning models with contrastive learning is a widely used technique in image analysis. Current findings indicate a strong potential for contrastive pre-training on medical images. However, further research is necessary to incorporate the particular characteristics of these images. We hypothesize that the similarity of medical images hinders the success of contrastive learning in the medical imaging domain. To this end, we investigate different strategies based on deep embedding, information theory, and hashing in order to identify and reduce redundancy in medical pre-training datasets. The effect of these different reduction strategies on contrastive learning is evaluated on two pre-training datasets and several downstream classification tasks. In all of our experiments, dataset reduction leads to a considerable performance gain in downstream tasks, e.g., an AUC score improvement from 0.78 to 0.83 for the COVID CT Classification Grand Challenge, 0.97 to 0.98 for the OrganSMNIST Classification Challenge and 0.73 to 0.83 for a brain hemorrhage classification task. Furthermore, pre-training is up to nine times faster due to the dataset reduction. In conclusion, the proposed approach highlights the importance of dataset quality and provides a transferable approach to improve contrastive pre-training for classification downstream tasks on medical images.
comment: Published in Computers in Biology and Medicine
☆ CLIP-VAD: Exploiting Vision-Language Models for Voice Activity Detection
Voice Activity Detection (VAD) is the process of automatically determining whether a person is speaking and identifying the timing of their speech in an audiovisual data. Traditionally, this task has been tackled by processing either audio signals or visual data, or by combining both modalities through fusion or joint learning. In our study, drawing inspiration from recent advancements in visual-language models, we introduce a novel approach leveraging Contrastive Language-Image Pretraining (CLIP) models. The CLIP visual encoder analyzes video segments composed of the upper body of an individual, while the text encoder handles textual descriptions automatically generated through prompt engineering. Subsequently, embeddings from these encoders are fused through a deep neural network to perform VAD. Our experimental analysis across three VAD benchmarks showcases the superior performance of our method compared to existing visual VAD approaches. Notably, our approach outperforms several audio-visual methods despite its simplicity, and without requiring pre-training on extensive audio-visual datasets.
☆ LEAD: Latent Realignment for Human Motion Diffusion
Our goal is to generate realistic human motion from natural language. Modern methods often face a trade-off between model expressiveness and text-to-motion alignment. Some align text and motion latent spaces but sacrifice expressiveness; others rely on diffusion models producing impressive motions, but lacking semantic meaning in their latent space. This may compromise realism, diversity, and applicability. Here, we address this by combining latent diffusion with a realignment mechanism, producing a novel, semantically structured space that encodes the semantics of language. Leveraging this capability, we introduce the task of textual motion inversion to capture novel motion concepts from a few examples. For motion synthesis, we evaluate LEAD on HumanML3D and KIT-ML and show comparable performance to the state-of-the-art in terms of realism, diversity, and text-motion consistency. Our qualitative analysis and user study reveal that our synthesized motions are sharper, more human-like and comply better with the text compared to modern methods. For motion textual inversion, our method demonstrates improved capacity in capturing out-of-distribution characteristics in comparison to traditional VAEs.
☆ Neural Real-Time Recalibration for Infrared Multi-Camera Systems
Currently, there are no learning-free or neural techniques for real-time recalibration of infrared multi-camera systems. In this paper, we address the challenge of real-time, highly-accurate calibration of multi-camera infrared systems, a critical task for time-sensitive applications. Unlike traditional calibration techniques that lack adaptability and struggle with on-the-fly recalibrations, we propose a neural network-based method capable of dynamic real-time calibration. The proposed method integrates a differentiable projection model that directly correlates 3D geometries with their 2D image projections and facilitates the direct optimization of both intrinsic and extrinsic camera parameters. Key to our approach is the dynamic camera pose synthesis with perturbations in camera parameters, emulating realistic operational challenges to enhance model robustness. We introduce two model variants: one designed for multi-camera systems with onboard processing of 2D points, utilizing the direct 2D projections of 3D fiducials, and another for image-based systems, employing color-coded projected points for implicitly establishing correspondence. Through rigorous experimentation, we demonstrate our method is more accurate than traditional calibration techniques with or without perturbations while also being real-time, marking a significant leap in the field of real-time multi-camera system calibration. The source code can be found at https://github.com/theICTlab/neural-recalibration
comment: real-time camera calibration, infrared camera, neural calibration
☆ An Integrated Deep Learning Model for Skin Cancer Detection Using Hybrid Feature Fusion Technique
Skin cancer is a serious and potentially fatal disease caused by DNA damage. Early detection significantly increases survival rates, making accurate diagnosis crucial. In this groundbreaking study, we present a hybrid framework based on Deep Learning (DL) that achieves precise classification of benign and malignant skin lesions. Our approach begins with dataset preprocessing to enhance classification accuracy, followed by training two separate pre-trained DL models, InceptionV3 and DenseNet121. By fusing the results of each model using the weighted sum rule, our system achieves exceptional accuracy rates. Specifically, we achieve a 92.27% detection accuracy rate, 92.33% sensitivity, 92.22% specificity, 90.81% precision, and 91.57% F1-score, outperforming existing models and demonstrating the robustness and trustworthiness of our hybrid approach. Our study represents a significant advance in skin cancer diagnosis and provides a promising foundation for further research in the field. With the potential to save countless lives through earlier detection, our hybrid deep-learning approach is a game-changer in the fight against skin cancer.
☆ How Do Training Methods Influence the Utilization of Vision Models? NeurIPS 2024
Not all learnable parameters (e.g., weights) contribute equally to a neural network's decision function. In fact, entire layers' parameters can sometimes be reset to random values with little to no impact on the model's decisions. We revisit earlier studies that examined how architecture and task complexity influence this phenomenon and ask: is this phenomenon also affected by how we train the model? We conducted experimental evaluations on a diverse set of ImageNet-1k classification models to explore this, keeping the architecture and training data constant but varying the training pipeline. Our findings reveal that the training method strongly influences which layers become critical to the decision function for a given task. For example, improved training regimes and self-supervised training increase the importance of early layers while significantly under-utilizing deeper layers. In contrast, methods such as adversarial training display an opposite trend. Our preliminary results extend previous findings, offering a more nuanced understanding of the inner mechanics of neural networks. Code: https://github.com/paulgavrikov/layer_criticality
comment: Accepted at the Interpretable AI: Past, Present and Future Workshop at NeurIPS 2024
☆ LUDVIG: Learning-free Uplifting of 2D Visual features to Gaussian Splatting scenes
We address the task of uplifting visual features or semantic masks from 2D vision models to 3D scenes represented by Gaussian Splatting. Whereas common approaches rely on iterative optimization-based procedures, we show that a simple yet effective aggregation technique yields excellent results. Applied to semantic masks from Segment Anything (SAM), our uplifting approach leads to segmentation quality comparable to the state of the art. We then extend this method to generic DINOv2 features, integrating 3D scene geometry through graph diffusion, and achieve competitive segmentation results despite DINOv2 not being trained on millions of annotated masks like SAM.
☆ Toward Generalizing Visual Brain Decoding to Unseen Subjects
Visual brain decoding aims to decode visual information from human brain activities. Despite the great progress, one critical limitation of current brain decoding research lies in the lack of generalization capability to unseen subjects. Prior works typically focus on decoding brain activity of individuals based on the observation that different subjects exhibit different brain activities, while it remains unclear whether brain decoding can be generalized to unseen subjects. This study aims to answer this question. We first consolidate an image-fMRI dataset consisting of stimulus-image and fMRI-response pairs, involving 177 subjects in the movie-viewing task of the Human Connectome Project (HCP). This dataset allows us to investigate the brain decoding performance with the increase of participants. We then present a learning paradigm that applies uniform processing across all subjects, instead of employing different network heads or tokenizers for individuals as in previous methods, which can accommodate a large number of subjects to explore the generalization capability across different subjects. A series of experiments are conducted and we have the following findings. First, the network exhibits clear generalization capabilities with the increase of training subjects. Second, the generalization capability is common to popular network architectures (MLP, CNN and Transformer). Third, the generalization performance is affected by the similarity between subjects. Our findings reveal the inherent similarities in brain activities across individuals. With the emerging of larger and more comprehensive datasets, it is possible to train a brain decoding foundation model in the future.Codes and models can be found at https://github.com/Xiangtaokong/TGBD.
☆ FashionR2R: Texture-preserving Rendered-to-Real Image Translation with Diffusion Models NeurIPS 2024
Modeling and producing lifelike clothed human images has attracted researchers' attention from different areas for decades, with the complexity from highly articulated and structured content. Rendering algorithms decompose and simulate the imaging process of a camera, while are limited by the accuracy of modeled variables and the efficiency of computation. Generative models can produce impressively vivid human images, however still lacking in controllability and editability. This paper studies photorealism enhancement of rendered images, leveraging generative power from diffusion models on the controlled basis of rendering. We introduce a novel framework to translate rendered images into their realistic counterparts, which consists of two stages: Domain Knowledge Injection (DKI) and Realistic Image Generation (RIG). In DKI, we adopt positive (real) domain finetuning and negative (rendered) domain embedding to inject knowledge into a pretrained Text-to-image (T2I) diffusion model. In RIG, we generate the realistic image corresponding to the input rendered image, with a Texture-preserving Attention Control (TAC) to preserve fine-grained clothing textures, exploiting the decoupled features encoded in the UNet structure. Additionally, we introduce SynFashion dataset, featuring high-quality digital clothing images with diverse textures. Extensive experimental results demonstrate the superiority and effectiveness of our method in rendered-to-real image translation.
comment: Accepted by NeurIPS 2024
☆ Integrating Deep Learning with Fundus and Optical Coherence Tomography for Cardiovascular Disease Prediction
Early identification of patients at risk of cardiovascular diseases (CVD) is crucial for effective preventive care, reducing healthcare burden, and improving patients' quality of life. This study demonstrates the potential of retinal optical coherence tomography (OCT) imaging combined with fundus photographs for identifying future adverse cardiac events. We used data from 977 patients who experienced CVD within a 5-year interval post-image acquisition, alongside 1,877 control participants without CVD, totaling 2,854 subjects. We propose a novel binary classification network based on a Multi-channel Variational Autoencoder (MCVAE), which learns a latent embedding of patients' fundus and OCT images to classify individuals into two groups: those likely to develop CVD in the future and those who are not. Our model, trained on both imaging modalities, achieved promising results (AUROC 0.78 +/- 0.02, accuracy 0.68 +/- 0.002, precision 0.74 +/- 0.02, sensitivity 0.73 +/- 0.02, and specificity 0.68 +/- 0.01), demonstrating its efficacy in identifying patients at risk of future CVD events based on their retinal images. This study highlights the potential of retinal OCT imaging and fundus photographs as cost-effective, non-invasive alternatives for predicting cardiovascular disease risk. The widespread availability of these imaging techniques in optometry practices and hospitals further enhances their potential for large-scale CVD risk screening. Our findings contribute to the development of standardized, accessible methods for early CVD risk identification, potentially improving preventive care strategies and patient outcomes.
comment: Part of the book series: Lecture Notes in Computer Science ((LNCS,volume 15155))
☆ Variable Aperture Bokeh Rendering via Customized Focal Plane Guidance
Bokeh rendering is one of the most popular techniques in photography. It can make photographs visually appealing, forcing users to focus their attentions on particular area of image. However, achieving satisfactory bokeh effect usually presents significant challenge, since mobile cameras with restricted optical systems are constrained, while expensive high-end DSLR lens with large aperture should be needed. Therefore, many deep learning-based computational photography methods have been developed to mimic the bokeh effect in recent years. Nevertheless, most of these methods were limited to rendering bokeh effect in certain single aperture. There lacks user-friendly bokeh rendering method that can provide precise focal plane control and customised bokeh generation. There as well lacks authentic realistic bokeh dataset that can potentially promote bokeh learning on variable apertures. To address these two issues, in this paper, we have proposed an effective controllable bokeh rendering method, and contributed a Variable Aperture Bokeh Dataset (VABD). In the proposed method, user can customize focal plane to accurately locate concerned subjects and select target aperture information for bokeh rendering. Experimental results on public EBB! benchmark dataset and our constructed dataset VABD have demonstrated that the customized focal plane together aperture prompt can bootstrap model to simulate realistic bokeh effect. The proposed method has achieved competitive state-of-the-art performance with only 4.4M parameters, which is much lighter than mainstream computational bokeh models. The contributed dataset and source codes will be released on github https://github.com/MoTong-AI-studio/VABM.
☆ Dynamic Negative Guidance of Diffusion Models ICLR 2025
Negative Prompting (NP) is widely utilized in diffusion models, particularly in text-to-image applications, to prevent the generation of undesired features. In this paper, we show that conventional NP is limited by the assumption of a constant guidance scale, which may lead to highly suboptimal results, or even complete failure, due to the non-stationarity and state-dependence of the reverse process. Based on this analysis, we derive a principled technique called Dynamic Negative Guidance, which relies on a near-optimal time and state dependent modulation of the guidance without requiring additional training. Unlike NP, negative guidance requires estimating the posterior class probability during the denoising process, which is achieved with limited additional computational overhead by tracking the discrete Markov Chain during the generative process. We evaluate the performance of DNG class-removal on MNIST and CIFAR10, where we show that DNG leads to higher safety, preservation of class balance and image quality when compared with baseline methods. Furthermore, we show that it is possible to use DNG with Stable Diffusion to obtain more accurate and less invasive guidance than NP.
comment: Paper currently under review. Submitted to ICLR 2025
☆ SurgeryV2: Bridging the Gap Between Model Merging and Multi-Task Learning with Deep Representation Surgery ICML 2024
Model merging-based multitask learning (MTL) offers a promising approach for performing MTL by merging multiple expert models without requiring access to raw training data. However, in this paper, we examine the merged model's representation distribution and uncover a critical issue of "representation bias". This bias arises from a significant distribution gap between the representations of the merged and expert models, leading to the suboptimal performance of the merged MTL model. To address this challenge, we first propose a representation surgery solution called Surgery. Surgery is a lightweight, task-specific module that aligns the final layer representations of the merged model with those of the expert models, effectively alleviating bias and improving the merged model's performance. Despite these improvements, a performance gap remains compared to the traditional MTL method. Further analysis reveals that representation bias phenomena exist at each layer of the merged model, and aligning representations only in the last layer is insufficient for fully reducing systemic bias because biases introduced at each layer can accumulate and interact in complex ways. To tackle this, we then propose a more comprehensive solution, deep representation surgery (also called SurgeryV2), which mitigates representation bias across all layers, and thus bridges the performance gap between model merging-based MTL and traditional MTL. Finally, we design an unsupervised optimization objective to optimize both the Surgery and SurgeryV2 modules. Our experimental results show that incorporating these modules into state-of-the-art (SOTA) model merging schemes leads to significant performance gains. Notably, our SurgeryV2 scheme reaches almost the same level as individual expert models or the traditional MTL model. The code is available at \url{https://github.com/EnnengYang/SurgeryV2}.
comment: This paper is an extended version of our previous work [arXiv:2402.02705] presented at ICML 2024
☆ AnomalyNCD: Towards Novel Anomaly Class Discovery in Industrial Scenarios
In the industrial scenario, anomaly detection could locate but cannot classify anomalies. To complete their capability, we study to automatically discover and recognize visual classes of industrial anomalies. In terms of multi-class anomaly classification, previous methods cluster anomalies represented by frozen pre-trained models but often fail due to poor discrimination. Novel class discovery (NCD) has the potential to tackle this. However, it struggles with non-prominent and semantically weak anomalies that challenge network learning focus. To address these, we introduce AnomalyNCD, a multi-class anomaly classification framework compatible with existing anomaly detection methods. This framework learns anomaly-specific features and classifies anomalies in a self-supervised manner. Initially, a technique called Main Element Binarization (MEBin) is first designed, which segments primary anomaly regions into masks to alleviate the impact of incorrect detections on learning. Subsequently, we employ mask-guided contrastive representation learning to improve feature discrimination, which focuses network attention on isolated anomalous regions and reduces the confusion of erroneous inputs through re-corrected pseudo labels. Finally, to enable flexible classification at both region and image levels during inference, we develop a region merging strategy that determines the overall image category based on the classified anomaly regions. Our method outperforms the state-of-the-art works on the MVTec AD and MTD datasets. Compared with the current methods, AnomalyNCD combined with zero-shot anomaly detection method achieves a 10.8% $F_1$ gain, 8.8% NMI gain, and 9.5% ARI gain on MVTec AD, 12.8% $F_1$ gain, 5.7% NMI gain, and 10.8% ARI gain on MTD. The source code is available at https://github.com/HUST-SLOW/AnomalyNCD.
☆ Impact of imperfect annotations on CNN training and performance for instance segmentation and classification in digital pathology
Segmentation and classification of large numbers of instances, such as cell nuclei, are crucial tasks in digital pathology for accurate diagnosis. However, the availability of high-quality datasets for deep learning methods is often limited due to the complexity of the annotation process. In this work, we investigate the impact of noisy annotations on the training and performance of a state-of-the-art CNN model for the combined task of detecting, segmenting and classifying nuclei in histopathology images. In this context, we investigate the conditions for determining an appropriate number of training epochs to prevent overfitting to annotation noise during training. Our results indicate that the utilisation of a small, correctly annotated validation set is instrumental in avoiding overfitting and maintaining model performance to a large extent. Additionally, our findings underscore the beneficial role of pre-training.
☆ 2D-3D Deformable Image Registration of Histology Slide and Micro-CT with ML-based Initialization
Recent developments in the registration of histology and micro-computed tomography ({\mu}CT) have broadened the perspective of pathological applications such as virtual histology based on {\mu}CT. This topic remains challenging because of the low image quality of soft tissue CT. Additionally, soft tissue samples usually deform during the histology slide preparation, making it difficult to correlate the structures between histology slide and {\mu}CT. In this work, we propose a novel 2D-3D multi-modal deformable image registration method. The method uses a machine learning (ML) based initialization followed by the registration. The registration is finalized by an analytical out-of-plane deformation refinement. The method is evaluated on datasets acquired from tonsil and tumor tissues. {\mu}CTs of both phase-contrast and conventional absorption modalities are investigated. The registration results from the proposed method are compared with those from intensity- and keypoint-based methods. The comparison is conducted using both visual and fiducial-based evaluations. The proposed method demonstrates superior performance compared to the other two methods.
comment: 12 pages, 4 figures
☆ Zero-shot Action Localization via the Confidence of Large Vision-Language Models
Precise action localization in untrimmed video is vital for fields such as professional sports and minimally invasive surgery, where the delineation of particular motions in recordings can dramatically enhance analysis. But in many cases, large scale datasets with video-label pairs for localization are unavailable, limiting the opportunity to fine-tune video-understanding models. Recent developments in large vision-language models (LVLM) address this need with impressive zero-shot capabilities in a variety of video understanding tasks. However, the adaptation of image-based LVLMs, with their powerful visual question answering capabilities, to action localization in long-form video is still relatively unexplored. To this end, we introduce a true ZEro-shot Action Localization method (ZEAL). Specifically, we leverage the built-in action knowledge of a large language model (LLM) to inflate actions into highly-detailed descriptions of the archetypal start and end of the action. These descriptions serve as queries to LVLM for generating frame-level confidence scores which can be aggregated to produce localization outputs. The simplicity and flexibility of our method lends it amenable to more capable LVLMs as they are developed, and we demonstrate remarkable results in zero-shot action localization on a challenging benchmark, without any training.
☆ Evaluating the evaluators: Towards human-aligned metrics for missing markers reconstruction
Animation data is often obtained through optical motion capture systems, which utilize a multitude of cameras to establish the position of optical markers. However, system errors or occlusions can result in missing markers, the manual cleaning of which can be time-consuming. This has sparked interest in machine learning-based solutions for missing marker reconstruction in the academic community. Most academic papers utilize a simplistic mean square error as the main metric. In this paper, we show that this metric does not correlate with subjective perception of the fill quality. We introduce and evaluate a set of better-correlated metrics that can drive progress in the field.
☆ Croc: Pretraining Large Multimodal Models with Cross-Modal Comprehension
Recent advances in Large Language Models (LLMs) have catalyzed the development of Large Multimodal Models (LMMs). However, existing research primarily focuses on tuning language and image instructions, ignoring the critical pretraining phase where models learn to process textual and visual modalities jointly. In this paper, we propose a new pretraining paradigm for LMMs to enhance the visual comprehension capabilities of LLMs by introducing a novel cross-modal comprehension stage. Specifically, we design a dynamically learnable prompt token pool and employ the Hungarian algorithm to replace part of the original visual tokens with the most relevant prompt tokens. Then, we conceptualize visual tokens as analogous to a "foreign language" for the LLMs and propose a mixed attention mechanism with bidirectional visual attention and unidirectional textual attention to comprehensively enhance the understanding of visual tokens. Meanwhile, we integrate a detailed caption generation task, leveraging rich descriptions to further facilitate LLMs in understanding visual semantic information. After pretraining on 1.5 million publicly accessible data, we present a new foundation model called Croc. Experimental results demonstrate that Croc achieves new state-of-the-art performance on massive vision-language benchmarks. To support reproducibility and facilitate further research, we release the training code and pre-trained model weights at https://github.com/deepglint/Croc.
comment: 18 pages, 11 figures
☆ Fast proxy centers for Jeffreys centroids: The Jeffreys-Fisher-Rao and the inductive Gauss-Bregman centers
The symmetric Kullback-Leibler centroid also called the Jeffreys centroid of a set of mutually absolutely continuous probability distributions on a measure space provides a notion of centrality which has proven useful in many tasks including information retrieval, information fusion, and clustering in image, video and sound processing. However, the Jeffreys centroid is not available in closed-form for sets of categorical or normal distributions, two widely used statistical models, and thus need to be approximated numerically in practice. In this paper, we first propose the new Jeffreys-Fisher-Rao center defined as the Fisher-Rao midpoint of the sided Kullback-Leibler centroids as a plug-in replacement of the Jeffreys centroid. This Jeffreys-Fisher-Rao center admits a generic formula for uni-parameter exponential family distributions, and closed-form formula for categorical and normal distributions, matches exactly the Jeffreys centroid for same-mean normal distributions, and is experimentally observed in practice to be close to the Jeffreys centroid. Second, we define a new type of inductive centers generalizing the principle of Gauss arithmetic-geometric double sequence mean for pairs of densities of any given exponential family. This center is shown experimentally to approximate very well the Jeffreys centroid and is suggested to use when the Jeffreys-Fisher-Rao center is not available in closed form. Moreover, this Gauss-Bregman inductive center always converges and matches the Jeffreys centroid for sets of same-mean normal distributions. We report on our experiments demonstrating the use of the Jeffreys-Fisher-Rao and Gauss-Bregman centers instead of the Jeffreys centroid. Finally, we conclude this work by reinterpreting these fast proxy centers of Jeffreys centroids under the lens of dually flat spaces in information geometry.
comment: 35 pages, 10 figures
☆ HiCo: Hierarchical Controllable Diffusion Model for Layout-to-image Generation NeurIPS2024
The task of layout-to-image generation involves synthesizing images based on the captions of objects and their spatial positions. Existing methods still struggle in complex layout generation, where common bad cases include object missing, inconsistent lighting, conflicting view angles, etc. To effectively address these issues, we propose a \textbf{Hi}erarchical \textbf{Co}ntrollable (HiCo) diffusion model for layout-to-image generation, featuring object seperable conditioning branch structure. Our key insight is to achieve spatial disentanglement through hierarchical modeling of layouts. We use a multi branch structure to represent hierarchy and aggregate them in fusion module. To evaluate the performance of multi-objective controllable layout generation in natural scenes, we introduce the HiCo-7K benchmark, derived from the GRIT-20M dataset and manually cleaned. https://github.com/360CVGroup/HiCo_T2I.
comment: NeurIPS2024
☆ Advanced Underwater Image Quality Enhancement via Hybrid Super-Resolution Convolutional Neural Networks and Multi-Scale Retinex-Based Defogging Techniques
The difficulties of underwater image degradation due to light scattering, absorption, and fog-like particles which lead to low resolution and poor visibility are discussed in this study report. We suggest a sophisticated hybrid strategy that combines Multi-Scale Retinex (MSR) defogging methods with Super-Resolution Convolutional Neural Networks (SRCNN) to address these problems. The Retinex algorithm mimics human visual perception to reduce uneven lighting and fogging, while the SRCNN component improves the spatial resolution of underwater photos.Through the combination of these methods, we are able to enhance the clarity, contrast, and colour restoration of underwater images, offering a reliable way to improve image quality in difficult underwater conditions. The research conducts extensive experiments on real-world underwater datasets to further illustrate the efficacy of the suggested approach. In terms of sharpness, visibility, and feature retention, quantitative evaluation which use metrics like the Structural Similarity Index Measure (SSIM) and Peak Signal-to-Noise Ratio (PSNR) demonstrates notable advances over conventional techniques.In real-time underwater applications like marine exploration, underwater robotics, and autonomous underwater vehicles, where clear and high-resolution imaging is crucial for operational success, the combination of deep learning and conventional image processing techniques offers a computationally efficient framework with superior results.
☆ Takin-ADA: Emotion Controllable Audio-Driven Animation with Canonical and Landmark Loss Optimization
Existing audio-driven facial animation methods face critical challenges, including expression leakage, ineffective subtle expression transfer, and imprecise audio-driven synchronization. We discovered that these issues stem from limitations in motion representation and the lack of fine-grained control over facial expressions. To address these problems, we present Takin-ADA, a novel two-stage approach for real-time audio-driven portrait animation. In the first stage, we introduce a specialized loss function that enhances subtle expression transfer while reducing unwanted expression leakage. The second stage utilizes an advanced audio processing technique to improve lip-sync accuracy. Our method not only generates precise lip movements but also allows flexible control over facial expressions and head motions. Takin-ADA achieves high-resolution (512x512) facial animations at up to 42 FPS on an RTX 4090 GPU, outperforming existing commercial solutions. Extensive experiments demonstrate that our model significantly surpasses previous methods in video quality, facial dynamics realism, and natural head movements, setting a new benchmark in the field of audio-driven facial animation.
comment: under review
☆ You Only Look Twice! for Failure Causes Identification of Drill Bits
Efficient identification of the root causes of drill bit failure is crucial due to potential impacts such as operational losses, safety threats, and delays. Early recognition of these failures enables proactive maintenance, reducing risks and financial losses associated with unforeseen breakdowns and prolonged downtime. Thus, our study investigates various causes of drill bit failure using images of different blades. The process involves annotating cutters with their respective locations and damage types, followed by the development of two YOLO Location and Damage Cutter Detection models, as well as multi-class multi-label Decision Tree and Random Forests models to identify the causes of failure by assessing the cutters' location and damage type. Additionally, RRFCI is proposed for the classification of failure causes. Notably, the cutter location detection model achieved a high score of 0.97 mPA, and the cutter damage detection model yielded a 0.49 mPA. The rule-based approach over-performed both DT and RF in failure cause identification, achieving a macro-average F1-score of 0.94 across all damage causes. The integration of the complete automated pipeline successfully identified 100\% of the 24 failure causes when tested on independent sets of ten drill bits, showcasing its potential to efficiently assist experts in identifying the root causes of drill bit damages.
☆ ClearSR: Latent Low-Resolution Image Embeddings Help Diffusion-Based Real-World Super Resolution Models See Clearer
We present ClearSR, a new method that can better take advantage of latent low-resolution image (LR) embeddings for diffusion-based real-world image super-resolution (Real-ISR). Previous Real-ISR models mostly focus on how to activate more generative priors of text-to-image diffusion models to make the output high-resolution (HR) images look better. However, since these methods rely too much on the generative priors, the content of the output images is often inconsistent with the input LR ones. To mitigate the above issue, in this work, we explore using latent LR embeddings to constrain the control signals from ControlNet, and extract LR information at both detail and structure levels. We show that the proper use of latent LR embeddings can produce higher-quality control signals, which enables the super-resolution results to be more consistent with the LR image and leads to clearer visual results. In addition, we also show that latent LR embeddings can be used to control the inference stage, allowing for the improvement of fidelity and generation ability simultaneously. Experiments demonstrate that our model can achieve better performance across multiple metrics on several test sets and generate more consistent SR results with LR images than existing methods. Our code will be made publicly available.
☆ HYPNOS : Highly Precise Foreground-focused Diffusion Finetuning for Inanimate Objects ACCV
In recent years, personalized diffusion-based text-to-image generative tasks have been a hot topic in computer vision studies. A robust diffusion model is determined by its ability to perform near-perfect reconstruction of certain product outcomes given few related input samples. Unfortunately, the current prominent diffusion-based finetuning technique falls short in maintaining the foreground object consistency while being constrained to produce diverse backgrounds in the image outcome. In the worst scenario, the overfitting issue may occur, meaning that the foreground object is less controllable due to the condition above, for example, the input prompt information is transferred ambiguously to both foreground and background regions, instead of the supposed background region only. To tackle the issues above, we proposed Hypnos, a highly precise foreground-focused diffusion finetuning technique. On the image level, this strategy works best for inanimate object generation tasks, and to do so, Hypnos implements two main approaches, namely: (i) a content-centric prompting strategy and (ii) the utilization of our additional foreground-focused discriminative module. The utilized module is connected with the diffusion model and finetuned with our proposed set of supervision mechanism. Combining the strategies above yielded to the foreground-background disentanglement capability of the diffusion model. Our experimental results showed that the proposed strategy gave a more robust performance and visually pleasing results compared to the former technique. For better elaborations, we also provided extensive studies to assess the fruitful outcomes above, which reveal how personalization behaves in regard to several training conditions.
comment: 26 pages, 12 figures, to appear on the Rich Media with Generative AI workshop in conjunction with Asian Conference on Computer Vision (ACCV) 2024
☆ Vision-Language Navigation with Energy-Based Policy
Vision-language navigation (VLN) requires an agent to execute actions following human instructions. Existing VLN models are optimized through expert demonstrations by supervised behavioural cloning or incorporating manual reward engineering. While straightforward, these efforts overlook the accumulation of errors in the Markov decision process, and struggle to match the distribution of the expert policy. Going beyond this, we propose an Energy-based Navigation Policy (ENP) to model the joint state-action distribution using an energy-based model. At each step, low energy values correspond to the state-action pairs that the expert is most likely to perform, and vice versa. Theoretically, the optimization objective is equivalent to minimizing the forward divergence between the occupancy measure of the expert and ours. Consequently, ENP learns to globally align with the expert policy by maximizing the likelihood of the actions and modeling the dynamics of the navigation states in a collaborative manner. With a variety of VLN architectures, ENP achieves promising performances on R2R, REVERIE, RxR, and R2R-CE, unleashing the power of existing VLN models.
☆ ERDDCI: Exact Reversible Diffusion via Dual-Chain Inversion for High-Quality Image Editing
Diffusion models (DMs) have been successfully applied to real image editing. These models typically invert images into latent noise vectors used to reconstruct the original images (known as inversion), and then edit them during the inference process. However, recent popular DMs often rely on the assumption of local linearization, where the noise injected during the inversion process is expected to approximate the noise removed during the inference process. While DM efficiently generates images under this assumption, it can also accumulate errors during the diffusion process due to the assumption, ultimately negatively impacting the quality of real image reconstruction and editing. To address this issue, we propose a novel method, referred to as ERDDCI (Exact Reversible Diffusion via Dual-Chain Inversion). ERDDCI uses the new Dual-Chain Inversion (DCI) for joint inference to derive an exact reversible diffusion process. By using DCI, our method effectively avoids the cumbersome optimization process in existing inversion approaches and achieves high-quality image editing. Additionally, to accommodate image operations under high guidance scales, we introduce a dynamic control strategy that enables more refined image reconstruction and editing. Our experiments demonstrate that ERDDCI significantly outperforms state-of-the-art methods in a 50-step diffusion process. It achieves rapid and precise image reconstruction with an SSIM of 0.999 and an LPIPS of 0.001, and also delivers competitive results in image editing.
☆ PReP: Efficient context-based shape retrieval for missing parts
In this paper we study the problem of shape part retrieval in the point cloud domain. Shape retrieval methods in the literature rely on the presence of an existing query object, but what if the part we are looking for is not available? We present Part Retrieval Pipeline (PReP), a pipeline that creatively utilizes metric learning techniques along with a trained classification model to measure the suitability of potential replacement parts from a database, as part of an application scenario targeting circular economy. Through an innovative training procedure with increasing difficulty, it is able to learn to recognize suitable parts relying only on shape context. Thanks to its low parameter size and computational requirements, it can be used to sort through a warehouse of potentially tens of thousand of spare parts in just a few seconds. We also establish an alternative baseline approach to compare against, and extensively document the unique challenges associated with this task, as well as identify the design choices to solve them.
☆ Pseudo-label Refinement for Improving Self-Supervised Learning Systems
Self-supervised learning systems have gained significant attention in recent years by leveraging clustering-based pseudo-labels to provide supervision without the need for human annotations. However, the noise in these pseudo-labels caused by the clustering methods poses a challenge to the learning process leading to degraded performance. In this work, we propose a pseudo-label refinement (SLR) algorithm to address this issue. The cluster labels from the previous epoch are projected to the current epoch cluster-labels space and a linear combination of the new label and the projected label is computed as a soft refined label containing the information from the previous epoch clusters as well as from the current epoch. In contrast to the common practice of using the maximum value as a cluster/class indicator, we employ hierarchical clustering on these soft pseudo-labels to generate refined hard-labels. This approach better utilizes the information embedded in the soft labels, outperforming the simple maximum value approach for hard label generation. The effectiveness of the proposed SLR algorithm is evaluated in the context of person re-identification (Re-ID) using unsupervised domain adaptation (UDA). Experimental results demonstrate that the modified Re-ID baseline, incorporating the SLR algorithm, achieves significantly improved mean Average Precision (mAP) performance in various UDA tasks, including real-to-synthetic, synthetic-to-real, and different real-to-real scenarios. These findings highlight the efficacy of the SLR algorithm in enhancing the performance of self-supervised learning systems.
☆ Storyboard guided Alignment for Fine-grained Video Action Recognition
Fine-grained video action recognition can be conceptualized as a video-text matching problem. Previous approaches often rely on global video semantics to consolidate video embeddings, which can lead to misalignment in video-text pairs due to a lack of understanding of action semantics at an atomic granularity level. To tackle this challenge, we propose a multi-granularity framework based on two observations: (i) videos with different global semantics may share similar atomic actions or appearances, and (ii) atomic actions within a video can be momentary, slow, or even non-directly related to the global video semantics. Inspired by the concept of storyboarding, which disassembles a script into individual shots, we enhance global video semantics by generating fine-grained descriptions using a pre-trained large language model. These detailed descriptions capture common atomic actions depicted in videos. A filtering metric is proposed to select the descriptions that correspond to the atomic actions present in both the videos and the descriptions. By employing global semantics and fine-grained descriptions, we can identify key frames in videos and utilize them to aggregate embeddings, thereby making the embedding more accurate. Extensive experiments on various video action recognition datasets demonstrate superior performance of our proposed method in supervised, few-shot, and zero-shot settings.
☆ MambaSCI: Efficient Mamba-UNet for Quad-Bayer Patterned Video Snapshot Compressive Imaging NeurIPS 2024
Color video snapshot compressive imaging (SCI) employs computational imaging techniques to capture multiple sequential video frames in a single Bayer-patterned measurement. With the increasing popularity of quad-Bayer pattern in mainstream smartphone cameras for capturing high-resolution videos, mobile photography has become more accessible to a wider audience. However, existing color video SCI reconstruction algorithms are designed based on the traditional Bayer pattern. When applied to videos captured by quad-Bayer cameras, these algorithms often result in color distortion and ineffective demosaicing, rendering them impractical for primary equipment. To address this challenge, we propose the MambaSCI method, which leverages the Mamba and UNet architectures for efficient reconstruction of quad-Bayer patterned color video SCI. To the best of our knowledge, our work presents the first algorithm for quad-Bayer patterned SCI reconstruction, and also the initial application of the Mamba model to this task. Specifically, we customize Residual-Mamba-Blocks, which residually connect the Spatial-Temporal Mamba (STMamba), Edge-Detail-Reconstruction (EDR) module, and Channel Attention (CA) module. Respectively, STMamba is used to model long-range spatial-temporal dependencies with linear complexity, EDR is for better edge-detail reconstruction, and CA is used to compensate for the missing channel information interaction in Mamba model. Experiments demonstrate that MambaSCI surpasses state-of-the-art methods with lower computational and memory costs. PyTorch style pseudo-code for the core modules is provided in the supplementary materials.
comment: NeurIPS 2024
☆ Shape Transformation Driven by Active Contour for Class-Imbalanced Semi-Supervised Medical Image Segmentation
Annotating 3D medical images demands expert knowledge and is time-consuming. As a result, semi-supervised learning (SSL) approaches have gained significant interest in 3D medical image segmentation. The significant size differences among various organs in the human body lead to imbalanced class distribution, which is a major challenge in the real-world application of these SSL approaches. To address this issue, we develop a novel Shape Transformation driven by Active Contour (STAC), that enlarges smaller organs to alleviate imbalanced class distribution across different organs. Inspired by curve evolution theory in active contour methods, STAC employs a signed distance function (SDF) as the level set function, to implicitly represent the shape of organs, and deforms voxels in the direction of the steepest descent of SDF (i.e., the normal vector). To ensure that the voxels far from expansion organs remain unchanged, we design an SDF-based weight function to control the degree of deformation for each voxel. We then use STAC as a data-augmentation process during the training stage. Experimental results on two benchmark datasets demonstrate that the proposed method significantly outperforms some state-of-the-art methods. Source code is publicly available at https://github.com/GuGuLL123/STAC.
☆ Text-to-Image Representativity Fairness Evaluation Framework
Text-to-Image generative systems are progressing rapidly to be a source of advertisement and media and could soon serve as image searches or artists. However, there is a significant concern about the representativity bias these models embody and how these biases can propagate in the social fabric after fine-tuning them. Therefore, continuously monitoring and evaluating these models for fairness is important. To address this issue, we propose Text-to-Image (TTI) Representativity Fairness Evaluation Framework. In this framework, we evaluate three aspects of a TTI system; diversity, inclusion, and quality. For each aspect, human-based and model-based approaches are proposed and evaluated for their ability to capture the bias and whether they can substitute each other. The framework starts by suggesting the prompts for generating the images for the evaluation based on the context and the sensitive attributes under study. Then the three aspects are evaluated using the proposed approaches. Based on the evaluation, a decision is made regarding the representativity bias within the TTI system. The evaluation of our framework on Stable Diffusion shows that the framework can effectively capture the bias in TTI systems. The results also confirm that our proposed model based-approaches can substitute human-based approaches in three out of four components with high correlation, which could potentially reduce costs and automate the process. The study suggests that continual learning of the model on more inclusive data across disadvantaged minorities such as Indians and Middle Easterners is essential to mitigate current stereotyping and lack of inclusiveness.
☆ E3D-GPT: Enhanced 3D Visual Foundation for Medical Vision-Language Model
The development of 3D medical vision-language models holds significant potential for disease diagnosis and patient treatment. However, compared to 2D medical images, 3D medical images, such as CT scans, face challenges related to limited training data and high dimension, which severely restrict the progress of 3D medical vision-language models. To address these issues, we collect a large amount of unlabeled 3D CT data and utilize self-supervised learning to construct a 3D visual foundation model for extracting 3D visual features. Then, we apply 3D spatial convolutions to aggregate and project high-level image features, reducing computational complexity while preserving spatial information. We also construct two instruction-tuning datasets based on BIMCV-R and CT-RATE to fine-tune the 3D vision-language model. Our model demonstrates superior performance compared to existing methods in report generation, visual question answering, and disease diagnosis. Code and data will be made publicly available soon.
☆ Rethinking Transformer for Long Contextual Histopathology Whole Slide Image Analysis NeurIPS-2024
Histopathology Whole Slide Image (WSI) analysis serves as the gold standard for clinical cancer diagnosis in the daily routines of doctors. To develop computer-aided diagnosis model for WSIs, previous methods typically employ Multi-Instance Learning to enable slide-level prediction given only slide-level labels. Among these models, vanilla attention mechanisms without pairwise interactions have traditionally been employed but are unable to model contextual information. More recently, self-attention models have been utilized to address this issue. To alleviate the computational complexity of long sequences in large WSIs, methods like HIPT use region-slicing, and TransMIL employs approximation of full self-attention. Both approaches suffer from suboptimal performance due to the loss of key information. Moreover, their use of absolute positional embedding struggles to effectively handle long contextual dependencies in shape-varying WSIs. In this paper, we first analyze how the low-rank nature of the long-sequence attention matrix constrains the representation ability of WSI modelling. Then, we demonstrate that the rank of attention matrix can be improved by focusing on local interactions via a local attention mask. Our analysis shows that the local mask aligns with the attention patterns in the lower layers of the Transformer. Furthermore, the local attention mask can be implemented during chunked attention calculation, reducing the quadratic computational complexity to linear with a small local bandwidth. Building on this, we propose a local-global hybrid Transformer for both computational acceleration and local-global information interactions modelling. Our method, Long-contextual MIL (LongMIL), is evaluated through extensive experiments on various WSI tasks to validate its superiority. Our code will be available at github.com/invoker-LL/Long-MIL.
comment: NeurIPS-2024. arXiv admin note: text overlap with arXiv:2311.12885
☆ Neural Signed Distance Function Inference through Splatting 3D Gaussians Pulled on Zero-Level Set NeurIPS 2024
It is vital to infer a signed distance function (SDF) in multi-view based surface reconstruction. 3D Gaussian splatting (3DGS) provides a novel perspective for volume rendering, and shows advantages in rendering efficiency and quality. Although 3DGS provides a promising neural rendering option, it is still hard to infer SDFs for surface reconstruction with 3DGS due to the discreteness, the sparseness, and the off-surface drift of 3D Gaussians. To resolve these issues, we propose a method that seamlessly merge 3DGS with the learning of neural SDFs. Our key idea is to more effectively constrain the SDF inference with the multi-view consistency. To this end, we dynamically align 3D Gaussians on the zero-level set of the neural SDF using neural pulling, and then render the aligned 3D Gaussians through the differentiable rasterization. Meanwhile, we update the neural SDF by pulling neighboring space to the pulled 3D Gaussians, which progressively refine the signed distance field near the surface. With both differentiable pulling and splatting, we jointly optimize 3D Gaussians and the neural SDF with both RGB and geometry constraints, which recovers more accurate, smooth, and complete surfaces with more geometry details. Our numerical and visual comparisons show our superiority over the state-of-the-art results on the widely used benchmarks.
comment: Accepted by NeurIPS 2024. Project page: https://wen-yuan-zhang.github.io/GS-Pull/
☆ MultiChartQA: Benchmarking Vision-Language Models on Multi-Chart Problems
Multimodal Large Language Models (MLLMs) have demonstrated impressive abilities across various tasks, including visual question answering and chart comprehension, yet existing benchmarks for chart-related tasks fall short in capturing the complexity of real-world multi-chart scenarios. Current benchmarks primarily focus on single-chart tasks, neglecting the multi-hop reasoning required to extract and integrate information from multiple charts, which is essential in practical applications. To fill this gap, we introduce MultiChartQA, a benchmark that evaluates MLLMs' capabilities in four key areas: direct question answering, parallel question answering, comparative reasoning, and sequential reasoning. Our evaluation of a wide range of MLLMs reveals significant performance gaps compared to humans. These results highlight the challenges in multi-chart comprehension and the potential of MultiChartQA to drive advancements in this field. Our code and data are available at https://github.com/Zivenzhu/Multi-chart-QA
comment: 18 pages, 9 figures
☆ Feature Augmentation based Test-Time Adaptation
Test-time adaptation (TTA) allows a model to be adapted to an unseen domain without accessing the source data. Due to the nature of practical environments, TTA has a limited amount of data for adaptation. Recent TTA methods further restrict this by filtering input data for reliability, making the effective data size even smaller and limiting adaptation potential. To address this issue, We propose Feature Augmentation based Test-time Adaptation (FATA), a simple method that fully utilizes the limited amount of input data through feature augmentation. FATA employs Normalization Perturbation to augment features and adapts the model using the FATA loss, which makes the outputs of the augmented and original features similar. FATA is model-agnostic and can be seamlessly integrated into existing models without altering the model architecture. We demonstrate the effectiveness of FATA on various models and scenarios on ImageNet-C and Office-Home, validating its superiority in diverse real-world conditions.
comment: 10 pages
☆ Learning autonomous driving from aerial imagery IROS 2024
In this work, we consider the problem of learning end to end perception to control for ground vehicles solely from aerial imagery. Photogrammetric simulators allow the synthesis of novel views through the transformation of pre-generated assets into novel views.However, they have a large setup cost, require careful collection of data and often human effort to create usable simulators. We use a Neural Radiance Field (NeRF) as an intermediate representation to synthesize novel views from the point of view of a ground vehicle. These novel viewpoints can then be used for several downstream autonomous navigation applications. In this work, we demonstrate the utility of novel view synthesis though the application of training a policy for end to end learning from images and depth data. In a traditional real to sim to real framework, the collected data would be transformed into a visual simulator which could then be used to generate novel views. In contrast, using a NeRF allows a compact representation and the ability to optimize over the parameters of the visual simulator as more data is gathered in the environment. We demonstrate the efficacy of our method in a custom built mini-city environment through the deployment of imitation policies on robotic cars. We additionally consider the task of place localization and demonstrate that our method is able to relocalize the car in the real world.
comment: Presented at IROS 2024
☆ DaRePlane: Direction-aware Representations for Dynamic Scene Reconstruction
Numerous recent approaches to modeling and re-rendering dynamic scenes leverage plane-based explicit representations, addressing slow training times associated with models like neural radiance fields (NeRF) and Gaussian splatting (GS). However, merely decomposing 4D dynamic scenes into multiple 2D plane-based representations is insufficient for high-fidelity re-rendering of scenes with complex motions. In response, we present DaRePlane, a novel direction-aware representation approach that captures scene dynamics from six different directions. This learned representation undergoes an inverse dual-tree complex wavelet transformation (DTCWT) to recover plane-based information. Within NeRF pipelines, DaRePlane computes features for each space-time point by fusing vectors from these recovered planes, then passed to a tiny MLP for color regression. When applied to Gaussian splatting, DaRePlane computes the features of Gaussian points, followed by a tiny multi-head MLP for spatial-time deformation prediction. Notably, to address redundancy introduced by the six real and six imaginary direction-aware wavelet coefficients, we introduce a trainable masking approach, mitigating storage issues without significant performance decline. To demonstrate the generality and efficiency of DaRePlane, we test it on both regular and surgical dynamic scenes, for both NeRF and GS systems. Extensive experiments show that DaRePlane yields state-of-the-art performance in novel view synthesis for various complex dynamic scenes.
comment: arXiv admin note: substantial text overlap with arXiv:2403.02265
☆ Optimal DLT-based Solutions for the Perspective-n-Point
We propose a modified normalized direct linear transform (DLT) algorithm for solving the perspective-n-point (PnP) problem with much better behavior than the conventional DLT. The modification consists of analytically weighting the different measurements in the linear system with a negligible increase in computational load. Our approach exhibits clear improvements -- in both performance and runtime -- when compared to popular methods such as EPnP, CPnP, RPnP, and OPnP. Our new non-iterative solution approaches that of the true optimal found via Gauss-Newton optimization, but at a fraction of the computational cost. Our optimal DLT (oDLT) implementation, as well as the experiments, are released in open source.
comment: 8 pages, 6 figures, 2 tables
☆ Unlabeled Action Quality Assessment Based on Multi-dimensional Adaptive Constrained Dynamic Time Warping
The growing popularity of online sports and exercise necessitates effective methods for evaluating the quality of online exercise executions. Previous action quality assessment methods, which relied on labeled scores from motion videos, exhibited slightly lower accuracy and discriminability. This limitation hindered their rapid application to newly added exercises. To address this problem, this paper presents an unlabeled Multi-Dimensional Exercise Distance Adaptive Constrained Dynamic Time Warping (MED-ACDTW) method for action quality assessment. Our approach uses an athletic version of DTW to compare features from template and test videos, eliminating the need for score labels during training. The result shows that utilizing both 2D and 3D spatial dimensions, along with multiple human body features, improves the accuracy by 2-3% compared to using either 2D or 3D pose estimation alone. Additionally, employing MED for score calculation enhances the precision of frame distance matching, which significantly boosts overall discriminability. The adaptive constraint scheme enhances the discriminability of action quality assessment by approximately 30%. Furthermore, to address the absence of a standardized perspective in sports class evaluations, we introduce a new dataset called BGym.
☆ Assessing Open-world Forgetting in Generative Image Model Customization
Recent advances in diffusion models have significantly enhanced image generation capabilities. However, customizing these models with new classes often leads to unintended consequences that compromise their reliability. We introduce the concept of open-world forgetting to emphasize the vast scope of these unintended alterations, contrasting it with the well-studied closed-world forgetting, which is measurable by evaluating performance on a limited set of classes or skills. Our research presents the first comprehensive investigation into open-world forgetting in diffusion models, focusing on semantic and appearance drift of representations. We utilize zero-shot classification to analyze semantic drift, revealing that even minor model adaptations lead to unpredictable shifts affecting areas far beyond newly introduced concepts, with dramatic drops in zero-shot classification of up to 60%. Additionally, we observe significant changes in texture and color of generated content when analyzing appearance drift. To address these issues, we propose a mitigation strategy based on functional regularization, designed to preserve original capabilities while accommodating new concepts. Our study aims to raise awareness of unintended changes due to model customization and advocates for the analysis of open-world forgetting in future research on model customization and finetuning methods. Furthermore, we provide insights for developing more robust adaptation methodologies.
comment: Project page: https://hecoding.github.io/open-world-forgetting/
☆ Fine-Grained Verifiers: Preference Modeling as Next-token Prediction in Vision-Language Alignment
The recent advancements in large language models (LLMs) and pre-trained vision models have accelerated the development of vision-language large models (VLLMs), enhancing the interaction between visual and linguistic modalities. Despite their notable success across various domains, VLLMs face challenges in modality alignment, which can lead to issues like hallucinations and unsafe content generation. Current alignment techniques often rely on coarse feedback and external datasets, limiting scalability and performance. In this paper, we propose FiSAO (Fine-Grained Self-Alignment Optimization), a novel self-alignment method that utilizes the model's own visual encoder as a fine-grained verifier to improve vision-language alignment without the need for additional data. By leveraging token-level feedback from the vision encoder, FiSAO significantly improves vision-language alignment, even surpassing traditional preference tuning methods that require additional data. Through both theoretical analysis and experimental validation, we demonstrate that FiSAO effectively addresses the misalignment problem in VLLMs, marking the first instance of token-level rewards being applied to such models.
comment: 23 pages
☆ Preview-based Category Contrastive Learning for Knowledge Distillation
Knowledge distillation is a mainstream algorithm in model compression by transferring knowledge from the larger model (teacher) to the smaller model (student) to improve the performance of student. Despite many efforts, existing methods mainly investigate the consistency between instance-level feature representation or prediction, which neglects the category-level information and the difficulty of each sample, leading to undesirable performance. To address these issues, we propose a novel preview-based category contrastive learning method for knowledge distillation (PCKD). It first distills the structural knowledge of both instance-level feature correspondence and the relation between instance features and category centers in a contrastive learning fashion, which can explicitly optimize the category representation and explore the distinct correlation between representations of instances and categories, contributing to discriminative category centers and better classification results. Besides, we introduce a novel preview strategy to dynamically determine how much the student should learn from each sample according to their difficulty. Different from existing methods that treat all samples equally and curriculum learning that simply filters out hard samples, our method assigns a small weight for hard instances as a preview to better guide the student training. Extensive experiments on several challenging datasets, including CIFAR-100 and ImageNet, demonstrate the superiority over state-of-the-art methods.
comment: 14 pages, 8 figures, Journal
☆ ProReason: Multi-Modal Proactive Reasoning with Decoupled Eyesight and Wisdom
Large vision-language models (LVLMs) have witnessed significant progress on visual understanding tasks. However, they often prioritize language knowledge over image information on visual reasoning tasks, incurring performance degradation. To tackle this issue, we first identify the drawbacks of existing solutions (i.e., insufficient and irrelevant visual descriptions, and limited multi-modal capacities). We then decompose visual reasoning process into two stages: visual perception (i.e., eyesight) and textual reasoning (i.e., wisdom), and introduce a novel visual reasoning framework named ProReason. This framework features multi-run proactive perception and decoupled vision-reasoning capabilities. Briefly, given a multi-modal question, ProReason iterates proactive information collection and reasoning until the answer can be concluded with necessary and sufficient visual descriptions. Notably, the disassociation of capabilities allows seamless integration of existing large language models (LLMs) to compensate for the reasoning deficits of LVLMs. Our extensive experiments demonstrate that ProReason outperforms both existing multi-step reasoning frameworks and passive peer methods on a wide range of benchmarks for both open-source and closed-source models. In addition, with the assistance of LLMs, ProReason achieves a performance improvement of up to 15% on MMMU benchmark. Our insights into existing solutions and the decoupled perspective for feasible integration of LLMs illuminate future research on visual reasoning techniques, especially LLM-assisted ones.
☆ ViConsFormer: Constituting Meaningful Phrases of Scene Texts using Transformer-based Method in Vietnamese Text-based Visual Question Answering
Text-based VQA is a challenging task that requires machines to use scene texts in given images to yield the most appropriate answer for the given question. The main challenge of text-based VQA is exploiting the meaning and information from scene texts. Recent studies tackled this challenge by considering the spatial information of scene texts in images via embedding 2D coordinates of their bounding boxes. In this study, we follow the definition of meaning from linguistics to introduce a novel method that effectively exploits the information from scene texts written in Vietnamese. Experimental results show that our proposed method obtains state-of-the-art results on two large-scale Vietnamese Text-based VQA datasets. The implementation can be found at this link.
☆ Deep Learning Applications in Medical Image Analysis: Advancements, Challenges, and Future Directions
Medical image analysis has emerged as an essential element of contemporary healthcare, facilitating physicians in achieving expedited and precise diagnosis. Recent breakthroughs in deep learning, a subset of artificial intelligence, have markedly revolutionized the analysis of medical pictures, improving the accuracy and efficiency of clinical procedures. Deep learning algorithms, especially convolutional neural networks (CNNs), have demonstrated remarkable proficiency in autonomously learning features from multidimensional medical pictures, including MRI, CT, and X-ray scans, without the necessity for manual feature extraction. These models have been utilized across multiple medical disciplines, including pathology, radiology, ophthalmology, and cardiology, where they aid in illness detection, classification, and segmentation tasks......
☆ Extreme Precipitation Nowcasting using Multi-Task Latent Diffusion Models
Deep learning models have made remarkable strides in precipitation prediction, yet they continue to struggle with capturing the spatial details of the features of radar images, particularly over high precipitation intensity areas. This shortcoming is evident in the form of low forecast accuracy in the spatial positioning of radar echo images across varying precipitation intensity regions. To address this challenge, we introduce the multi-task latent diffusion model(MTLDM), a novel approach for precipitation prediction. The basic concept of the MTLDM is based on the understanding that the radar image representing precipitation is the result of multiple factors. Therefore, we adopt a divide-and-conquer approach, that is, we decompose the radar image using decomposition technology and then predict the decomposed sub-images separately. We conceptualize the precipitation image as a composition of various components corresponding to different precipitation intensities. The MTLDM decomposes the precipitation image into these distinct components and employs a dedicated task to predict each one. This method enables spatiotemporally consistent prediction of real-world precipitation areas up to 5-80 min in advance, outperforming existing state-of-the-art techniques across multiple evaluation metrics.
comment: 12 pages, 6figures
☆ Enhancing In-vehicle Multiple Object Tracking Systems with Embeddable Ising Machines
A cognitive function of tracking multiple objects, needed in autonomous mobile vehicles, comprises object detection and their temporal association. While great progress owing to machine learning has been recently seen for elaborating the similarity matrix between the objects that have been recognized and the objects detected in a current video frame, less for the assignment problem that finally determines the temporal association, which is a combinatorial optimization problem. Here we show an in-vehicle multiple object tracking system with a flexible assignment function for tracking through multiple long-term occlusion events. To solve the flexible assignment problem formulated as a nondeterministic polynomial time-hard problem, the system relies on an embeddable Ising machine based on a quantum-inspired algorithm called simulated bifurcation. Using a vehicle-mountable computing platform, we demonstrate a realtime system-wide throughput (23 frames per second on average) with the enhanced functionality.
comment: 18 pages, 7 figures, 2 tables
♻ ☆ Learning Generative Interactive Environments By Trained Agent Exploration
World models are increasingly pivotal in interpreting and simulating the rules and actions of complex environments. Genie, a recent model, excels at learning from visually diverse environments but relies on costly human-collected data. We observe that their alternative method of using random agents is too limited to explore the environment. We propose to improve the model by employing reinforcement learning based agents for data generation. This approach produces diverse datasets that enhance the model's ability to adapt and perform well across various scenarios and realistic actions within the environment. In this paper, we first release the model GenieRedux - an implementation based on Genie. Additionally, we introduce GenieRedux-G, a variant that uses the agent's readily available actions to factor out action prediction uncertainty during validation. Our evaluation, including a replication of the Coinrun case study, shows that GenieRedux-G achieves superior visual fidelity and controllability using the trained agent exploration. The proposed approach is reproducable, scalable and adaptable to new types of environments. Our codebase is available at https://github.com/insait-institute/GenieRedux .
♻ ☆ EVER: Exact Volumetric Ellipsoid Rendering for Real-time View Synthesis
We present Exact Volumetric Ellipsoid Rendering (EVER), a method for real-time differentiable emission-only volume rendering. Unlike recent rasterization based approach by 3D Gaussian Splatting (3DGS), our primitive based representation allows for exact volume rendering, rather than alpha compositing 3D Gaussian billboards. As such, unlike 3DGS our formulation does not suffer from popping artifacts and view dependent density, but still achieves frame rates of $\sim\!30$ FPS at 720p on an NVIDIA RTX4090. Since our approach is built upon ray tracing it enables effects such as defocus blur and camera distortion (e.g. such as from fisheye cameras), which are difficult to achieve by rasterization. We show that our method is more accurate with fewer blending issues than 3DGS and follow-up work on view-consistent rendering, especially on the challenging large-scale scenes from the Zip-NeRF dataset where it achieves sharpest results among real-time techniques.
comment: Project page: https://half-potato.gitlab.io/posts/ever
♻ ☆ Movie101v2: Improved Movie Narration Benchmark
Automatic movie narration aims to generate video-aligned plot descriptions to assist visually impaired audiences. Unlike standard video captioning, it involves not only describing key visual details but also inferring plots that unfold across multiple movie shots, presenting distinct and complex challenges. To advance this field, we introduce Movie101v2, a large-scale, bilingual dataset with enhanced data quality specifically designed for movie narration. Revisiting the task, we propose breaking down the ultimate goal of automatic movie narration into three progressive stages, offering a clear roadmap with corresponding evaluation metrics. Based on our new benchmark, we baseline a range of large vision-language models, including GPT-4V, and conduct an in-depth analysis of the challenges in narration generation. Our findings highlight that achieving applicable movie narration generation is a fascinating goal that requires significant research.
♻ ☆ Harnessing Shared Relations via Multimodal Mixup Contrastive Learning for Multimodal Classification NeurIPS 2024
Deep multimodal learning has shown remarkable success by leveraging contrastive learning to capture explicit one-to-one relations across modalities. However, real-world data often exhibits shared relations beyond simple pairwise associations. We propose M3CoL, a Multimodal Mixup Contrastive Learning approach to capture nuanced shared relations inherent in multimodal data. Our key contribution is a Mixup-based contrastive loss that learns robust representations by aligning mixed samples from one modality with their corresponding samples from other modalities thereby capturing shared relations between them. For multimodal classification tasks, we introduce a framework that integrates a fusion module with unimodal prediction modules for auxiliary supervision during training, complemented by our proposed Mixup-based contrastive loss. Through extensive experiments on diverse datasets (N24News, ROSMAP, BRCA, and Food-101), we demonstrate that M3CoL effectively captures shared multimodal relations and generalizes across domains. It outperforms state-of-the-art methods on N24News, ROSMAP, and BRCA, while achieving comparable performance on Food-101. Our work highlights the significance of learning shared relations for robust multimodal learning, opening up promising avenues for future research.
comment: RK and RS contributed equally to this work, 20 Pages, 8 Figures, 9 Tables. Another version of the paper accepted at NeurIPS 2024 Workshop on Unifying Representations in Neural Models (UniReps)
♻ ☆ IncEventGS: Pose-Free Gaussian Splatting from a Single Event Camera
Implicit neural representation and explicit 3D Gaussian Splatting (3D-GS) for novel view synthesis have achieved remarkable progress with frame-based camera (e.g. RGB and RGB-D cameras) recently. Compared to frame-based camera, a novel type of bio-inspired visual sensor, i.e. event camera, has demonstrated advantages in high temporal resolution, high dynamic range, low power consumption and low latency. Due to its unique asynchronous and irregular data capturing process, limited work has been proposed to apply neural representation or 3D Gaussian splatting for an event camera. In this work, we present IncEventGS, an incremental 3D Gaussian Splatting reconstruction algorithm with a single event camera. To recover the 3D scene representation incrementally, we exploit the tracking and mapping paradigm of conventional SLAM pipelines for IncEventGS. Given the incoming event stream, the tracker firstly estimates an initial camera motion based on prior reconstructed 3D-GS scene representation. The mapper then jointly refines both the 3D scene representation and camera motion based on the previously estimated motion trajectory from the tracker. The experimental results demonstrate that IncEventGS delivers superior performance compared to prior NeRF-based methods and other related baselines, even we do not have the ground-truth camera poses. Furthermore, our method can also deliver better performance compared to state-of-the-art event visual odometry methods in terms of camera motion estimation. Code is publicly available at: https://github.com/wu-cvgl/IncEventGS.
comment: Code Page: https://github.com/wu-cvgl/IncEventGS
♻ ☆ Scalable Drift Monitoring in Medical Imaging AI
The integration of artificial intelligence (AI) into medical imaging has advanced clinical diagnostics but poses challenges in managing model drift and ensuring long-term reliability. To address these challenges, we develop MMC+, an enhanced framework for scalable drift monitoring, building upon the CheXstray framework that introduced real-time drift detection for medical imaging AI models using multi-modal data concordance. This work extends the original framework's methodologies, providing a more scalable and adaptable solution for real-world healthcare settings and offers a reliable and cost-effective alternative to continuous performance monitoring addressing limitations of both continuous and periodic monitoring methods. MMC+ introduces critical improvements to the original framework, including more robust handling of diverse data streams, improved scalability with the integration of foundation models like MedImageInsight for high-dimensional image embeddings without site-specific training, and the introduction of uncertainty bounds to better capture drift in dynamic clinical environments. Validated with real-world data from Massachusetts General Hospital during the COVID-19 pandemic, MMC+ effectively detects significant data shifts and correlates them with model performance changes. While not directly predicting performance degradation, MMC+ serves as an early warning system, indicating when AI systems may deviate from acceptable performance bounds and enabling timely interventions. By emphasizing the importance of monitoring diverse data streams and evaluating data shifts alongside model performance, this work contributes to the broader adoption and integration of AI solutions in clinical settings.
♻ ☆ Fundus to Fluorescein Angiography Video Generation as a Retinal Generative Foundation Model
Fundus fluorescein angiography (FFA) is crucial for diagnosing and monitoring retinal vascular issues but is limited by its invasive nature and restricted accessibility compared to color fundus (CF) imaging. Existing methods that convert CF images to FFA are confined to static image generation, missing the dynamic lesional changes. We introduce Fundus2Video, an autoregressive generative adversarial network (GAN) model that generates dynamic FFA videos from single CF images. Fundus2Video excels in video generation, achieving an FVD of 1497.12 and a PSNR of 11.77. Clinical experts have validated the fidelity of the generated videos. Additionally, the model's generator demonstrates remarkable downstream transferability across ten external public datasets, including blood vessel segmentation, retinal disease diagnosis, systemic disease prediction, and multimodal retrieval, showcasing impressive zero-shot and few-shot capabilities. These findings position Fundus2Video as a powerful, non-invasive alternative to FFA exams and a versatile retinal generative foundation model that captures both static and temporal retinal features, enabling the representation of complex inter-modality relationships.
♻ ☆ Video-XL: Extra-Long Vision Language Model for Hour-Scale Video Understanding
Although current Multi-modal Large Language Models (MLLMs) demonstrate promising results in video understanding, processing extremely long videos remains an ongoing challenge. Typically, MLLMs struggle with handling thousands of visual tokens that exceed the maximum context length, and they suffer from the information decay due to token aggregation. Another challenge is the high computational cost stemming from the large number of video tokens. To tackle these issues, we propose Video-XL, an extra-long vision language model designed for efficient hour-scale video understanding. Specifically, we argue that LLMs can be adapted as effective visual condensers and propose Visual Context Latent Summarization which condenses visual contexts into highly compact forms. Extensive experiments demonstrate that our model achieves promising results on popular long video understanding benchmarks. For example, Video-XL outperforms the current state-of-the-art method on VNBench by nearly 10\% in accuracy. Moreover, Video-XL presents an impressive balance between efficiency and effectiveness, processing 2048 frames on a single 80GB GPU while achieving nearly 95% accuracy in the Needle-in-a-Haystack evaluation.
♻ ☆ On Efficient Variants of Segment Anything Model: A Survey
The Segment Anything Model (SAM) is a foundational model for image segmentation tasks, known for its strong generalization across diverse applications. However, its impressive performance comes with significant computational and resource demands, making it challenging to deploy in resource-limited environments such as edge devices. To address this, a variety of SAM variants have been proposed to enhance efficiency while keeping accuracy. This survey provides the first comprehensive review of these efficient SAM variants. We begin by exploring the motivations driving this research. We then present core techniques used in SAM and model acceleration. This is followed by a detailed exploration of SAM acceleration strategies, categorized by approach, and a discussion of several future research directions. Finally, we offer a unified and extensive evaluation of these methods across various hardware, assessing their efficiency and accuracy on representative benchmarks, and providing a clear comparison of their overall performance.
♻ ☆ Deep Implicit Optimization for Robust and Flexible Image Registration
Deep Learning in Image Registration (DLIR) methods have been tremendously successful in image registration due to their speed and ability to incorporate weak label supervision at training time. However, DLIR methods forego many of the benefits of classical optimization-based methods. The functional nature of deep networks do not guarantee that the predicted transformation is a local minima of the registration objective, the representation of the transformation (displacement/velocity field/affine) is fixed, and the networks are not robust to domain shift. Our method aims to bridge this gap between classical and learning methods by incorporating optimization as a layer in a deep network. A deep network is trained to predict multi-scale dense feature images that are registered using a black box iterative optimization solver. This optimal warp is then used to minimize image and label alignment errors. By implicitly differentiating end-to-end through an iterative optimization solver, our learned features are registration and label-aware, and the warp functions are guaranteed to be local minima of the registration objective in the feature space. Our framework shows excellent performance on in-domain datasets, and is agnostic to domain shift such as anisotropy and varying intensity profiles. For the first time, our method allows switching between arbitrary transformation representations (free-form to diffeomorphic) at test time with zero retraining. End-to-end feature learning also facilitates interpretability of features, and out-of-the-box promptability using additional label-fidelity terms at inference.
♻ ☆ Quantization Effects on Neural Networks Perception: How would quantization change the perceptual field of vision models?
Neural network quantization is a critical technique for deploying models on resource-limited devices. Despite its widespread use, the impact of quantization on model perceptual fields, particularly in relation to class activation maps (CAMs), remains underexplored. This study investigates how quantization influences the spatial recognition abilities of vision models by examining the alignment between CAMs and visual salient objects maps across various architectures. Utilizing a dataset of 10,000 images from ImageNet, we conduct a comprehensive evaluation of six diverse CNN architectures: VGG16, ResNet50, EfficientNet, MobileNet, SqueezeNet, and DenseNet. Through the systematic application of quantization techniques, we identify subtle changes in CAMs and their alignment with Salient object maps. Our results demonstrate the differing sensitivities of these architectures to quantization and highlight its implications for model performance and interpretability in real-world applications. This work primarily contributes to a deeper understanding of neural network quantization, offering insights essential for deploying efficient and interpretable models in practical settings.
comment: Accepted & presented at IPTA 2024
♻ ☆ MicroDreamer: Efficient 3D Generation in $\sim$20 Seconds by Score-based Iterative Reconstruction
Optimization-based approaches, such as score distillation sampling (SDS), show promise in zero-shot 3D generation but suffer from low efficiency, primarily due to the high number of function evaluations (NFEs) required for each sample and the limitation of optimization confined to latent space. This paper introduces score-based iterative reconstruction (SIR), an efficient and general algorithm mimicking a differentiable 3D reconstruction process to reduce the NFEs and enable optimization in pixel space. Given a single set of images sampled from a multi-view score-based diffusion model, SIR repeatedly optimizes 3D parameters, unlike the single-step optimization in SDS. With other improvements in training, we present an efficient approach called MicroDreamer that generally applies to various 3D representations and 3D generation tasks. In particular, MicroDreamer is 5-20 times faster than SDS in generating neural radiance field while retaining a comparable performance and takes about 20 seconds to create meshes from 3D Gaussian splatting on a single A100 GPU, halving the time of the fastest optimization-based baseline DreamGaussian with significantly superior performance compared to the measurement standard deviation. Our code is available at https://github.com/ML-GSAI/MicroDreamer.
♻ ☆ Similarity and Quality Metrics for MR Image-To-Image Translation
Image-to-image translation can create large impact in medical imaging, as images can be synthetically transformed to other modalities, sequence types, higher resolutions or lower noise levels. To ensure patient safety, these methods should be validated by human readers, which requires a considerable amount of time and costs. Quantitative metrics can effectively complement such studies and provide reproducible and objective assessment of synthetic images. If a reference is available, the similarity of MR images is frequently evaluated by SSIM and PSNR metrics, even though these metrics are not or too sensitive regarding specific distortions. When reference images to compare with are not available, non-reference quality metrics can reliably detect specific distortions, such as blurriness. To provide an overview on distortion sensitivity, we quantitatively analyze 11 similarity (reference) and 12 quality (non-reference) metrics for assessing synthetic images. We additionally include a metric on a downstream segmentation task. We investigate the sensitivity regarding 11 kinds of distortions and typical MR artifacts, and analyze the influence of different normalization methods on each metric and distortion. Finally, we derive recommendations for effective usage of the analyzed similarity and quality metrics for evaluation of image-to-image translation models.
comment: 21 pages, 8 figures, supplement with 16 pages, 10 figures, submitted to Nature Scientific Reports
♻ ☆ LED: Light Enhanced Depth Estimation at Night
Nighttime camera-based depth estimation is a highly challenging task, especially for autonomous driving applications, where accurate depth perception is essential for ensuring safe navigation. We aim to improve the reliability of perception systems at night time, where models trained on daytime data often fail in the absence of precise but costly LiDAR sensors. In this work, we introduce Light Enhanced Depth (LED), a novel cost-effective approach that significantly improves depth estimation in low-light environments by harnessing a pattern projected by high definition headlights available in modern vehicles. LED leads to significant performance boosts across multiple depth-estimation architectures (encoder-decoder, Adabins, DepthFormer) both on synthetic and real datasets. Furthermore, increased performances beyond illuminated areas reveal a holistic enhancement in scene understanding. Finally, we release the Nighttime Synthetic Drive Dataset, a new synthetic and photo-realistic nighttime dataset, which comprises 49,990 comprehensively annotated images.
comment: Preprint. Code and dataset available on the project page : https://simondemoreau.github.io/LED/
♻ ☆ DiTFastAttn: Attention Compression for Diffusion Transformer Models
Diffusion Transformers (DiT) excel at image and video generation but face computational challenges due to the quadratic complexity of self-attention operators. We propose DiTFastAttn, a post-training compression method to alleviate the computational bottleneck of DiT. We identify three key redundancies in the attention computation during DiT inference: (1) spatial redundancy, where many attention heads focus on local information; (2) temporal redundancy, with high similarity between the attention outputs of neighboring steps; (3) conditional redundancy, where conditional and unconditional inferences exhibit significant similarity. We propose three techniques to reduce these redundancies: (1) Window Attention with Residual Sharing to reduce spatial redundancy; (2) Attention Sharing across Timesteps to exploit the similarity between steps; (3) Attention Sharing across CFG to skip redundant computations during conditional generation. We apply DiTFastAttn to DiT, PixArt-Sigma for image generation tasks, and OpenSora for video generation tasks. Our results show that for image generation, our method reduces up to 76% of the attention FLOPs and achieves up to 1.8x end-to-end speedup at high-resolution (2k x 2k) generation.
♻ ☆ Prompt Tuning of Deep Neural Networks for Speaker-adaptive Visual Speech Recognition
Visual Speech Recognition (VSR) aims to infer speech into text depending on lip movements alone. As it focuses on visual information to model the speech, its performance is inherently sensitive to personal lip appearances and movements, and this makes the VSR models show degraded performance when they are applied to unseen speakers. In this paper, to remedy the performance degradation of the VSR model on unseen speakers, we propose prompt tuning methods of Deep Neural Networks (DNNs) for speaker-adaptive VSR. Specifically, motivated by recent advances in Natural Language Processing (NLP), we finetune prompts on adaptation data of target speakers instead of modifying the pre-trained model parameters. Different from the previous prompt tuning methods mainly limited to Transformer variant architecture, we explore different types of prompts, the addition, the padding, and the concatenation form prompts that can be applied to the VSR model which is composed of CNN and Transformer in general. With the proposed prompt tuning, we show that the performance of the pre-trained VSR model on unseen speakers can be largely improved by using a small amount of adaptation data (e.g., less than 5 minutes), even if the pre-trained model is already developed with large speaker variations. Moreover, by analyzing the performance and parameters of different types of prompts, we investigate when the prompt tuning is preferred over the finetuning methods. The effectiveness of the proposed method is evaluated on both word- and sentence-level VSR databases, LRW-ID and GRID.
comment: IEEE TPAMI
♻ ☆ Harnessing Webpage UIs for Text-Rich Visual Understanding
Text-rich visual understanding-the ability to process environments where dense textual content is integrated with visuals-is crucial for multimodal large language models (MLLMs) to interact effectively with structured environments. To enhance this capability, we propose synthesizing general multimodal instructions from webpage UIs using text-based large language models (LLMs). Despite lacking direct visual input, text-based LLMs are able to process structured text representations from webpage accessibility trees. These instructions are then paired with UI screenshots to train multimodal models. We introduce MultiUI, a dataset containing 7.3 million samples from 1 million websites, covering diverse multimodal tasks and UI layouts. Models trained on MultiUI not only excel in web UI tasks-achieving up to a 48% improvement on VisualWebBench and a 19.1% boost in element accuracy on a web agent dataset Mind2Web-but also generalize surprisingly well to non-web UI tasks and even to non-UI domains, such as document understanding, OCR, and chart interpretation. These results highlight the broad applicability of web UI data for advancing text-rich visual understanding across various scenarios.
♻ ☆ Dating ancient manuscripts using radiocarbon and AI-based writing style analysis
Determining the chronology of ancient handwritten manuscripts is essential for reconstructing the evolution of ideas. For the Dead Sea Scrolls, this is particularly important. However, there is an almost complete lack of date-bearing manuscripts evenly distributed across the timeline and written in similar scripts available for palaeographic comparison. Here, we present Enoch, a state-of-the-art AI-based date-prediction model, trained on the basis of new radiocarbon-dated samples of the scrolls. Enoch uses established handwriting-style descriptors and applies Bayesian ridge regression. The challenge of this study is that the number of radiocarbon-dated manuscripts is small, while current machine learning requires an abundance of training data. We show that by using combined angular and allographic writing style feature vectors and applying Bayesian ridge regression, Enoch could predict the radiocarbon-based dates from style, supported by leave-one-out validation, with varied MAEs of 27.9 to 30.7 years relative to the radiocarbon dating. Enoch was then used to estimate the dates of 135 unseen manuscripts, revealing that 79 per cent of the samples were considered 'realistic' upon palaeographic post-hoc evaluation. We present a new chronology of the scrolls. The radiocarbon ranges and Enoch's style-based predictions are often older than the traditionally assumed palaeographic estimates. In the range of 300-50 BCE, Enoch's date prediction provides an improved granularity. The study is in line with current developments in multimodal machine-learning techniques, and the methods can be used for date prediction in other partially-dated manuscript collections. This research shows how Enoch's quantitative, probability-based approach can be a tool for palaeographers and historians, re-dating ancient Jewish key texts and contributing to current debates on Jewish and Christian origins.
comment: 16 pages of main article, 103 pages of supplementary materials; the first version of this article is originally prepared in July 2023 after the completion of all the experiments
♻ ☆ Distribution Guidance Network for Weakly Supervised Point Cloud Semantic Segmentation
Despite alleviating the dependence on dense annotations inherent to fully supervised methods, weakly supervised point cloud semantic segmentation suffers from inadequate supervision signals. In response to this challenge, we introduce a novel perspective that imparts auxiliary constraints by regulating the feature space under weak supervision. Our initial investigation identifies which distributions accurately characterize the feature space, subsequently leveraging this priori to guide the alignment of the weakly supervised embeddings. Specifically, we analyze the superiority of the mixture of von Mises-Fisher distributions (moVMF) among several common distribution candidates. Accordingly, we develop a Distribution Guidance Network (DGNet), which comprises a weakly supervised learning branch and a distribution alignment branch. Leveraging reliable clustering initialization derived from the weakly supervised learning branch, the distribution alignment branch alternately updates the parameters of the moVMF and the network, ensuring alignment with the moVMF-defined latent space. Extensive experiments validate the rationality and effectiveness of our distribution choice and network design. Consequently, DGNet achieves state-of-the-art performance under multiple datasets and various weakly supervised settings.
♻ ☆ SatSwinMAE: Efficient Autoencoding for Multiscale Time-series Satellite Imagery
Recent advancements in foundation models have significantly impacted various fields, including natural language processing, computer vision, and multi-modal tasks. One area that stands to benefit greatly is Earth observation, where these models can efficiently process large-scale, unlabeled geospatial data. In this work we extend the SwinMAE model to integrate temporal information for satellite time-series data. The architecture employs a hierarchical 3D Masked Autoencoder (MAE) with Video Swin Transformer blocks to effectively capture multi-scale spatio-temporal dependencies in satellite imagery. To enhance transfer learning, we incorporate both encoder and decoder pretrained weights, along with skip connections to preserve scale-specific information. This forms an architecture similar to SwinUNet with an additional temporal component. Our approach shows significant performance improvements over existing state-of-the-art foundation models for all the evaluated downstream tasks: land cover segmentation, building density prediction, flood mapping, wildfire scar mapping and multi-temporal crop segmentation. Particularly, in the land cover segmentation task of the PhilEO Bench dataset, it outperforms other geospatial foundation models with a 10.4% higher accuracy.
♻ ☆ TotalVibeSegmentator: Full Body MRI Segmentation for the NAKO and UK Biobank
Objectives: To present a publicly available torso segmentation network for large epidemiology datasets on volumetric interpolated breath-hold examination (VIBE) images. Materials & Methods: We extracted preliminary segmentations from TotalSegmentator, spine, and body composition networks for VIBE images, then improved them iteratively and retrained a nnUNet network. Using subsets of NAKO (85 subjects) and UK Biobank (16 subjects), we evaluated with Dice-score on a holdout set (12 subjects) and existing organ segmentation approach (1000 subjects), generating 71 semantic segmentation types for VIBE images. We provide an additional network for the vertebra segments 22 individual vertebra types. Results: We achieved an average Dice score of 0.89 +- 0.07 overall 71 segmentation labels. We scored > 0.90 Dice-score on the abdominal organs except for the pancreas with a Dice of 0.70. Conclusion: Our work offers a detailed and refined publicly available full torso segmentation on VIBE images.
comment: https://github.com/robert-graf/TotalVibeSegmentator
♻ ☆ Enhanced Prompt-leveraged Weakly Supervised Cancer Segmentation based on Segment Anything
This work proposes a novel approach beyond supervised learning for effective pathological image analysis, addressing the challenge of limited robust labeled data. Pathological diagnosis of diseases like cancer has conventionally relied on the evaluation of morphological features by physicians and pathologists. However, recent advancements in compute-aided diagnosis (CAD) systems are gaining significant attention as diagnostic support tools. Although the advancement of deep learning has improved CAD significantly, segmentation models typically require large pixel-level annotated dataset, and such labeling is expensive. Existing studies not based on supervised approaches still struggle with limited generalization, and no practical approach has emerged yet. To address this issue, we present a weakly supervised semantic segmentation (WSSS) model by combining class activation map and Segment Anything Model (SAM)-based pseudo-labeling. For effective pretraining, we adopt the SAM-a foundation model that is pretrained on large datasets and operates in zero-shot configurations using only coarse prompts. The proposed approach transfer enhanced Attention Dropout Layer's knowledge to SAM, thereby generating pseudo-labels. To demonstrate the superiority of the proposed method, experimental studies are conducted on histopathological breast cancer datasets. The proposed method outperformed other WSSS methods across three datasets, demonstrating its efficiency by achieving this with only 12GB of GPU memory during training. Our code is available at : https://github.com/QI-NemoSong/EPLC-SAM
comment: 10 pages, 7 figures
♻ ☆ Encode-Store-Retrieve: Augmenting Human Memory through Language-Encoded Egocentric Perception
We depend on our own memory to encode, store, and retrieve our experiences. However, memory lapses can occur. One promising avenue for achieving memory augmentation is through the use of augmented reality head-mounted displays to capture and preserve egocentric videos, a practice commonly referred to as lifelogging. However, a significant challenge arises from the sheer volume of video data generated through lifelogging, as the current technology lacks the capability to encode and store such large amounts of data efficiently. Further, retrieving specific information from extensive video archives requires substantial computational power, further complicating the task of quickly accessing desired content. To address these challenges, we propose a memory augmentation agent that involves leveraging natural language encoding for video data and storing them in a vector database. This approach harnesses the power of large vision language models to perform the language encoding process. Additionally, we propose using large language models to facilitate natural language querying. Our agent underwent extensive evaluation using the QA-Ego4D dataset and achieved state-of-the-art results with a BLEU score of 8.3, outperforming conventional machine learning models that scored between 3.4 and 5.8. Additionally, we conducted a user study in which participants interacted with the human memory augmentation agent through episodic memory and open-ended questions. The results of this study show that the agent results in significantly better recall performance on episodic memory tasks compared to human participants. The results also highlight the agent's practical applicability and user acceptance.
♻ ☆ Object Pose Estimation via the Aggregation of Diffusion Features CVPR2024
Estimating the pose of objects from images is a crucial task of 3D scene understanding, and recent approaches have shown promising results on very large benchmarks. However, these methods experience a significant performance drop when dealing with unseen objects. We believe that it results from the limited generalizability of image features. To address this problem, we have an in-depth analysis on the features of diffusion models, e.g. Stable Diffusion, which hold substantial potential for modeling unseen objects. Based on this analysis, we then innovatively introduce these diffusion features for object pose estimation. To achieve this, we propose three distinct architectures that can effectively capture and aggregate diffusion features of different granularity, greatly improving the generalizability of object pose estimation. Our approach outperforms the state-of-the-art methods by a considerable margin on three popular benchmark datasets, LM, O-LM, and T-LESS. In particular, our method achieves higher accuracy than the previous best arts on unseen objects: 97.9% vs. 93.5% on Unseen LM, 85.9% vs. 76.3% on Unseen O-LM, showing the strong generalizability of our method. Our code is released at https://github.com/Tianfu18/diff-feats-pose.
comment: Accepted to CVPR2024, fix typo
♻ ☆ VLFeedback: A Large-Scale AI Feedback Dataset for Large Vision-Language Models Alignment EMNLP 2024
As large vision-language models (LVLMs) evolve rapidly, the demand for high-quality and diverse data to align these models becomes increasingly crucial. However, the creation of such data with human supervision proves costly and time-intensive. In this paper, we investigate the efficacy of AI feedback to scale supervision for aligning LVLMs. We introduce VLFeedback, the first large-scale vision-language feedback dataset, comprising over 82K multi-modal instructions and comprehensive rationales generated by off-the-shelf models without human annotations. To evaluate the effectiveness of AI feedback for vision-language alignment, we train Silkie, an LVLM fine-tuned via direct preference optimization on VLFeedback. Silkie showcases exceptional performance regarding helpfulness, visual faithfulness, and safety metrics. It outperforms its base model by 6.9\% and 9.5\% in perception and cognition tasks, reduces hallucination issues on MMHal-Bench, and exhibits enhanced resilience against red-teaming attacks. Furthermore, our analysis underscores the advantage of AI feedback, particularly in fostering preference diversity to deliver more comprehensive improvements. Our dataset, training code and models are available at https://vlf-silkie.github.io.
comment: EMNLP 2024 Main Conference camera-ready version (fixed small typos). This article supersedes arXiv:2312.10665
♻ ☆ X-Fi: A Modality-Invariant Foundation Model for Multimodal Human Sensing
Human sensing, which employs various sensors and advanced deep learning technologies to accurately capture and interpret human body information, has significantly impacted fields like public security and robotics. However, current human sensing primarily depends on modalities such as cameras and LiDAR, each of which has its own strengths and limitations. Furthermore, existing multi-modal fusion solutions are typically designed for fixed modality combinations, requiring extensive retraining when modalities are added or removed for diverse scenarios. In this paper, we propose a modality-invariant foundation model for all modalities, X-Fi, to address this issue. X-Fi enables the independent or combinatory use of sensor modalities without additional training by utilizing a transformer structure to accommodate variable input sizes and incorporating a novel "X-fusion" mechanism to preserve modality-specific features during multimodal integration. This approach not only enhances adaptability but also facilitates the learning of complementary features across modalities. Extensive experiments conducted on the MM-Fi and XRF55 datasets, employing six distinct modalities, demonstrate that X-Fi achieves state-of-the-art performance in human pose estimation (HPE) and human activity recognition (HAR) tasks. The findings indicate that our proposed model can efficiently support a wide range of human sensing applications, ultimately contributing to the evolution of scalable, multimodal sensing technologies.
♻ ☆ Suppress Content Shift: Better Diffusion Features via Off-the-Shelf Generation Techniques
Diffusion models are powerful generative models, and this capability can also be applied to discrimination. The inner activations of a pre-trained diffusion model can serve as features for discriminative tasks, namely, diffusion feature. We discover that diffusion feature has been hindered by a hidden yet universal phenomenon that we call content shift. To be specific, there are content differences between features and the input image, such as the exact shape of a certain object. We locate the cause of content shift as one inherent characteristic of diffusion models, which suggests the broad existence of this phenomenon in diffusion feature. Further empirical study also indicates that its negative impact is not negligible even when content shift is not visually perceivable. Hence, we propose to suppress content shift to enhance the overall quality of diffusion features. Specifically, content shift is related to the information drift during the process of recovering an image from the noisy input, pointing out the possibility of turning off-the-shelf generation techniques into tools for content shift suppression. We further propose a practical guideline named GATE to efficiently evaluate the potential benefit of a technique and provide an implementation of our methodology. Despite the simplicity, the proposed approach has achieved superior results on various tasks and datasets, validating its potential as a generic booster for diffusion features. Our code is available at https://github.com/Darkbblue/diffusion-content-shift.
comment: arXiv admin note: substantial text overlap with arXiv:2410.03558
♻ ☆ Not All Diffusion Model Activations Have Been Evaluated as Discriminative Features
Diffusion models are initially designed for image generation. Recent research shows that the internal signals within their backbones, named activations, can also serve as dense features for various discriminative tasks such as semantic segmentation. Given numerous activations, selecting a small yet effective subset poses a fundamental problem. To this end, the early study of this field performs a large-scale quantitative comparison of the discriminative ability of the activations. However, we find that many potential activations have not been evaluated, such as the queries and keys used to compute attention scores. Moreover, recent advancements in diffusion architectures bring many new activations, such as those within embedded ViT modules. Both combined, activation selection remains unresolved but overlooked. To tackle this issue, this paper takes a further step with a much broader range of activations evaluated. Considering the significant increase in activations, a full-scale quantitative comparison is no longer operational. Instead, we seek to understand the properties of these activations, such that the activations that are clearly inferior can be filtered out in advance via simple qualitative evaluation. After careful analysis, we discover three properties universal among diffusion models, enabling this study to go beyond specific models. On top of this, we present effective feature selection solutions for several popular diffusion models. Finally, the experiments across multiple discriminative tasks validate the superiority of our method over the SOTA competitors. Our code is available at https://github.com/Darkbblue/generic-diffusion-feature.
♻ ☆ UniG: Modelling Unitary 3D Gaussians for View-consistent 3D Reconstruction
In this work, we present UniG, a view-consistent 3D reconstruction and novel view synthesis model that generates a high-fidelity representation of 3D Gaussians from sparse images. Existing 3D Gaussians-based methods usually regress Gaussians per-pixel of each view, create 3D Gaussians per view separately, and merge them through point concatenation. Such a view-independent reconstruction approach often results in a view inconsistency issue, where the predicted positions of the same 3D point from different views may have discrepancies. To address this problem, we develop a DETR (DEtection TRansformer)-like framework, which treats 3D Gaussians as decoder queries and updates their parameters layer by layer by performing multi-view cross-attention (MVDFA) over multiple input images. In this way, multiple views naturally contribute to modeling a unitary representation of 3D Gaussians, thereby making 3D reconstruction more view-consistent. Moreover, as the number of 3D Gaussians used as decoder queries is irrespective of the number of input views, allow an arbitrary number of input images without causing memory explosion. Extensive experiments validate the advantages of our approach, showcasing superior performance over existing methods quantitatively (improving PSNR by 4.2 dB when trained on Objaverse and tested on the GSO benchmark) and qualitatively. The code will be released at https://github.com/jwubz123/UNIG.
♻ ☆ ViLCo-Bench: VIdeo Language COntinual learning Benchmark NeurIPS
Video language continual learning involves continuously adapting to information from video and text inputs, enhancing a model's ability to handle new tasks while retaining prior knowledge. This field is a relatively under-explored area, and establishing appropriate datasets is crucial for facilitating communication and research in this field. In this study, we present the first dedicated benchmark, ViLCo-Bench, designed to evaluate continual learning models across a range of video-text tasks. The dataset comprises ten-minute-long videos and corresponding language queries collected from publicly available datasets. Additionally, we introduce a novel memory-efficient framework that incorporates self-supervised learning and mimics long-term and short-term memory effects. This framework addresses challenges including memory complexity from long video clips, natural language complexity from open queries, and text-video misalignment. We posit that ViLCo-Bench, with greater complexity compared to existing continual learning benchmarks, would serve as a critical tool for exploring the video-language domain, extending beyond conventional class-incremental tasks, and addressing complex and limited annotation issues. The curated data, evaluations, and our novel method are available at https://github.com/cruiseresearchgroup/ViLCo.
comment: 14 pages, 4 figures, 8 tables, Accepted at NeurIPS Dataset and Benchmark Track 2024
♻ ☆ MK-SGN: A Spiking Graph Convolutional Network with Multimodal Fusion and Knowledge Distillation for Skeleton-based Action Recognition
In recent years, skeleton-based action recognition, leveraging multimodal Graph Convolutional Networks (GCN), has achieved remarkable results. However, due to their deep structure and reliance on continuous floating-point operations, GCN-based methods are energy-intensive. We propose an innovative Spiking Graph Convolutional Network with Multimodal Fusion and Knowledge Distillation (MK-SGN) to address this issue. By merging the energy efficiency of Spiking Neural Network (SNN) with the graph representation capability of GCN, the proposed MK-SGN reduces energy consumption while maintaining recognition accuracy. Firstly, we convert Graph Convolutional Networks (GCN) into Spiking Graph Convolutional Networks (SGN) establishing a new benchmark and paving the way for future research exploration. During this process, we introduce a spiking attention mechanism and design a Spiking-Spatio Graph Convolution module with a Spatial Global Spiking Attention mechanism (SA-SGC), enhancing feature learning capability. Secondly, we propose a Spiking Multimodal Fusion module (SMF), leveraging mutual information to process multimodal data more efficiently. Lastly, we delve into knowledge distillation methods from multimodal GCN to SGN and propose a novel, integrated method that simultaneously focuses on both intermediate layer distillation and soft label distillation to improve the performance of SGN. MK-SGN outperforms the state-of-the-art GCN-like frameworks on three challenging datasets for skeleton-based action recognition in reducing energy consumption. It also outperforms the state-of-the-art SNN frameworks in accuracy. Specifically, our method reduces energy consumption by more than 98% compared to typical GCN-based methods, while maintaining competitive accuracy on the NTU-RGB+D 60 cross-subject split using 4-time steps.
♻ ☆ Hard Region Aware Network for Remote Sensing Change Detection
Change detection (CD) is essential for various real-world applications, such as urban management and disaster assessment. Numerous CD methods have been proposed, and considerable results have been achieved recently. However, detecting changes in hard regions, i.e., the change boundary and irrelevant pseudo changes caused by background clutters, remains difficult for these methods, since they pose equal attention for all regions in bi-temporal images. This paper proposes a novel change detection network, termed as HRANet, which provides accurate change maps via hard region mining. Specifically, an online hard region estimation branch is constructed to model the pixel-wise hard samples, supervised by the error between predicted change maps and corresponding ground truth during the training process. A cross-layer knowledge review module is introduced to distill temporal change information from low-level to high-level features, thereby enhancing the feature representation capabilities. Finally, the hard region aware features extracted from the online hard region estimation branch and multi-level temporal difference features are aggregated into a unified feature representation to improve the accuracy of CD. Experimental results on two benchmark datasets demonstrate the superior performance of HRANet in the CD task.
♻ ☆ Signal-SGN: A Spiking Graph Convolutional Network for Skeletal Action Recognition via Learning Temporal-Frequency Dynamics
In skeletal-based action recognition, Graph Convolutional Networks (GCNs) based methods face limitations due to their complexity and high energy consumption. Spiking Neural Networks (SNNs) have gained attention in recent years for their low energy consumption, but existing methods combining GCNs and SNNs fail to fully utilize the temporal characteristics of skeletal sequences, leading to increased storage and computational costs. To address this issue, we propose a Signal-SGN(Spiking Graph Convolutional Network), which leverages the temporal dimension of skeletal sequences as the spiking timestep and treats features as discrete stochastic signals. The core of the network consists of a 1D Spiking Graph Convolutional Network (1D-SGN) and a Frequency Spiking Convolutional Network (FSN). The SGN performs graph convolution on single frames and incorporates spiking network characteristics to capture inter-frame temporal relationships, while the FSN uses Fast Fourier Transform (FFT) and complex convolution to extract temporal-frequency features. We also introduce a multi-scale wavelet transform feature fusion module(MWTF) to capture spectral features of temporal signals, enhancing the model's classification capability. We propose a pluggable temporal-frequency spatial semantic feature extraction module(TFSM) to enhance the model's ability to distinguish features without increasing inference-phase consumption. Our numerous experiments on the NTU RGB+D, NTU RGB+D 120, and NW-UCLA datasets demonstrate that the proposed models not only surpass existing SNN-based methods in accuracy but also reduce computational and storage costs during training. Furthermore, they achieve competitive accuracy compared to corresponding GCN-based methods, which is quite remarkable.
♻ ☆ Scene Prior Filtering for Depth Super-Resolution
Multi-modal fusion is vital to the success of super-resolution of depth maps. However, commonly used fusion strategies, such as addition and concatenation, fall short of effectively bridging the modal gap. As a result, guided image filtering methods have been introduced to mitigate this issue. Nevertheless, it is observed that their filter kernels usually encounter significant texture interference and edge inaccuracy. To tackle these two challenges, we introduce a Scene Prior Filtering network, SPFNet, which utilizes the priors surface normal and semantic map from large-scale models. Specifically, we design an All-in-one Prior Propagation that computes the similarity between multi-modal scene priors, i.e., RGB, normal, semantic, and depth, to reduce the texture interference. In addition, we present a One-to-one Prior Embedding that continuously embeds each single-modal prior into depth using Mutual Guided Filtering, further alleviating the texture interference while enhancing edges. Our SPFNet has been extensively evaluated on both real and synthetic datasets, achieving state-of-the-art performance.
comment: 13 pages
♻ ☆ LatentExplainer: Explaining Latent Representations in Deep Generative Models with Multi-modal Foundation Models
Deep generative models like VAEs and diffusion models have advanced various generation tasks by leveraging latent variables to learn data distributions and generate high-quality samples. Despite the field of explainable AI making strides in interpreting machine learning models, understanding latent variables in generative models remains challenging. This paper introduces \textit{LatentExplainer}, a framework for automatically generating semantically meaningful explanations of latent variables in deep generative models. \textit{LatentExplainer} tackles three main challenges: inferring the meaning of latent variables, aligning explanations with inductive biases, and handling varying degrees of explainability. Our approach perturbs latent variables, interpreting changes in generated data, and uses multi-modal large language models (MLLMs) to produce human-understandable explanations. We evaluate our proposed method on several real-world and synthetic datasets, and the results demonstrate superior performance in generating high-quality explanations for latent variables. The results highlight the effectiveness of incorporating inductive biases and uncertainty quantification, significantly enhancing model interpretability.
♻ ☆ From Introspection to Best Practices: Principled Analysis of Demonstrations in Multimodal In-Context Learning
Motivated by in-context learning (ICL) capabilities of Large Language models (LLMs), multimodal LLMs with additional visual modality are also exhibited with similar ICL abilities when multiple image-text pairs are provided as demonstrations. However, relatively less work has been done to investigate the principles behind how and why multimodal ICL works. We conduct a systematic and principled evaluation of multimodal ICL for models of different scales on a broad spectrum of new yet critical tasks. Through perturbations over different modality information, we show that modalities matter differently across tasks in multimodal ICL. Guided by task-specific modality impact, we recommend modality-driven demonstration strategies to boost ICL performance. We also find that models may follow inductive biases from multimodal ICL even if they are rarely seen in or contradict semantic priors from pretraining data. Our principled analysis provides a comprehensive way of understanding the role of demonstrations in multimodal in-context learning, and sheds light on effectively improving multimodal ICL on a wide range of tasks.
♻ ☆ PAPL-SLAM: Principal Axis-Anchored Monocular Point-Line SLAM
In point-line SLAM systems, the utilization of line structural information and the optimization of lines are two significant problems. The former is usually addressed through structural regularities, while the latter typically involves using minimal parameter representations of lines in optimization. However, separating these two steps leads to the loss of constraint information to each other. We anchor lines with similar directions to a principal axis and optimize them with $n+2$ parameters for $n$ lines, solving both problems together. Our method considers scene structural information, which can be easily extended to different world hypotheses while significantly reducing the number of line parameters to be optimized, enabling rapid and accurate mapping and tracking. To further enhance the system's robustness and avoid mismatch, we have modeled the line-axis probabilistic data association and provided the algorithm for axis creation, updating, and optimization. Additionally, considering that most real-world scenes conform to the Atlanta World hypothesis, we provide a structural line detection strategy based on vertical priors and vanishing points. Experimental results and ablation studies on various indoor and outdoor datasets demonstrate the effectiveness of our system.
comment: 8 pages, 4 figures
♻ ☆ PredFormer: Transformers Are Effective Spatial-Temporal Predictive Learners
Spatiotemporal predictive learning methods generally fall into two categories: recurrent-based approaches, which face challenges in parallelization and performance, and recurrent-free methods, which employ convolutional neural networks (CNNs) as encoder-decoder architectures. These methods benefit from strong inductive biases but often at the expense of scalability and generalization. This paper proposes PredFormer, a pure transformer-based framework for spatiotemporal predictive learning. Motivated by the Vision Transformers (ViT) design, PredFormer leverages carefully designed Gated Transformer blocks, following a comprehensive analysis of 3D attention mechanisms, including full-, factorized-, and interleaved-spatial-temporal attention. With its recurrent-free, transformer-based design, PredFormer is both simple and efficient, significantly outperforming previous methods by large margins. Extensive experiments on synthetic and real-world datasets demonstrate that PredFormer achieves state-of-the-art performance. On Moving MNIST, PredFormer achieves a 51.3% reduction in MSE relative to SimVP. For TaxiBJ, the model decreases MSE by 33.1% and boosts FPS from 533 to 2364. Additionally, on WeatherBench, it reduces MSE by 11.1% while enhancing FPS from 196 to 404. These performance gains in both accuracy and efficiency demonstrate PredFormer's potential for real-world applications. The source code will be released at https://github.com/yyyujintang/PredFormer .
comment: 15 pages, 7 figures
♻ ☆ Biometric Authentication Based on Enhanced Remote Photoplethysmography Signal Morphology
Remote photoplethysmography (rPPG) is a non-contact method for measuring cardiac signals from facial videos, offering a convenient alternative to contact photoplethysmography (cPPG) obtained from contact sensors. Recent studies have shown that each individual possesses a unique cPPG signal morphology that can be utilized as a biometric identifier, which has inspired us to utilize the morphology of rPPG signals extracted from facial videos for person authentication. Since the facial appearance and rPPG are mixed in the facial videos, we first de-identify facial videos to remove facial appearance while preserving the rPPG information, which protects facial privacy and guarantees that only rPPG is used for authentication. The de-identified videos are fed into an rPPG model to get the rPPG signal morphology for authentication. In the first training stage, unsupervised rPPG training is performed to get coarse rPPG signals. In the second training stage, an rPPG-cPPG hybrid training is performed by incorporating external cPPG datasets to achieve rPPG biometric authentication and enhance rPPG signal morphology. Our approach needs only de-identified facial videos with subject IDs to train rPPG authentication models. The experimental results demonstrate that rPPG signal morphology hidden in facial videos can be used for biometric authentication. The code is available at https://github.com/zhaodongsun/rppg_biometrics.
comment: accepted by IJCB 2024, Best Paper Runner-Up Award
♻ ☆ DAWN: Dynamic Frame Avatar with Non-autoregressive Diffusion Framework for Talking Head Video Generation
Talking head generation intends to produce vivid and realistic talking head videos from a single portrait and speech audio clip. Although significant progress has been made in diffusion-based talking head generation, almost all methods rely on autoregressive strategies, which suffer from limited context utilization beyond the current generation step, error accumulation, and slower generation speed. To address these challenges, we present DAWN (Dynamic frame Avatar With Non-autoregressive diffusion), a framework that enables all-at-once generation of dynamic-length video sequences. Specifically, it consists of two main components: (1) audio-driven holistic facial dynamics generation in the latent motion space, and (2) audio-driven head pose and blink generation. Extensive experiments demonstrate that our method generates authentic and vivid videos with precise lip motions, and natural pose/blink movements. Additionally, with a high generation speed, DAWN possesses strong extrapolation capabilities, ensuring the stable production of high-quality long videos. These results highlight the considerable promise and potential impact of DAWN in the field of talking head video generation. Furthermore, we hope that DAWN sparks further exploration of non-autoregressive approaches in diffusion models. Our code will be publicly available at https://github.com/Hanbo-Cheng/DAWN-pytorch.
♻ ☆ Diffusion Curriculum: Synthetic-to-Real Generative Curriculum Learning via Image-Guided Diffusion
Low-quality or scarce data has posed significant challenges for training deep neural networks in practice. While classical data augmentation cannot contribute very different new data, diffusion models opens up a new door to build self-evolving AI by generating high-quality and diverse synthetic data through text-guided prompts. However, text-only guidance cannot control synthetic images' proximity to the original images, resulting in out-of-distribution data detrimental to the model performance. To overcome the limitation, we study image guidance to achieve a spectrum of interpolations between synthetic and real images. With stronger image guidance, the generated images are similar to the training data but hard to learn. While with weaker image guidance, the synthetic images will be easier for model but contribute to a larger distribution gap with the original data. The generated full spectrum of data enables us to build a novel "Diffusion Curriculum (DisCL)". DisCL adjusts the image guidance level of image synthesis for each training stage: It identifies and focuses on hard samples for the model and assesses the most effective guidance level of synthetic images to improve hard data learning. We apply DisCL to two challenging tasks: long-tail (LT) classification and learning from low-quality data. It focuses on lower-guidance images of high-quality to learn prototypical features as a warm-up of learning higher-guidance images that might be weak on diversity or quality. Extensive experiments showcase a gain of 2.7% and 2.1% in OOD and ID macro-accuracy when applying DisCL to iWildCam dataset. On ImageNet-LT, DisCL improves the base model's tail-class accuracy from 4.4% to 23.64% and leads to a 4.02% improvement in all-class accuracy.
comment: 23 pages, including references and appendix. Code is available at http://github.com/tianyi-lab/DisCL
♻ ☆ ExACT: Teaching AI Agents to Explore with Reflective-MCTS and Exploratory Learning
Autonomous agents have demonstrated significant potential in automating complex multistep decision-making tasks. However, even state-of-the-art vision-language models (VLMs), such as GPT-4o, still fall short of human-level performance, particularly in intricate web environments and long-horizon tasks. To address these limitations, we present ExACT, an approach to combine test-time search and self-learning to build o1-like models for agentic applications. We first introduce Reflective Monte Carlo Tree Search (R-MCTS), a novel test time algorithm designed to enhance AI agents' ability to explore decision space on the fly. R-MCTS extends traditional MCTS by 1) incorporating contrastive reflection, allowing agents to learn from past interactions and dynamically improve their search efficiency; and 2) using multi-agent debate for reliable state evaluation. Next, we introduce Exploratory Learning, a novel learning strategy to teach agents to search at inference time without relying on any external search algorithms. On the challenging VisualWebArena benchmark, our GPT-4o based R-MCTS agent achieves a 6% to 30% relative improvement across various tasks compared to the previous state-of-the-art. Additionally, we show that the knowledge and experience gained from test-time search can be effectively transferred back to GPT-4o via fine-tuning. After Exploratory Learning, GPT-4o 1) demonstrates the ability to explore the environment, evaluate a state, and backtrack to viable ones when it detects that the current state cannot lead to success, and 2) matches 87% of R-MCTS's performance while using significantly less compute. Notably, our work demonstrates the compute scaling properties in both training - data collection with R-MCTS - and testing time. These results suggest a promising research direction to enhance VLMs' capabilities for agentic applications via test-time search and self-learning.
♻ ☆ Double-Condensing Attention Condenser: Leveraging Attention in Deep Learning to Detect Skin Cancer from Skin Lesion Images
Skin cancer is the most common type of cancer in the United States and is estimated to affect one in five Americans. Recent advances have demonstrated strong performance on skin cancer detection, as exemplified by state of the art performance in the SIIM-ISIC Melanoma Classification Challenge; however these solutions leverage ensembles of complex deep neural architectures requiring immense storage and compute costs, and therefore may not be tractable. A recent movement for TinyML applications is integrating Double-Condensing Attention Condensers (DC-AC) into a self-attention neural network backbone architecture to allow for faster and more efficient computation. This paper explores leveraging an efficient self-attention structure to detect skin cancer in skin lesion images and introduces a deep neural network design with DC-AC customized for skin cancer detection from skin lesion images. The final model is publicly available as a part of a global open-source initiative dedicated to accelerating advancement in machine learning to aid clinicians in the fight against cancer. Future work of this research includes iterating on the design of the selected network architecture and refining the approach to generalize to other forms of cancer.
♻ ☆ Residual-INR: Communication Efficient On-Device Learning Using Implicit Neural Representation
Edge computing is a distributed computing paradigm that collects and processes data at or near the source of data generation. The on-device learning at edge relies on device-to-device wireless communication to facilitate real-time data sharing and collaborative decision-making among multiple devices. This significantly improves the adaptability of the edge computing system to the changing environments. However, as the scale of the edge computing system is getting larger, communication among devices is becoming the bottleneck because of the limited bandwidth of wireless communication leads to large data transfer latency. To reduce the amount of device-to-device data transmission and accelerate on-device learning, in this paper, we propose Residual-INR, a fog computing-based communication-efficient on-device learning framework by utilizing implicit neural representation (INR) to compress images/videos into neural network weights. Residual-INR enhances data transfer efficiency by collecting JPEG images from edge devices, compressing them into INR format at the fog node, and redistributing them for on-device learning. By using a smaller INR for full image encoding and a separate object INR for high-quality object region reconstruction through residual encoding, our technique can reduce the encoding redundancy while maintaining the object quality. Residual-INR is a promising solution for edge on-device learning because it reduces data transmission by up to 5.16 x across a network of 10 edge devices. It also facilitates CPU-free accelerated on-device learning, achieving up to 2.9 x speedup without sacrificing accuracy. Our code is available at: https://github.com/sharclab/Residual-INR.
comment: This paper has been accepted by ICCAD 2024
♻ ☆ Embedded Prompt Tuning: Towards Enhanced Calibration of Pretrained Models for Medical Images
Foundation models pre-trained on large-scale data have been widely witnessed to achieve success in various natural imaging downstream tasks. Parameter-efficient fine-tuning (PEFT) methods aim to adapt foundation models to new domains by updating only a small portion of parameters in order to reduce computational overhead. However, the effectiveness of these PEFT methods, especially in cross-domain few-shot scenarios, e.g., medical image analysis, has not been fully explored. In this work, we facilitate the study of the performance of PEFT when adapting foundation models to medical image classification tasks. Furthermore, to alleviate the limitations of prompt introducing ways and approximation capabilities on Transformer architectures of mainstream prompt tuning methods, we propose the Embedded Prompt Tuning (EPT) method by embedding prompt tokens into the expanded channels. We also find that there are anomalies in the feature space distribution of foundation models during pre-training process, and prompt tuning can help mitigate this negative impact. To explain this phenomenon, we also introduce a novel perspective to understand prompt tuning: Prompt tuning is a distribution calibrator. And we support it by analyzing patch-wise scaling and feature separation operations contained in EPT. Our experiments show that EPT outperforms several state-of-the-art fine-tuning methods by a significant margin on few-shot medical image classification tasks, and completes the fine-tuning process within highly competitive time, indicating EPT is an effective PEFT method. The source code is available at github.com/zuwenqiang/EPT.
comment: 16 pages, 7 figures. arXiv admin note: text overlap with arXiv:2306.09579, arXiv:2203.12119 by other authors
♻ ☆ Polyhedral Complex Derivation from Piecewise Trilinear Networks NeurIPS 2024
Recent advancements in visualizing deep neural networks provide insights into their structures and mesh extraction from Continuous Piecewise Affine (CPWA) functions. Meanwhile, developments in neural surface representation learning incorporate non-linear positional encoding, addressing issues like spectral bias; however, this poses challenges in applying mesh extraction techniques based on CPWA functions. Focusing on trilinear interpolating methods as positional encoding, we present theoretical insights and an analytical mesh extraction, showing the transformation of hypersurfaces to flat planes within the trilinear region under the eikonal constraint. Moreover, we introduce a method for approximating intersecting points among three hypersurfaces contributing to broader applications. We empirically validate correctness and parsimony through chamfer distance and efficiency, and angular distance, while examining the correlation between the eikonal loss and the planarity of the hypersurfaces.
comment: Accepted at NeurIPS 2024. Updated with the camera-ready version
♻ ☆ MOS: Model Synergy for Test-Time Adaptation on LiDAR-Based 3D Object Detection
LiDAR-based 3D object detection is crucial for various applications but often experiences performance degradation in real-world deployments due to domain shifts. While most studies focus on cross-dataset shifts, such as changes in environments and object geometries, practical corruptions from sensor variations and weather conditions remain underexplored. In this work, we propose a novel online test-time adaptation framework for 3D detectors that effectively tackles these shifts, including a challenging cross-corruption scenario where cross-dataset shifts and corruptions co-occur. By leveraging long-term knowledge from previous test batches, our approach mitigates catastrophic forgetting and adapts effectively to diverse shifts. Specifically, we propose a Model Synergy (MOS) strategy that dynamically selects historical checkpoints with diverse knowledge and assembles them to best accommodate the current test batch. This assembly is directed by our proposed Synergy Weights (SW), which perform a weighted averaging of the selected checkpoints, minimizing redundancy in the composite model. The SWs are computed by evaluating the similarity of predicted bounding boxes on the test data and the independence of features between checkpoint pairs in the model bank. To maintain an efficient and informative model bank, we discard checkpoints with the lowest average SW scores, replacing them with newly updated models. Our method was rigorously tested against existing test-time adaptation strategies across three datasets and eight types of corruptions, demonstrating superior adaptability to dynamic scenes and conditions. Notably, it achieved a 67.3% improvement in a challenging cross-corruption scenario, offering a more comprehensive benchmark for adaptation. The source code will be made publicly available.
♻ ☆ Action Selection Learning for Multi-label Multi-view Action Recognition
Multi-label multi-view action recognition aims to recognize multiple concurrent or sequential actions from untrimmed videos captured by multiple cameras. Existing work has focused on multi-view action recognition in a narrow area with strong labels available, where the onset and offset of each action are labeled at the frame-level. This study focuses on real-world scenarios where cameras are distributed to capture a wide-range area with only weak labels available at the video-level. We propose the method named Multi-view Action Selection Learning (MultiASL), which leverages action selection learning to enhance view fusion by selecting the most useful information from different viewpoints. The proposed method includes a Multi-view Spatial-Temporal Transformer video encoder to extract spatial and temporal features from multi-viewpoint videos. Action Selection Learning is employed at the frame-level, using pseudo ground-truth obtained from weak labels at the video-level, to identify the most relevant frames for action recognition. Experiments in a real-world office environment using the MM-Office dataset demonstrate the superior performance of the proposed method compared to existing methods. The source code is available at https://github.com/thanhhff/MultiASL/.
comment: ACM Multimedia Asia 2024
Information Retrieval 20
☆ SIMformer: Single-Layer Vanilla Transformer Can Learn Free-Space Trajectory Similarity
Free-space trajectory similarity calculation, e.g., DTW, Hausdorff, and Frechet, often incur quadratic time complexity, thus learning-based methods have been proposed to accelerate the computation. The core idea is to train an encoder to transform trajectories into representation vectors and then compute vector similarity to approximate the ground truth. However, existing methods face dual challenges of effectiveness and efficiency: 1) they all utilize Euclidean distance to compute representation similarity, which leads to the severe curse of dimensionality issue -- reducing the distinguishability among representations and significantly affecting the accuracy of subsequent similarity search tasks; 2) most of them are trained in triplets manner and often necessitate additional information which downgrades the efficiency; 3) previous studies, while emphasizing the scalability in terms of efficiency, overlooked the deterioration of effectiveness when the dataset size grows. To cope with these issues, we propose a simple, yet accurate, fast, scalable model that only uses a single-layer vanilla transformer encoder as the feature extractor and employs tailored representation similarity functions to approximate various ground truth similarity measures. Extensive experiments demonstrate our model significantly mitigates the curse of dimensionality issue and outperforms the state-of-the-arts in effectiveness, efficiency, and scalability.
☆ Enhancing AI Accessibility in Veterinary Medicine: Linking Classifiers and Electronic Health Records
In the rapidly evolving landscape of veterinary healthcare, integrating machine learning (ML) clinical decision-making tools with electronic health records (EHRs) promises to improve diagnostic accuracy and patient care. However, the seamless integration of ML classifiers into existing EHRs in veterinary medicine is frequently hindered by the rigidity of EHR systems or the limited availability of IT resources. To address this shortcoming, we present Anna, a freely-available software solution that provides ML classifier results for EHR laboratory data in real-time.
☆ DiSCo Meets LLMs: A Unified Approach for Sparse Retrieval and Contextual Distillation in Conversational Search
Conversational Search (CS) is the task of retrieving relevant documents from a corpus within a conversational context, combining retrieval with conversational context modeling. With the explosion of Large Language Models (LLMs), the CS field has seen major improvements with LLMs rewriting user queries, accounting for conversational context. However, engaging LLMs at inference time harms efficiency. Current methods address this by distilling embeddings from human-rewritten queries to learn the context modeling task. Yet, these approaches predominantly focus on context modeling, and only treat the contrastive component of the retrieval task within a distillation-independent loss term. To address these limitations, we propose a new distillation method, as a relaxation of the previous objective, unifying retrieval and context modeling. We relax the existing training objectives by distilling similarity scores between conversations and documents, rather than relying solely on representation learning. Our proposed distillation objective allows for more freedom in the representation space and leverages the contrastive nature of document relevance. Through experiments on Learned Sparse Retrieval (LSR) across 5 CS datasets, our approach demonstrates substantial improvements in both in-domain and out-of-domain retrieval performance, outperforming state-of-the-art with gains of up to 6 points in recall for out-of-domain datasets. Additionally, through the relaxation of the objective, we propose a multi-teacher distillation, using multiple LLMs as teachers, yielding additional gains, and outperforming the teachers themselves in in-domain experiments. Finally, analysis of the sparsity of the models reveals that our distillation allows for better control over the sparsity of the trained models.
☆ RAG-ConfusionQA: A Benchmark for Evaluating LLMs on Confusing Questions
Conversational AI agents use Retrieval Augmented Generation (RAG) to provide verifiable document-grounded responses to user inquiries. However, many natural questions do not have good answers: about 25\% contain false assumptions~\cite{Yu2023:CREPE}, and over 50\% are ambiguous~\cite{Min2020:AmbigQA}. RAG agents need high-quality data to improve their responses to confusing questions. This paper presents a novel synthetic data generation method to efficiently create a diverse set of context-grounded confusing questions from a given document corpus. We conduct an empirical comparative evaluation of several large language models as RAG agents to measure the accuracy of confusion detection and appropriate response generation. We contribute a benchmark dataset to the public domain.
comment: under review
☆ SPFresh: Incremental In-Place Update for Billion-Scale Vector Search SOSP 23
Approximate Nearest Neighbor Search (ANNS) is now widely used in various applications, ranging from information retrieval, question answering, and recommendation, to search for similar high-dimensional vectors. As the amount of vector data grows continuously, it becomes important to support updates to vector index, the enabling technique that allows for efficient and accurate ANNS on vectors. Because of the curse of high dimensionality, it is often costly to identify the right neighbors of a single new vector, a necessary process for index update. To amortize update costs, existing systems maintain a secondary index to accumulate updates, which are merged by the main index by global rebuilding the entire index periodically. However, this approach has high fluctuations of search latency and accuracy, not even to mention that it requires substantial resources and is extremely time-consuming for rebuilds. We introduce SPFresh, a system that supports in-place vector updates. At the heart of SPFresh is LIRE, a lightweight incremental rebalancing protocol to split vector partitions and reassign vectors in the nearby partitions to adapt to data distribution shift. LIRE achieves low-overhead vector updates by only reassigning vectors at the boundary between partitions, where in a high-quality vector index the amount of such vectors are deemed small. With LIRE, SPFresh provides superior query latency and accuracy to solutions based on global rebuild, with only 1% of DRAM and less than 10% cores needed at the peak compared to the state-of-the-art, in a billion scale vector index with 1% of daily vector update rate.
comment: SOSP 23
☆ ChartifyText: Automated Chart Generation from Data-Involved Texts via LLM
Text documents with numerical values involved are widely used in various applications such as scientific research, economy, public health and journalism. However, it is difficult for readers to quickly interpret such data-involved texts and gain deep insights. To fill this research gap, this work aims to automatically generate charts to accurately convey the underlying data and ideas to readers, which is essentially a challenging task. The challenges originate from text ambiguities, intrinsic sparsity and uncertainty of data in text documents, and subjective sentiment differences. Specifically, we propose ChartifyText, a novel fully-automated approach that leverages Large Language Models (LLMs) to convert complex data-involved texts to expressive charts. It consists of two major modules: tabular data inference and expressive chart generation. The tabular data inference module employs systematic prompt engineering to guide the LLM (e.g., GPT-4) to infer table data, where data ranges, uncertainties, missing data values and corresponding subjective sentiments are explicitly considered. The expressive chart generation module augments standard charts with intuitive visual encodings and concise texts to accurately convey the underlying data and insights. We extensively evaluate the effectiveness of ChartifyText on real-world data-involved text documents through case studies, in-depth interviews with three visualization experts, and a carefully-designed user study with 15 participants. The results demonstrate the usefulness and effectiveness of ChartifyText in helping readers efficiently and effectively make sense of data-involved texts.
☆ Graph Neural Patching for Cold-Start Recommendations
The cold start problem in recommender systems remains a critical challenge. Current solutions often train hybrid models on auxiliary data for both cold and warm users/items, potentially degrading the experience for the latter. This drawback limits their viability in practical scenarios where the satisfaction of existing warm users/items is paramount. Although graph neural networks (GNNs) excel at warm recommendations by effective collaborative signal modeling, they haven't been effectively leveraged for the cold-start issue within a user-item graph, which is largely due to the lack of initial connections for cold user/item entities. Addressing this requires a GNN adept at cold-start recommendations without sacrificing performance for existing ones. To this end, we introduce Graph Neural Patching for Cold-Start Recommendations (GNP), a customized GNN framework with dual functionalities: GWarmer for modeling collaborative signal on existing warm users/items and Patching Networks for simulating and enhancing GWarmer's performance on cold-start recommendations. Extensive experiments on three benchmark datasets confirm GNP's superiority in recommending both warm and cold users/items.
comment: 13 pages, accepted by Australasian Database Conference 2024. arXiv admin note: substantial text overlap with arXiv:2209.12215
☆ Personalized Image Generation with Large Multimodal Models
Personalized content filtering, such as recommender systems, has become a critical infrastructure to alleviate information overload. However, these systems merely filter existing content and are constrained by its limited diversity, making it difficult to meet users' varied content needs. To address this limitation, personalized content generation has emerged as a promising direction with broad applications. Nevertheless, most existing research focuses on personalized text generation, with relatively little attention given to personalized image generation. The limited work in personalized image generation faces challenges in accurately capturing users' visual preferences and needs from noisy user-interacted images and complex multimodal instructions. Worse still, there is a lack of supervised data for training personalized image generation models. To overcome the challenges, we propose a Personalized Image Generation Framework named Pigeon, which adopts exceptional large multimodal models with three dedicated modules to capture users' visual preferences and needs from noisy user history and multimodal instructions. To alleviate the data scarcity, we introduce a two-stage preference alignment scheme, comprising masked preference reconstruction and pairwise preference alignment, to align Pigeon with the personalized image generation task. We apply Pigeon to personalized sticker and movie poster generation, where extensive quantitative results and human evaluation highlight its superiority over various generative baselines.
☆ Optimizing Retrieval-Augmented Generation with Elasticsearch for Enhanced Question-Answering Systems
This study aims to improve the accuracy and quality of large-scale language models (LLMs) in answering questions by integrating Elasticsearch into the Retrieval Augmented Generation (RAG) framework. The experiment uses the Stanford Question Answering Dataset (SQuAD) version 2.0 as the test dataset and compares the performance of different retrieval methods, including traditional methods based on keyword matching or semantic similarity calculation, BM25-RAG and TF-IDF- RAG, and the newly proposed ES-RAG scheme. The results show that ES-RAG not only has obvious advantages in retrieval efficiency but also performs well in key indicators such as accuracy, which is 0.51 percentage points higher than TF-IDF-RAG. In addition, Elasticsearch's powerful search capabilities and rich configuration options enable the entire question-answering system to better handle complex queries and provide more flexible and efficient responses based on the diverse needs of users. Future research directions can further explore how to optimize the interaction mechanism between Elasticsearch and LLM, such as introducing higher-level semantic understanding and context-awareness capabilities, to achieve a more intelligent and humanized question-answering experience.
☆ Towards Robust Transcription: Exploring Noise Injection Strategies for Training Data Augmentation
Recent advancements in Automatic Piano Transcription (APT) have significantly improved system performance, but the impact of noisy environments on the system performance remains largely unexplored. This study investigates the impact of white noise at various Signal-to-Noise Ratio (SNR) levels on state-of-the-art APT models and evaluates the performance of the Onsets and Frames model when trained on noise-augmented data. We hope this research provides valuable insights as preliminary work toward developing transcription models that maintain consistent performance across a range of acoustic conditions.
comment: Accepted to the Late-Breaking Demo Session of the 25th International Society for Music Information Retrieval (ISMIR) Conference, 2024
☆ The S2 Hierarchical Discrete Global Grid as a Nexus for Data Representation, Integration, and Querying Across Geospatial Knowledge Graphs
Geospatial Knowledge Graphs (GeoKGs) have become integral to the growing field of Geospatial Artificial Intelligence. Initiatives like the U.S. National Science Foundation's Open Knowledge Network program aim to create an ecosystem of nation-scale, cross-disciplinary GeoKGs that provide AI-ready geospatial data aligned with FAIR principles. However, building this infrastructure presents key challenges, including 1) managing large volumes of data, 2) the computational complexity of discovering topological relations via SPARQL, and 3) conflating multi-scale raster and vector data. Discrete Global Grid Systems (DGGS) help tackle these issues by offering efficient data integration and representation strategies. The KnowWhereGraph utilizes Google's S2 Geometry -- a DGGS framework -- to enable efficient multi-source data processing, qualitative spatial querying, and cross-graph integration. This paper outlines the implementation of S2 within KnowWhereGraph, emphasizing its role in topologically enriching and semantically compressing data. Ultimately, this work demonstrates the potential of DGGS frameworks, particularly S2, for building scalable GeoKGs.
♻ ☆ EasyRec: Simple yet Effective Language Models for Recommendation
Deep neural networks have become a powerful technique for learning representations from user-item interaction data in collaborative filtering (CF) for recommender systems. However, many existing methods heavily rely on unique user and item IDs, which limits their ability to perform well in practical zero-shot learning scenarios where sufficient training data may be unavailable. Inspired by the success of language models (LMs) and their strong generalization capabilities, a crucial question arises: How can we harness the potential of language models to empower recommender systems and elevate its generalization capabilities to new heights? In this study, we propose EasyRec - an effective and easy-to-use approach that seamlessly integrates text-based semantic understanding with collaborative signals. EasyRec employs a text-behavior alignment framework, which combines contrastive learning with collaborative language model tuning, to ensure a strong alignment between the text-enhanced semantic space and the collaborative behavior information. Extensive empirical evaluations across diverse real-world datasets demonstrate the superior performance of EasyRec compared to state-of-the-art alternative models, particularly in the challenging text-based zero-shot recommendation scenarios. Furthermore, the study highlights the potential of seamlessly integrating EasyRec as a plug-and-play component into text-enhanced collaborative filtering frameworks, thereby empowering existing recommender systems to elevate their recommendation performance and adapt to the evolving user preferences in dynamic environments. For better result reproducibility of our EasyRec framework, the model implementation details, source code, and datasets are available at the link: https://github.com/HKUDS/EasyRec.
♻ ☆ Improving Retrieval in Sponsored Search by Leveraging Query Context Signals EMNLP 2024
Accurately retrieving relevant bid keywords for user queries is critical in Sponsored Search but remains challenging, particularly for short, ambiguous queries. Existing dense and generative retrieval models often fail to capture nuanced user intent in these cases. To address this, we propose an approach to enhance query understanding by augmenting queries with rich contextual signals derived from web search results and large language models, stored in an online cache. Specifically, we use web search titles and snippets to ground queries in real-world information and utilize GPT-4 to generate query rewrites and explanations that clarify user intent. These signals are efficiently integrated through a Fusion-in-Decoder based Unity architecture, enabling both dense and generative retrieval with serving costs on par with traditional context-free models. To address scenarios where context is unavailable in the cache, we introduce context glancing, a curriculum learning strategy that improves model robustness and performance even without contextual signals during inference. Extensive offline experiments demonstrate that our context-aware approach substantially outperforms context-free models. Furthermore, online A/B testing on a prominent search engine across 160+ countries shows significant improvements in user engagement and revenue.
comment: Accepted to EMNLP 2024 Industry Track. 10 pages, 10 tables, 1 figure
♻ ☆ Revisiting BPR: A Replicability Study of a Common Recommender System Baseline RecSys '24
Bayesian Personalized Ranking (BPR), a collaborative filtering approach based on matrix factorization, frequently serves as a benchmark for recommender systems research. However, numerous studies often overlook the nuances of BPR implementation, claiming that it performs worse than newly proposed methods across various tasks. In this paper, we thoroughly examine the features of the BPR model, indicating their impact on its performance, and investigate open-source BPR implementations. Our analysis reveals inconsistencies between these implementations and the original BPR paper, leading to a significant decrease in performance of up to 50% for specific implementations. Furthermore, through extensive experiments on real-world datasets under modern evaluation settings, we demonstrate that with proper tuning of its hyperparameters, the BPR model can achieve performance levels close to state-of-the-art methods on the top-n recommendation tasks and even outperform them on specific datasets. Specifically, on the Million Song Dataset, the BPR model with hyperparameters tuning statistically significantly outperforms Mult-VAE by 10% in NDCG@100 with binary relevance function.
comment: This paper is accepted at the Reproducibility track of the ACM RecSys '24 conference
♻ ☆ Generate and Instantiate What You Prefer: Text-Guided Diffusion for Sequential Recommendation
Recent advancements in generative recommendation systems, particularly in the realm of sequential recommendation tasks, have shown promise in enhancing generalization to new items. Among these approaches, diffusion-based generative recommendation has emerged as an effective tool, leveraging its ability to capture data distributions and generate high-quality samples. Despite effectiveness, two primary challenges have been identified: 1) the lack of consistent modeling of data distribution for oracle items; and 2) the difficulty in scaling to more informative control signals beyond historical interactions. These issues stem from the uninformative nature of ID embeddings, which necessitate random initialization and limit the incorporation of additional control signals. To address these limitations, we propose iDreamRec to involve more concrete prior knowledge to establish item embeddings, particularly through detailed item text descriptions and advanced Text Embedding Models (TEM). More importantly, by converting item descriptions into embeddings aligned with TEM, we enable the integration of intention instructions as control signals to guide the generation of oracle items. Experimental results on four datasets demonstrate that iDreamRec not only outperforms existing diffusion-based generative recommenders but also facilitates the incorporation of intention instructions for more precise and effective recommendation generation.
♻ ☆ Conversational Recommender System and Large Language Model Are Made for Each Other in E-commerce Pre-sales Dialogue EMNLP 2023
E-commerce pre-sales dialogue aims to understand and elicit user needs and preferences for the items they are seeking so as to provide appropriate recommendations. Conversational recommender systems (CRSs) learn user representation and provide accurate recommendations based on dialogue context, but rely on external knowledge. Large language models (LLMs) generate responses that mimic pre-sales dialogues after fine-tuning, but lack domain-specific knowledge for accurate recommendations. Intuitively, the strengths of LLM and CRS in E-commerce pre-sales dialogues are complementary, yet no previous work has explored this. This paper investigates the effectiveness of combining LLM and CRS in E-commerce pre-sales dialogues, proposing two collaboration methods: CRS assisting LLM and LLM assisting CRS. We conduct extensive experiments on a real-world dataset of Ecommerce pre-sales dialogues. We analyze the impact of two collaborative approaches with two CRSs and two LLMs on four tasks of Ecommerce pre-sales dialogue. We find that collaborations between CRS and LLM can be very effective in some cases.
comment: EMNLP 2023 Findings
♻ ☆ Starbucks: Improved Training for 2D Matryoshka Embeddings
Effective approaches that can scale embedding model depth (i.e. layers) and embedding size allow for the creation of models that are highly scalable across different computational resources and task requirements. While the recently proposed 2D Matryoshka training approach can efficiently produce a single embedding model such that its sub-layers and sub-dimensions can measure text similarity, its effectiveness is significantly worse than if smaller models were trained separately. To address this issue, we propose Starbucks, a new training strategy for Matryoshka-like embedding models, which encompasses both the fine-tuning and pre-training phases. For the fine-tuning phase, we discover that, rather than sampling a random sub-layer and sub-dimensions for each training steps, providing a fixed list of layer-dimension pairs, from small size to large sizes, and computing the loss across all pairs significantly improves the effectiveness of 2D Matryoshka embedding models, bringing them on par with their separately trained counterparts. To further enhance performance, we introduce a new pre-training strategy, which applies masked autoencoder language modelling to sub-layers and sub-dimensions during pre-training, resulting in a stronger backbone for subsequent fine-tuning of the embedding model. Experimental results on both semantic text similarity and retrieval benchmarks demonstrate that the proposed pre-training and fine-tuning strategies significantly improved the effectiveness over 2D Matryoshka models, enabling Starbucks models to perform more efficiently and effectively than separately trained models.
♻ ☆ Graph Neural Network Enhanced Retrieval for Question Answering of LLMs
Retrieval augmented generation has revolutionized large language model (LLM) outputs by providing factual supports. Nevertheless, it struggles to capture all the necessary knowledge for complex reasoning questions. Existing retrieval methods typically divide reference documents into passages, treating them in isolation. These passages, however, are often interrelated, such as passages that are contiguous or share the same keywords. Therefore, it is crucial to recognize such relatedness for enhancing the retrieval process. In this paper, we propose a novel retrieval method, called GNN-Ret, which leverages graph neural networks (GNNs) to enhance retrieval by exploiting the relatedness between passages. Specifically, we first construct a graph of passages by connecting passages that are structure-related or keyword-related. A graph neural network (GNN) is then leveraged to exploit the relationships between passages and improve the retrieval of supporting passages. Furthermore, we extend our method to handle multi-hop reasoning questions using a recurrent graph neural network (RGNN), named RGNN-Ret. At each step, RGNN-Ret integrates the graphs of passages from previous steps, thereby enhancing the retrieval of supporting passages. Extensive experiments on benchmark datasets demonstrate that GNN-Ret achieves higher accuracy for question answering with a single query of LLMs than strong baselines that require multiple queries, and RGNN-Ret further improves accuracy and achieves state-of-the-art performance, with up to 10.4% accuracy improvement on the 2WikiMQA dataset.
comment: Under review
♻ ☆ FINED: Feed Instance-Wise Information Need with Essential and Disentangled Parametric Knowledge from the Past
Recommender models play a vital role in various industrial scenarios, while often faced with the catastrophic forgetting problem caused by the fast shifting data distribution. To alleviate this problem, a common approach is to reuse knowledge from the historical data. However, preserving the vast and fast-accumulating data is hard, which causes dramatic storage overhead. Memorizing old data through a parametric knowledge base is then proposed, which compresses the vast amount of raw data into model parameters. Despite the flexibility, how to improve the memorization and generalization capabilities of the parametric knowledge base and suit the flexible information need of each instance are challenging. In this paper, we propose FINED to Feed INstance-wise information need with Essential and Disentangled parametric knowledge from past data for recommendation enhancement. Concretely, we train a knowledge extractor that extracts knowledge patterns of arbitrary order from past data and a knowledge encoder that memorizes the arbitrary order patterns, which serves as the retrieval key generator and memory network respectively in the following knowledge reusing phase. The whole process is regularized by the proposed two constraints, which improve the capabilities of the parametric knowledge base without increasing the size of it. The essential principle helps to compress the input into representative vectors that capture the task-relevant information and filter out the noisy information. The disentanglement principle reduces the redundancy of stored information and pushes the knowledge base to focus on capturing the disentangled invariant patterns. These two rules together promote rational compression of information for robust and generalized knowledge representations. Extensive experiments on two datasets justify the effectiveness of the proposed method.
♻ ☆ Retrieval-Enhanced Machine Learning: Synthesis and Opportunities
In the field of language modeling, models augmented with retrieval components have emerged as a promising solution to address several challenges faced in the natural language processing (NLP) field, including knowledge grounding, interpretability, and scalability. Despite the primary focus on NLP, we posit that the paradigm of retrieval-enhancement can be extended to a broader spectrum of machine learning (ML) such as computer vision, time series prediction, and computational biology. Therefore, this work introduces a formal framework of this paradigm, Retrieval-Enhanced Machine Learning (REML), by synthesizing the literature in various domains in ML with consistent notations which is missing from the current literature. Also, we found that while a number of studies employ retrieval components to augment their models, there is a lack of integration with foundational Information Retrieval (IR) research. We bridge this gap between the seminal IR research and contemporary REML studies by investigating each component that comprises the REML framework. Ultimately, the goal of this work is to equip researchers across various disciplines with a comprehensive, formally structured framework of retrieval-enhanced models, thereby fostering interdisciplinary future research.
Machine Learning 200
Self-supervised contrastive learning performs non-linear system identification
Self-supervised learning (SSL) approaches have brought tremendous success across many tasks and domains. It has been argued that these successes can be attributed to a link between SSL and identifiable representation learning: Temporal structure and auxiliary variables ensure that latent representations are related to the true underlying generative factors of the data. Here, we deepen this connection and show that SSL can perform system identification in latent space. We propose DynCL, a framework to uncover linear, switching linear and non-linear dynamics under a non-linear observation model, give theoretical guarantees and validate them empirically.
☆ Decomposing The Dark Matter of Sparse Autoencoders
Sparse autoencoders (SAEs) are a promising technique for decomposing language model activations into interpretable linear features. However, current SAEs fall short of completely explaining model performance, resulting in "dark matter": unexplained variance in activations. This work investigates dark matter as an object of study in its own right. Surprisingly, we find that much of SAE dark matter--about half of the error vector itself and >90% of its norm--can be linearly predicted from the initial activation vector. Additionally, we find that the scaling behavior of SAE error norms at a per token level is remarkably predictable: larger SAEs mostly struggle to reconstruct the same contexts as smaller SAEs. We build on the linear representation hypothesis to propose models of activations that might lead to these observations, including postulating a new type of "introduced error"; these insights imply that the part of the SAE error vector that cannot be linearly predicted ("nonlinear" error) might be fundamentally different from the linearly predictable component. To validate this hypothesis, we empirically analyze nonlinear SAE error and show that 1) it contains fewer not yet learned features, 2) SAEs trained on it are quantitatively worse, 3) it helps predict SAE per-token scaling behavior, and 4) it is responsible for a proportional amount of the downstream increase in cross entropy loss when SAE activations are inserted into the model. Finally, we examine two methods to reduce nonlinear SAE error at a fixed sparsity: inference time gradient pursuit, which leads to a very slight decrease in nonlinear error, and linear transformations from earlier layer SAE outputs, which leads to a larger reduction.
comment: Code at https://github.com/JoshEngels/SAE-Dark-Matter
☆ Stochastic Gradient Descent Jittering for Inverse Problems: Alleviating the Accuracy-Robustness Tradeoff
Inverse problems aim to reconstruct unseen data from corrupted or perturbed measurements. While most work focuses on improving reconstruction quality, generalization accuracy and robustness are equally important, especially for safety-critical applications. Model-based architectures (MBAs), such as loop unrolling methods, are considered more interpretable and achieve better reconstructions. Empirical evidence suggests that MBAs are more robust to perturbations than black-box solvers, but the accuracy-robustness tradeoff in MBAs remains underexplored. In this work, we propose a simple yet effective training scheme for MBAs, called SGD jittering, which injects noise iteration-wise during reconstruction. We theoretically demonstrate that SGD jittering not only generalizes better than the standard mean squared error training but is also more robust to average-case attacks. We validate SGD jittering using denoising toy examples, seismic deconvolution, and single-coil MRI reconstruction. The proposed method achieves cleaner reconstructions for out-of-distribution data and demonstrates enhanced robustness to adversarial attacks.
☆ DiscoGraMS: Enhancing Movie Screen-Play Summarization using Movie Character-Aware Discourse Graph
Summarizing movie screenplays presents a unique set of challenges compared to standard document summarization. Screenplays are not only lengthy, but also feature a complex interplay of characters, dialogues, and scenes, with numerous direct and subtle relationships and contextual nuances that are difficult for machine learning models to accurately capture and comprehend. Recent attempts at screenplay summarization focus on fine-tuning transformer-based pre-trained models, but these models often fall short in capturing long-term dependencies and latent relationships, and frequently encounter the "lost in the middle" issue. To address these challenges, we introduce DiscoGraMS, a novel resource that represents movie scripts as a movie character-aware discourse graph (CaD Graph). This approach is well-suited for various downstream tasks, such as summarization, question-answering, and salience detection. The model aims to preserve all salient information, offering a more comprehensive and faithful representation of the screenplay's content. We further explore a baseline method that combines the CaD Graph with the corresponding movie script through a late fusion of graph and text modalities, and we present very initial promising results.
☆ Online Reinforcement Learning with Passive Memory
This paper considers an online reinforcement learning algorithm that leverages pre-collected data (passive memory) from the environment for online interaction. We show that using passive memory improves performance and further provide theoretical guarantees for regret that turns out to be near-minimax optimal. Results show that the quality of passive memory determines sub-optimality of the incurred regret. The proposed approach and results hold in both continuous and discrete state-action spaces.
☆ A Large Language Model-Driven Reward Design Framework via Dynamic Feedback for Reinforcement Learning
Large Language Models (LLMs) have shown significant potential in designing reward functions for Reinforcement Learning (RL) tasks. However, obtaining high-quality reward code often involves human intervention, numerous LLM queries, or repetitive RL training. To address these issues, we propose CARD, a LLM-driven Reward Design framework that iteratively generates and improves reward function code. Specifically, CARD includes a Coder that generates and verifies the code, while a Evaluator provides dynamic feedback to guide the Coder in improving the code, eliminating the need for human feedback. In addition to process feedback and trajectory feedback, we introduce Trajectory Preference Evaluation (TPE), which evaluates the current reward function based on trajectory preferences. If the code fails the TPE, the Evaluator provides preference feedback, avoiding RL training at every iteration and making the reward function better aligned with the task objective. Empirical results on Meta-World and ManiSkill2 demonstrate that our method achieves an effective balance between task performance and token efficiency, outperforming or matching the baselines across all tasks. On 10 out of 12 tasks, CARD shows better or comparable performance to policies trained with expert-designed rewards, and our method even surpasses the oracle on 3 tasks.
☆ Harnessing Causality in Reinforcement Learning With Bagged Decision Times
We consider reinforcement learning (RL) for a class of problems with bagged decision times. A bag contains a finite sequence of consecutive decision times. The transition dynamics are non-Markovian and non-stationary within a bag. Further, all actions within a bag jointly impact a single reward, observed at the end of the bag. Our goal is to construct an online RL algorithm to maximize the discounted sum of the bag-specific rewards. To handle non-Markovian transitions within a bag, we utilize an expert-provided causal directed acyclic graph (DAG). Based on the DAG, we construct the states as a dynamical Bayesian sufficient statistic of the observed history, which results in Markovian state transitions within and across bags. We then frame this problem as a periodic Markov decision process (MDP) that allows non-stationarity within a period. An online RL algorithm based on Bellman-equations for stationary MDPs is generalized to handle periodic MDPs. To justify the proposed RL algorithm, we show that our constructed state achieves the maximal optimal value function among all state constructions for a periodic MDP. Further we prove the Bellman optimality equations for periodic MDPs. We evaluate the proposed method on testbed variants, constructed with real data from a mobile health clinical trial.
☆ Bridging the Training-Inference Gap in LLMs by Leveraging Self-Generated Tokens
Language models are often trained to maximize the likelihood of the next token given past tokens in the training dataset. However, during inference time, they are utilized differently, generating text sequentially and auto-regressively by using previously generated tokens as input to predict the next one. Marginal differences in predictions at each step can cascade over successive steps, resulting in different distributions from what the models were trained for and potentially leading to unpredictable behavior. This paper proposes two simple approaches based on model own generation to address this discrepancy between the training and inference time. Our first approach is Batch-Scheduled Sampling, where, during training, we stochastically choose between the ground-truth token from the dataset and the model's own generated token as input to predict the next token. This is done in an offline manner, modifying the context window by interleaving ground-truth tokens with those generated by the model. Our second approach is Reference-Answer-based Correction, where we explicitly incorporate a self-correction capability into the model during training. This enables the model to effectively self-correct the gaps between the generated sequences and the ground truth data without relying on an external oracle model. By incorporating our proposed strategies during training, we have observed an overall improvement in performance compared to baseline methods, as demonstrated by our extensive experiments using summarization, general question-answering, and math question-answering tasks.
☆ EvoPress: Towards Optimal Dynamic Model Compression via Evolutionary Search
The high computational costs of large language models (LLMs) have led to a flurry of research on LLM compression, via methods such as quantization, sparsification, or structured pruning. A new frontier in this area is given by \emph{dynamic, non-uniform} compression methods, which adjust the compression levels (e.g., sparsity) per-block or even per-layer in order to minimize accuracy loss, while guaranteeing a global compression threshold. Yet, current methods rely on heuristics for identifying the "importance" of a given layer towards the loss, based on assumptions such as \emph{error monotonicity}, i.e. that the end-to-end model compression error is proportional to the sum of layer-wise errors. In this paper, we revisit this area, and propose a new and general approach for dynamic compression that is provably optimal in a given input range. We begin from the motivating observation that, in general, \emph{error monotonicity does not hold for LLMs}: compressed models with lower sum of per-layer errors can perform \emph{worse} than models with higher error sums. To address this, we propose a new general evolutionary framework for dynamic LLM compression called EvoPress, which has provable convergence, and low sample and evaluation complexity. We show that these theoretical guarantees lead to highly competitive practical performance for dynamic compression of Llama, Mistral and Phi models. Via EvoPress, we set new state-of-the-art results across all compression approaches: structural pruning (block/layer dropping), unstructured sparsity, as well as quantization with dynamic bitwidths. Our code is available at https://github.com/IST-DASLab/EvoPress.
☆ HR-Bandit: Human-AI Collaborated Linear Recourse Bandit
Human doctors frequently recommend actionable recourses that allow patients to modify their conditions to access more effective treatments. Inspired by such healthcare scenarios, we propose the Recourse Linear UCB ($\textsf{RLinUCB}$) algorithm, which optimizes both action selection and feature modifications by balancing exploration and exploitation. We further extend this to the Human-AI Linear Recourse Bandit ($\textsf{HR-Bandit}$), which integrates human expertise to enhance performance. $\textsf{HR-Bandit}$ offers three key guarantees: (i) a warm-start guarantee for improved initial performance, (ii) a human-effort guarantee to minimize required human interactions, and (iii) a robustness guarantee that ensures sublinear regret even when human decisions are suboptimal. Empirical results, including a healthcare case study, validate its superior performance against existing benchmarks.
comment: 18 pages
☆ Convergence of Manifold Filter-Combine Networks NeurIPS
In order to better understand manifold neural networks (MNNs), we introduce Manifold Filter-Combine Networks (MFCNs). The filter-combine framework parallels the popular aggregate-combine paradigm for graph neural networks (GNNs) and naturally suggests many interesting families of MNNs which can be interpreted as the manifold analog of various popular GNNs. We then propose a method for implementing MFCNs on high-dimensional point clouds that relies on approximating the manifold by a sparse graph. We prove that our method is consistent in the sense that it converges to a continuum limit as the number of data points tends to infinity.
comment: Accepted to NeurIPS Workshop on Symmetry and Geometry in Neural Representations (Extended Abstract Track)
☆ Parallel Backpropagation for Inverse of a Convolution with Application to Normalizing Flows
Inverse of an invertible convolution is an important operation that comes up in Normalizing Flows, Image Deblurring, etc. The naive algorithm for backpropagation of this operation using Gaussian elimination has running time $O(n^3)$ where $n$ is the number of pixels in the image. We give a fast parallel backpropagation algorithm with running time $O(\sqrt{n})$ for a square image and provide a GPU implementation of the same. Inverse Convolutions are usually used in Normalizing Flows in the sampling pass, making them slow. We propose to use Inverse Convolutions in the forward (image to latent vector) pass of the Normalizing flow. Since the sampling pass is the inverse of the forward pass, it will use convolutions only, resulting in efficient sampling times. We use our parallel backpropagation algorithm for optimizing the inverse convolution layer resulting in fast training times also. We implement this approach in various Normalizing Flow backbones, resulting in our Inverse-Flow models. We benchmark Inverse-Flow on standard datasets and show significantly improved sampling times with similar bits per dimension compared to previous models.
comment: Preprint
☆ On the Regularization of Learnable Embeddings for Time Series Processing
In processing multiple time series, accounting for the individual features of each sequence can be challenging. To address this, modern deep learning methods for time series analysis combine a shared (global) model with local layers, specific to each time series, often implemented as learnable embeddings. Ideally, these local embeddings should encode meaningful representations of the unique dynamics of each sequence. However, when these are learned end-to-end as parameters of a forecasting model, they may end up acting as mere sequence identifiers. Shared processing blocks may then become reliant on such identifiers, limiting their transferability to new contexts. In this paper, we address this issue by investigating methods to regularize the learning of local learnable embeddings for time series processing. Specifically, we perform the first extensive empirical study on the subject and show how such regularizations consistently improve performance in widely adopted architectures. Furthermore, we show that methods preventing the co-adaptation of local and global parameters are particularly effective in this context. This hypothesis is validated by comparing several methods preventing the downstream models from relying on sequence identifiers, going as far as completely resetting the embeddings during training. The obtained results provide an important contribution to understanding the interplay between learnable local parameters and shared processing layers: a key challenge in modern time series processing models and a step toward developing effective foundation models for time series.
☆ SIMformer: Single-Layer Vanilla Transformer Can Learn Free-Space Trajectory Similarity
Free-space trajectory similarity calculation, e.g., DTW, Hausdorff, and Frechet, often incur quadratic time complexity, thus learning-based methods have been proposed to accelerate the computation. The core idea is to train an encoder to transform trajectories into representation vectors and then compute vector similarity to approximate the ground truth. However, existing methods face dual challenges of effectiveness and efficiency: 1) they all utilize Euclidean distance to compute representation similarity, which leads to the severe curse of dimensionality issue -- reducing the distinguishability among representations and significantly affecting the accuracy of subsequent similarity search tasks; 2) most of them are trained in triplets manner and often necessitate additional information which downgrades the efficiency; 3) previous studies, while emphasizing the scalability in terms of efficiency, overlooked the deterioration of effectiveness when the dataset size grows. To cope with these issues, we propose a simple, yet accurate, fast, scalable model that only uses a single-layer vanilla transformer encoder as the feature extractor and employs tailored representation similarity functions to approximate various ground truth similarity measures. Extensive experiments demonstrate our model significantly mitigates the curse of dimensionality issue and outperforms the state-of-the-arts in effectiveness, efficiency, and scalability.
☆ Enhancing AI Accessibility in Veterinary Medicine: Linking Classifiers and Electronic Health Records
In the rapidly evolving landscape of veterinary healthcare, integrating machine learning (ML) clinical decision-making tools with electronic health records (EHRs) promises to improve diagnostic accuracy and patient care. However, the seamless integration of ML classifiers into existing EHRs in veterinary medicine is frequently hindered by the rigidity of EHR systems or the limited availability of IT resources. To address this shortcoming, we present Anna, a freely-available software solution that provides ML classifier results for EHR laboratory data in real-time.
☆ syren-new: Precise formulae for the linear and nonlinear matter power spectra with massive neutrinos and dynamical dark energy
Current and future large scale structure surveys aim to constrain the neutrino mass and the equation of state of dark energy. We aim to construct accurate and interpretable symbolic approximations to the linear and nonlinear matter power spectra as a function of cosmological parameters in extended $\Lambda$CDM models which contain massive neutrinos and non-constant equations of state for dark energy. This constitutes an extension of the syren-halofit emulators to incorporate these two effects, which we call syren-new (SYmbolic-Regression-ENhanced power spectrum emulator with NEutrinos and $W_0-w_a$). We also obtain a simple approximation to the derived parameter $\sigma_8$ as a function of the cosmological parameters for these models. Our results for the linear power spectrum are designed to emulate CLASS, whereas for the nonlinear case we aim to match the results of EuclidEmulator2. We compare our results to existing emulators and $N$-body simulations. Our analytic emulators for $\sigma_8$, the linear and nonlinear power spectra achieve root mean squared errors of 0.1%, 0.3% and 1.3%, respectively, across a wide range of cosmological parameters, redshifts and wavenumbers. We verify that emulator-related discrepancies are subdominant compared to observational errors and other modelling uncertainties when computing shear power spectra for LSST-like surveys. Our expressions have similar accuracy to existing (numerical) emulators, but are at least an order of magnitude faster, both on a CPU and GPU. Our work greatly improves the accuracy, speed and range of applicability of current symbolic approximations to the linear and nonlinear matter power spectra. We provide publicly available code for all symbolic approximations found.
comment: 18 pages, 15 figures
☆ JAMUN: Transferable Molecular Conformational Ensemble Generation with Walk-Jump Sampling
Conformational ensembles of protein structures are immensely important both to understanding protein function, and for drug discovery in novel modalities such as cryptic pockets. Current techniques for sampling ensembles are computationally inefficient, or do not transfer to systems outside their training data. We present walk-Jump Accelerated Molecular ensembles with Universal Noise (JAMUN), a step towards the goal of efficiently sampling the Boltzmann distribution of arbitrary proteins. By extending Walk-Jump Sampling to point clouds, JAMUN enables ensemble generation at orders of magnitude faster rates than traditional molecular dynamics or state-of-the-art ML methods. Further, JAMUN is able to predict the stable basins of small peptides that were not seen during training.
☆ Benchmarking Deep Reinforcement Learning for Navigation in Denied Sensor Environments
Deep Reinforcement learning (DRL) is used to enable autonomous navigation in unknown environments. Most research assume perfect sensor data, but real-world environments may contain natural and artificial sensor noise and denial. Here, we present a benchmark of both well-used and emerging DRL algorithms in a navigation task with configurable sensor denial effects. In particular, we are interested in comparing how different DRL methods (e.g. model-free PPO vs. model-based DreamerV3) are affected by sensor denial. We show that DreamerV3 outperforms other methods in the visual end-to-end navigation task with a dynamic goal - and other methods are not able to learn this. Furthermore, DreamerV3 generally outperforms other methods in sensor-denied environments. In order to improve robustness, we use adversarial training and demonstrate an improved performance in denied environments, although this generally comes with a performance cost on the vanilla environments. We anticipate this benchmark of different DRL methods and the usage of adversarial training to be a starting point for the development of more elaborate navigation strategies that are capable of dealing with uncertain and denied sensor readings.
comment: 31 pages, 19 figures. For associated code, see https://github.com/mazqtpopx/cranfield-navigation-gym
☆ Asymptotically Optimal Change Detection for Unnormalized Pre- and Post-Change Distributions
This paper addresses the problem of detecting changes when only unnormalized pre- and post-change distributions are accessible. This situation happens in many scenarios in physics such as in ferromagnetism, crystallography, magneto-hydrodynamics, and thermodynamics, where the energy models are difficult to normalize. Our approach is based on the estimation of the Cumulative Sum (CUSUM) statistics, which is known to produce optimal performance. We first present an intuitively appealing approximation method. Unfortunately, this produces a biased estimator of the CUSUM statistics and may cause performance degradation. We then propose the Log-Partition Approximation Cumulative Sum (LPA-CUSUM) algorithm based on thermodynamic integration (TI) in order to estimate the log-ratio of normalizing constants of pre- and post-change distributions. It is proved that this approach gives an unbiased estimate of the log-partition function and the CUSUM statistics, and leads to an asymptotically optimal performance. Moreover, we derive a relationship between the required sample size for thermodynamic integration and the desired detection delay performance, offering guidelines for practical parameter selection. Numerical studies are provided demonstrating the efficacy of our approach.
☆ Streaming Deep Reinforcement Learning Finally Works
Natural intelligence processes experience as a continuous stream, sensing, acting, and learning moment-by-moment in real time. Streaming learning, the modus operandi of classic reinforcement learning (RL) algorithms like Q-learning and TD, mimics natural learning by using the most recent sample without storing it. This approach is also ideal for resource-constrained, communication-limited, and privacy-sensitive applications. However, in deep RL, learners almost always use batch updates and replay buffers, making them computationally expensive and incompatible with streaming learning. Although the prevalence of batch deep RL is often attributed to its sample efficiency, a more critical reason for the absence of streaming deep RL is its frequent instability and failure to learn, which we refer to as stream barrier. This paper introduces the stream-x algorithms, the first class of deep RL algorithms to overcome stream barrier for both prediction and control and match sample efficiency of batch RL. Through experiments in Mujoco Gym, DM Control Suite, and Atari Games, we demonstrate stream barrier in existing algorithms and successful stable learning with our stream-x algorithms: stream Q, stream AC, and stream TD, achieving the best model-free performance in DM Control Dog environments. A set of common techniques underlies the stream-x algorithms, enabling their success with a single set of hyperparameters and allowing for easy extension to other algorithms, thereby reviving streaming RL.
☆ Learning to Control the Smoothness of Graph Convolutional Network Features
The pioneering work of Oono and Suzuki [ICLR, 2020] and Cai and Wang [arXiv:2006.13318] initializes the analysis of the smoothness of graph convolutional network (GCN) features. Their results reveal an intricate empirical correlation between node classification accuracy and the ratio of smooth to non-smooth feature components. However, the optimal ratio that favors node classification is unknown, and the non-smooth features of deep GCN with ReLU or leaky ReLU activation function diminish. In this paper, we propose a new strategy to let GCN learn node features with a desired smoothness -- adapting to data and tasks -- to enhance node classification. Our approach has three key steps: (1) We establish a geometric relationship between the input and output of ReLU or leaky ReLU. (2) Building on our geometric insights, we augment the message-passing process of graph convolutional layers (GCLs) with a learnable term to modulate the smoothness of node features with computational efficiency. (3) We investigate the achievable ratio between smooth and non-smooth feature components for GCNs with the augmented message-passing scheme. Our extensive numerical results show that the augmented message-passing schemes significantly improve node classification for GCN and some related models.
comment: 48 pages
☆ How Does Data Diversity Shape the Weight Landscape of Neural Networks?
To enhance the generalization of machine learning models to unseen data, techniques such as dropout, weight decay ($L_2$ regularization), and noise augmentation are commonly employed. While regularization methods (i.e., dropout and weight decay) are geared toward adjusting model parameters to prevent overfitting, data augmentation increases the diversity of the input training set, a method purported to improve accuracy and calibration error. In this paper, we investigate the impact of each of these techniques on the parameter space of neural networks, with the goal of understanding how they alter the weight landscape in transfer learning scenarios. To accomplish this, we employ Random Matrix Theory to analyze the eigenvalue distributions of pre-trained models, fine-tuned using these techniques but using different levels of data diversity, for the same downstream tasks. We observe that diverse data influences the weight landscape in a similar fashion as dropout. Additionally, we compare commonly used data augmentation methods with synthetic data created by generative models. We conclude that synthetic data can bring more diversity into real input data, resulting in a better performance on out-of-distribution test instances.
☆ Contractivity and linear convergence in bilinear saddle-point problems: An operator-theoretic approach
We study the convex-concave bilinear saddle-point problem $\min_x \max_y f(x) + y^\top Ax - g(y)$, where both, only one, or none of the functions $f$ and $g$ are strongly convex, and suitable rank conditions on the matrix $A$ hold. The solution of this problem is at the core of many machine learning tasks. By employing tools from operator theory, we systematically prove the contractivity (in turn, the linear convergence) of several first-order primal-dual algorithms, including the Chambolle-Pock method. Our approach results in concise and elegant proofs, and it yields new convergence guarantees and tighter bounds compared to known results.
☆ A Lipschitz spaces view of infinitely wide shallow neural networks
We revisit the mean field parametrization of shallow neural networks, using signed measures on unbounded parameter spaces and duality pairings that take into account the regularity and growth of activation functions. This setting directly leads to the use of unbalanced Kantorovich-Rubinstein norms defined by duality with Lipschitz functions, and of spaces of measures dual to those of continuous functions with controlled growth. These allow to make transparent the need for total variation and moment bounds or penalization to obtain existence of minimizers of variational formulations, under which we prove a compactness result in strong Kantorovich-Rubinstein norm, and in the absence of which we show several examples demonstrating undesirable behavior. Further, the Kantorovich-Rubinstein setting enables us to combine the advantages of a completely linear parametrization and ensuing reproducing kernel Banach space framework with optimal transport insights. We showcase this synergy with representer theorems and uniform large data limits for empirical risk minimization, and in proposed formulations for distillation and fusion applications.
comment: 39 pages, 1 table
☆ Learning With Multi-Group Guarantees For Clusterable Subpopulations
A canonical desideratum for prediction problems is that performance guarantees should hold not just on average over the population, but also for meaningful subpopulations within the overall population. But what constitutes a meaningful subpopulation? In this work, we take the perspective that relevant subpopulations should be defined with respect to the clusters that naturally emerge from the distribution of individuals for which predictions are being made. In this view, a population refers to a mixture model whose components constitute the relevant subpopulations. We suggest two formalisms for capturing per-subgroup guarantees: first, by attributing each individual to the component from which they were most likely drawn, given their features; and second, by attributing each individual to all components in proportion to their relative likelihood of having been drawn from each component. Using online calibration as a case study, we study a \variational algorithm that provides guarantees for each of these formalisms by handling all plausible underlying subpopulation structures simultaneously, and achieve an $O(T^{1/2})$ rate even when the subpopulations are not well-separated. In comparison, the more natural cluster-then-predict approach that first recovers the structure of the subpopulations and then makes predictions suffers from a $O(T^{2/3})$ rate and requires the subpopulations to be separable. Along the way, we prove that providing per-subgroup calibration guarantees for underlying clusters can be easier than learning the clusters: separation between median subgroup features is required for the latter but not the former.
☆ Neuro-Symbolic Traders: Assessing the Wisdom of AI Crowds in Markets
Deep generative models are becoming increasingly used as tools for financial analysis. However, it is unclear how these models will influence financial markets, especially when they infer financial value in a semi-autonomous way. In this work, we explore the interplay between deep generative models and market dynamics. We develop a form of virtual traders that use deep generative models to make buy/sell decisions, which we term neuro-symbolic traders, and expose them to a virtual market. Under our framework, neuro-symbolic traders are agents that use vision-language models to discover a model of the fundamental value of an asset. Agents develop this model as a stochastic differential equation, calibrated to market data using gradient descent. We test our neuro-symbolic traders on both synthetic data and real financial time series, including an equity stock, commodity, and a foreign exchange pair. We then expose several groups of neuro-symbolic traders to a virtual market environment. This market environment allows for feedback between the traders belief of the underlying value to the observed price dynamics. We find that this leads to price suppression compared to the historical data, highlighting a future risk to market stability. Our work is a first step towards quantifying the effect of deep generative agents on markets dynamics and sets out some of the potential risks and benefits of this approach in the future.
comment: 8 pages, 4 figures, ACM format
☆ Neural Combinatorial Clustered Bandits for Recommendation Systems
We consider the contextual combinatorial bandit setting where in each round, the learning agent, e.g., a recommender system, selects a subset of "arms," e.g., products, and observes rewards for both the individual base arms, which are a function of known features (called "context"), and the super arm (the subset of arms), which is a function of the base arm rewards. The agent's goal is to simultaneously learn the unknown reward functions and choose the highest-reward arms. For example, the "reward" may represent a user's probability of clicking on one of the recommended products. Conventional bandit models, however, employ restrictive reward function models in order to obtain performance guarantees. We make use of deep neural networks to estimate and learn the unknown reward functions and propose Neural UCB Clustering (NeUClust), which adopts a clustering approach to select the super arm in every round by exploiting underlying structure in the context space. Unlike prior neural bandit works, NeUClust uses a neural network to estimate the super arm reward and select the super arm, thus eliminating the need for a known optimization oracle. We non-trivially extend prior neural combinatorial bandit works to prove that NeUClust achieves $\widetilde{O}\left(\widetilde{d}\sqrt{T}\right)$ regret, where $\widetilde{d}$ is the effective dimension of a neural tangent kernel matrix, $T$ the number of rounds. Experiments on real world recommendation datasets show that NeUClust achieves better regret and reward than other contextual combinatorial and neural bandit algorithms.
☆ Optimizing Attention with Mirror Descent: Generalized Max-Margin Token Selection
Attention mechanisms have revolutionized several domains of artificial intelligence, such as natural language processing and computer vision, by enabling models to selectively focus on relevant parts of the input data. While recent work has characterized the optimization dynamics of gradient descent (GD) in attention-based models and the structural properties of its preferred solutions, less is known about more general optimization algorithms such as mirror descent (MD). In this paper, we investigate the convergence properties and implicit biases of a family of MD algorithms tailored for softmax attention mechanisms, with the potential function chosen as the $p$-th power of the $\ell_p$-norm. Specifically, we show that these algorithms converge in direction to a generalized hard-margin SVM with an $\ell_p$-norm objective when applied to a classification problem using a softmax attention model. Notably, our theoretical results reveal that the convergence rate is comparable to that of traditional GD in simpler models, despite the highly nonlinear and nonconvex nature of the present problem. Additionally, we delve into the joint optimization dynamics of the key-query matrix and the decoder, establishing conditions under which this complex joint optimization converges to their respective hard-margin SVM solutions. Lastly, our numerical experiments on real data demonstrate that MD algorithms improve generalization over standard GD and excel in optimal token selection.
☆ Towards Unsupervised Validation of Anomaly-Detection Models
Unsupervised validation of anomaly-detection models is a highly challenging task. While the common practices for model validation involve a labeled validation set, such validation sets cannot be constructed when the underlying datasets are unlabeled. The lack of robust and efficient unsupervised model-validation techniques presents an acute challenge in the implementation of automated anomaly-detection pipelines, especially when there exists no prior knowledge of the model's performance on similar datasets. This work presents a new paradigm to automated validation of anomaly-detection models, inspired by real-world, collaborative decision-making mechanisms. We focus on two commonly-used, unsupervised model-validation tasks -- model selection and model evaluation -- and provide extensive experimental results that demonstrate the accuracy and robustness of our approach on both tasks.
☆ Large Language Models Are Overparameterized Text Encoders
Large language models (LLMs) demonstrate strong performance as text embedding models when finetuned with supervised contrastive training. However, their large size balloons inference time and memory requirements. In this paper, we show that by pruning the last $p\%$ layers of an LLM before supervised training for only 1000 steps, we can achieve a proportional reduction in memory and inference time. We evaluate four different state-of-the-art LLMs on text embedding tasks and find that our method can prune up to 30\% of layers with negligible impact on performance and up to 80\% with only a modest drop. With only three lines of code, our method is easily implemented in any pipeline for transforming LLMs to text encoders. We also propose $\text{L}^3 \text{Prune}$, a novel layer-pruning strategy based on the model's initial loss that provides two optimal pruning configurations: a large variant with negligible performance loss and a small variant for resource-constrained settings. On average, the large variant prunes 21\% of the parameters with a $-0.3$ performance drop, and the small variant only suffers from a $-5.1$ decrease while pruning 74\% of the model. We consider these results strong evidence that LLMs are overparameterized for text embedding tasks, and can be easily pruned.
comment: 8 pages of content + 1 for limitations and ethical considerations, 14 pages in total including references and appendix, 5+1 figures
☆ MomentumSMoE: Integrating Momentum into Sparse Mixture of Experts NeurIPS 2024
Sparse Mixture of Experts (SMoE) has become the key to unlocking unparalleled scalability in deep learning. SMoE has the potential to exponentially increase parameter count while maintaining the efficiency of the model by only activating a small subset of these parameters for a given sample. However, it has been observed that SMoE suffers from unstable training and has difficulty adapting to new distributions, leading to the model's lack of robustness to data contamination. To overcome these limitations, we first establish a connection between the dynamics of the expert representations in SMoEs and gradient descent on a multi-objective optimization problem. Leveraging our framework, we then integrate momentum into SMoE and propose a new family of SMoEs named MomentumSMoE. We theoretically prove and numerically demonstrate that MomentumSMoE is more stable and robust than SMoE. In particular, we verify the advantages of MomentumSMoE over SMoE on a variety of practical tasks including ImageNet-1K object recognition and WikiText-103 language modeling. We demonstrate the applicability of MomentumSMoE to many types of SMoE models, including those in the Sparse MoE model for vision (V-MoE) and the Generalist Language Model (GLaM). We also show that other advanced momentum-based optimization methods, such as Adam, can be easily incorporated into the MomentumSMoE framework for designing new SMoE models with even better performance, almost negligible additional computation cost, and simple implementations.
comment: 10 pages in the main text. Published at NeurIPS 2024. The code is available at https://github.com/rachtsy/MomentumSMoE
☆ Building Trust in Black-box Optimization: A Comprehensive Framework for Explainability
Optimizing costly black-box functions within a constrained evaluation budget presents significant challenges in many real-world applications. Surrogate Optimization (SO) is a common resolution, yet its proprietary nature introduced by the complexity of surrogate models and the sampling core (e.g., acquisition functions) often leads to a lack of explainability and transparency. While existing literature has primarily concentrated on enhancing convergence to global optima, the practical interpretation of newly proposed strategies remains underexplored, especially in batch evaluation settings. In this paper, we propose \emph{Inclusive} Explainability Metrics for Surrogate Optimization (IEMSO), a comprehensive set of model-agnostic metrics designed to enhance the transparency, trustworthiness, and explainability of the SO approaches. Through these metrics, we provide both intermediate and post-hoc explanations to practitioners before and after performing expensive evaluations to gain trust. We consider four primary categories of metrics, each targeting a specific aspect of the SO process: Sampling Core Metrics, Batch Properties Metrics, Optimization Process Metrics, and Feature Importance. Our experimental evaluations demonstrate the significant potential of the proposed metrics across different benchmarks.
☆ Understanding the difficulty of low-precision post-training quantization of large language models
Large language models of high parameter counts are computationally expensive, yet can be made much more efficient by compressing their weights to very low numerical precision. This can be achieved either through post-training quantization by minimizing local, layer-wise quantization errors, or through quantization-aware fine-tuning by minimizing the global loss function. In this study, we discovered that, under the same data constraint, the former approach nearly always fared worse than the latter, a phenomenon particularly prominent when the numerical precision is very low. We further showed that this difficulty of post-training quantization arose from stark misalignment between optimization of the local and global objective functions. Our findings explains limited utility in minimization of local quantization error and the importance of direct quantization-aware fine-tuning, in the regime of large models at very low precision.
☆ Measuring Diversity: Axioms and Challenges
The concept of diversity is widely used in various applications: from image or molecule generation to recommender systems. Thus, being able to properly measure diversity is important. This paper addresses the problem of quantifying diversity for a set of objects. First, we make a systematic review of existing diversity measures and explore their undesirable behavior in some cases. Based on this review, we formulate three desirable properties (axioms) of a reliable diversity measure: monotonicity, uniqueness, and continuity. We show that none of the existing measures has all three properties and thus these measures are not suitable for quantifying diversity. Then, we construct two examples of measures that have all the desirable properties, thus proving that the list of axioms is not self-contradicting. Unfortunately, the constructed examples are too computationally complex for practical use, thus we pose an open problem of constructing a diversity measure that has all the listed properties and can be computed in practice.
comment: 17 pages, 7 figures
☆ Boosting K-means for Big Data by Fusing Data Streaming with Global Optimization
K-means clustering is a cornerstone of data mining, but its efficiency deteriorates when confronted with massive datasets. To address this limitation, we propose a novel heuristic algorithm that leverages the Variable Neighborhood Search (VNS) metaheuristic to optimize K-means clustering for big data. Our approach is based on the sequential optimization of the partial objective function landscapes obtained by restricting the Minimum Sum-of-Squares Clustering (MSSC) formulation to random samples from the original big dataset. Within each landscape, systematically expanding neighborhoods of the currently best (incumbent) solution are explored by reinitializing all degenerate and a varying number of additional centroids. Extensive and rigorous experimentation on a large number of real-world datasets reveals that by transforming the traditional local search into a global one, our algorithm significantly enhances the accuracy and efficiency of K-means clustering in big data environments, becoming the new state of the art in the field.
☆ Diffusion-based Semi-supervised Spectral Algorithm for Regression on Manifolds
We introduce a novel diffusion-based spectral algorithm to tackle regression analysis on high-dimensional data, particularly data embedded within lower-dimensional manifolds. Traditional spectral algorithms often fall short in such contexts, primarily due to the reliance on predetermined kernel functions, which inadequately address the complex structures inherent in manifold-based data. By employing graph Laplacian approximation, our method uses the local estimation property of heat kernel, offering an adaptive, data-driven approach to overcome this obstacle. Another distinct advantage of our algorithm lies in its semi-supervised learning framework, enabling it to fully use the additional unlabeled data. This ability enhances the performance by allowing the algorithm to dig the spectrum and curvature of the data manifold, providing a more comprehensive understanding of the dataset. Moreover, our algorithm performs in an entirely data-driven manner, operating directly within the intrinsic manifold structure of the data, without requiring any predefined manifold information. We provide a convergence analysis of our algorithm. Our findings reveal that the algorithm achieves a convergence rate that depends solely on the intrinsic dimension of the underlying manifold, thereby avoiding the curse of dimensionality associated with the higher ambient dimension.
☆ Comparing Differentiable and Dynamic Ray Tracing: Introducing the Multipath Lifetime Map
With the increasing presence of dynamic scenarios, such as Vehicle-to-Vehicle communications, radio propagation modeling tools must adapt to the rapidly changing nature of the radio channel. Recently, both Differentiable and Dynamic Ray Tracing frameworks have emerged to address these challenges. However, there is often confusion about how these approaches differ and which one should be used in specific contexts. In this paper, we provide an overview of these two techniques and a comparative analysis against two state-of-the-art tools: 3DSCAT from UniBo and Sionna from NVIDIA. To provide a more precise characterization of the scope of these methods, we introduce a novel simulation-based metric, the Multipath Lifetime Map, which enables the evaluation of spatial and temporal coherence in radio channels only based on the geometrical description of the environment. Finally, our metrics are evaluated on a classic urban street canyon scenario, yielding similar results to those obtained from measurement campaigns.
comment: 5 pages, 5 figures, 1 table, submitted to EuCAP 2025
☆ The Traveling Bandit: A Framework for Bayesian Optimization with Movement Costs
This paper introduces a framework for Bayesian Optimization (BO) with metric movement costs, addressing a critical challenge in practical applications where input alterations incur varying costs. Our approach is a convenient plug-in that seamlessly integrates with the existing literature on batched algorithms, where designs within batches are observed following the solution of a Traveling Salesman Problem. The proposed method provides a theoretical guarantee of convergence in terms of movement costs for BO. Empirically, our method effectively reduces average movement costs over time while maintaining comparable regret performance to conventional BO methods. This framework also shows promise for broader applications in various bandit settings with movement costs.
☆ Using Sentiment and Technical Analysis to Predict Bitcoin with Machine Learning
Cryptocurrencies have gained significant attention in recent years due to their decentralized nature and potential for financial innovation. Thus, the ability to accurately predict its price has become a subject of great interest for investors, traders, and researchers. Some works in the literature show how Bitcoin's market sentiment correlates with its price fluctuations in the market. However, papers that consider the sentiment of the market associated with financial Technical Analysis indicators in order to predict Bitcoin's price are still scarce. In this paper, we present a novel approach for predicting Bitcoin price movements by combining the Fear & Greedy Index, a measure of market sentiment, Technical Analysis indicators, and the potential of Machine Learning algorithms. This work represents a preliminary study on the importance of sentiment metrics in cryptocurrency forecasting. Our initial experiments demonstrate promising results considering investment returns, surpassing the Buy & Hold baseline, and offering valuable insights about the combination of indicators of sentiment and market in a cryptocurrency prediction model.
☆ Domain Adaptive Safety Filters via Deep Operator Learning
Learning-based approaches for constructing Control Barrier Functions (CBFs) are increasingly being explored for safety-critical control systems. However, these methods typically require complete retraining when applied to unseen environments, limiting their adaptability. To address this, we propose a self-supervised deep operator learning framework that learns the mapping from environmental parameters to the corresponding CBF, rather than learning the CBF directly. Our approach leverages the residual of a parametric Partial Differential Equation (PDE), where the solution defines a parametric CBF approximating the maximal control invariant set. This framework accommodates complex safety constraints, higher relative degrees, and actuation limits. We demonstrate the effectiveness of the method through numerical experiments on navigation tasks involving dynamic obstacles.
comment: 63rd IEEE Conference on Decision and Control (CDC)
☆ Rethinking Distance Metrics for Counterfactual Explainability
Counterfactual explanations have been a popular method of post-hoc explainability for a variety of settings in Machine Learning. Such methods focus on explaining classifiers by generating new data points that are similar to a given reference, while receiving a more desirable prediction. In this work, we investigate a framing for counterfactual generation methods that considers counterfactuals not as independent draws from a region around the reference, but as jointly sampled with the reference from the underlying data distribution. Through this framing, we derive a distance metric, tailored for counterfactual similarity that can be applied to a broad range of settings. Through both quantitative and qualitative analyses of counterfactual generation methods, we show that this framing allows us to express more nuanced dependencies among the covariates.
comment: 13 pages, 3 figures, 1 table
☆ Efficient Annotator Reliability Assessment and Sample Weighting for Knowledge-Based Misinformation Detection on Social Media
Misinformation spreads rapidly on social media, confusing the truth and targetting potentially vulnerable people. To effectively mitigate the negative impact of misinformation, it must first be accurately detected before applying a mitigation strategy, such as X's community notes, which is currently a manual process. This study takes a knowledge-based approach to misinformation detection, modelling the problem similarly to one of natural language inference. The EffiARA annotation framework is introduced, aiming to utilise inter- and intra-annotator agreement to understand the reliability of each annotator and influence the training of large language models for classification based on annotator reliability. In assessing the EffiARA annotation framework, the Russo-Ukrainian Conflict Knowledge-Based Misinformation Classification Dataset (RUC-MCD) was developed and made publicly available. This study finds that sample weighting using annotator reliability performs the best, utilising both inter- and intra-annotator agreement and soft-label training. The highest classification performance achieved using Llama-3.2-1B was a macro-F1 of 0.757 and 0.740 using TwHIN-BERT-large.
comment: 8 pages, 3 figures, 3 tables. Code available here: https://github.com/MiniEggz/ruc-misinfo
☆ An Integrated Deep Learning Model for Skin Cancer Detection Using Hybrid Feature Fusion Technique
Skin cancer is a serious and potentially fatal disease caused by DNA damage. Early detection significantly increases survival rates, making accurate diagnosis crucial. In this groundbreaking study, we present a hybrid framework based on Deep Learning (DL) that achieves precise classification of benign and malignant skin lesions. Our approach begins with dataset preprocessing to enhance classification accuracy, followed by training two separate pre-trained DL models, InceptionV3 and DenseNet121. By fusing the results of each model using the weighted sum rule, our system achieves exceptional accuracy rates. Specifically, we achieve a 92.27% detection accuracy rate, 92.33% sensitivity, 92.22% specificity, 90.81% precision, and 91.57% F1-score, outperforming existing models and demonstrating the robustness and trustworthiness of our hybrid approach. Our study represents a significant advance in skin cancer diagnosis and provides a promising foundation for further research in the field. With the potential to save countless lives through earlier detection, our hybrid deep-learning approach is a game-changer in the fight against skin cancer.
☆ ANT: Adaptive Noise Schedule for Time Series Diffusion Models NeurIPS 2024
Advances in diffusion models for generative artificial intelligence have recently propagated to the time series (TS) domain, demonstrating state-of-the-art performance on various tasks. However, prior works on TS diffusion models often borrow the framework of existing works proposed in other domains without considering the characteristics of TS data, leading to suboptimal performance. In this work, we propose Adaptive Noise schedule for Time series diffusion models (ANT), which automatically predetermines proper noise schedules for given TS datasets based on their statistics representing non-stationarity. Our intuition is that an optimal noise schedule should satisfy the following desiderata: 1) It linearly reduces the non-stationarity of TS data so that all diffusion steps are equally meaningful, 2) the data is corrupted to the random noise at the final step, and 3) the number of steps is sufficiently large. The proposed method is practical for use in that it eliminates the necessity of finding the optimal noise schedule with a small additional cost to compute the statistics for given datasets, which can be done offline before training. We validate the effectiveness of our method across various tasks, including TS forecasting, refinement, and generation, on datasets from diverse domains. Code is available at this repository: https://github.com/seunghan96/ANT.
comment: NeurIPS 2024
☆ CaTs and DAGs: Integrating Directed Acyclic Graphs with Transformers and Fully-Connected Neural Networks for Causally Constrained Predictions
Artificial Neural Networks (ANNs), including fully-connected networks and transformers, are highly flexible and powerful function approximators, widely applied in fields like computer vision and natural language processing. However, their inability to inherently respect causal structures can limit their robustness, making them vulnerable to covariate shift and difficult to interpret/explain. This poses significant challenges for their reliability in real-world applications. In this paper, we introduce Causal Fully-Connected Neural Networks (CFCNs) and Causal Transformers (CaTs), two general model families designed to operate under predefined causal constraints, as specified by a Directed Acyclic Graph (DAG). These models retain the powerful function approximation abilities of traditional neural networks while adhering to the underlying structural constraints, improving robustness, reliability, and interpretability at inference time. This approach opens new avenues for deploying neural networks in more demanding, real-world scenarios where robustness and explainability is critical.
☆ Transfer Reinforcement Learning in Heterogeneous Action Spaces using Subgoal Mapping
In this paper, we consider a transfer reinforcement learning problem involving agents with different action spaces. Specifically, for any new unseen task, the goal is to use a successful demonstration of this task by an expert agent in its action space to enable a learner agent learn an optimal policy in its own different action space with fewer samples than those required if the learner was learning on its own. Existing transfer learning methods across different action spaces either require handcrafted mappings between those action spaces provided by human experts, which can induce bias in the learning procedure, or require the expert agent to share its policy parameters with the learner agent, which does not generalize well to unseen tasks. In this work, we propose a method that learns a subgoal mapping between the expert agent policy and the learner agent policy. Since the expert agent and the learner agent have different action spaces, their optimal policies can have different subgoal trajectories. We learn this subgoal mapping by training a Long Short Term Memory (LSTM) network for a distribution of tasks and then use this mapping to predict the learner subgoal sequence for unseen tasks, thereby improving the speed of learning by biasing the agent's policy towards the predicted learner subgoal sequence. Through numerical experiments, we demonstrate that the proposed learning scheme can effectively find the subgoal mapping underlying the given distribution of tasks. Moreover, letting the learner agent imitate the expert agent's policy with the learnt subgoal mapping can significantly improve the sample efficiency and training time of the learner agent in unseen new tasks.
☆ Spectral Representations for Accurate Causal Uncertainty Quantification with Gaussian Processes
Accurate uncertainty quantification for causal effects is essential for robust decision making in complex systems, but remains challenging in non-parametric settings. One promising framework represents conditional distributions in a reproducing kernel Hilbert space and places Gaussian process priors on them to infer posteriors on causal effects, but requires restrictive nuclear dominant kernels and approximations that lead to unreliable uncertainty estimates. In this work, we introduce a method, IMPspec, that addresses these limitations via a spectral representation of the Hilbert space. We show that posteriors in this model can be obtained explicitly, by extending a result in Hilbert space regression theory. We also learn the spectral representation to optimise posterior calibration. Our method achieves state-of-the-art performance in uncertainty quantification and causal Bayesian optimisation across simulations and a healthcare application.
☆ Backdoored Retrievers for Prompt Injection Attacks on Retrieval Augmented Generation of Large Language Models
Large Language Models (LLMs) have demonstrated remarkable capabilities in generating coherent text but remain limited by the static nature of their training data. Retrieval Augmented Generation (RAG) addresses this issue by combining LLMs with up-to-date information retrieval, but also expand the attack surface of the system. This paper investigates prompt injection attacks on RAG, focusing on malicious objectives beyond misinformation, such as inserting harmful links, promoting unauthorized services, and initiating denial-of-service behaviors. We build upon existing corpus poisoning techniques and propose a novel backdoor attack aimed at the fine-tuning process of the dense retriever component. Our experiments reveal that corpus poisoning can achieve significant attack success rates through the injection of a small number of compromised documents into the retriever corpus. In contrast, backdoor attacks demonstrate even higher success rates but necessitate a more complex setup, as the victim must fine-tune the retriever using the attacker poisoned dataset.
comment: 12 pages, 5 figures
☆ Laplace Transform Based Low-Complexity Learning of Continuous Markov Semigroups
Markov processes serve as a universal model for many real-world random processes. This paper presents a data-driven approach for learning these models through the spectral decomposition of the infinitesimal generator (IG) of the Markov semigroup. The unbounded nature of IGs complicates traditional methods such as vector-valued regression and Hilbert-Schmidt operator analysis. Existing techniques, including physics-informed kernel regression, are computationally expensive and limited in scope, with no recovery guarantees for transfer operator methods when the time-lag is small. We propose a novel method that leverages the IG's resolvent, characterized by the Laplace transform of transfer operators. This approach is robust to time-lag variations, ensuring accurate eigenvalue learning even for small time-lags. Our statistical analysis applies to a broader class of Markov processes than current methods while reducing computational complexity from quadratic to linear in the state dimension. Finally, we illustrate the behaviour of our method in two experiments.
comment: 35 pages
☆ Enhancing Cryptocurrency Market Forecasting: Advanced Machine Learning Techniques and Industrial Engineering Contributions
Cryptocurrencies, as decentralized digital assets, have experienced rapid growth and adoption, with over 23,000 cryptocurrencies and a market capitalization nearing \$1.1 trillion (about \$3,400 per person in the US) as of 2023. This dynamic market presents significant opportunities and risks, highlighting the need for accurate price prediction models to manage volatility. This chapter comprehensively reviews machine learning (ML) techniques applied to cryptocurrency price prediction from 2014 to 2024. We explore various ML algorithms, including linear models, tree-based approaches, and advanced deep learning architectures such as transformers and large language models. Additionally, we examine the role of sentiment analysis in capturing market sentiment from textual data like social media posts and news articles to anticipate price fluctuations. With expertise in optimizing complex systems and processes, industrial engineers are pivotal in enhancing these models. They contribute by applying principles of process optimization, efficiency, and risk mitigation to improve computational performance and data management. This chapter highlights the evolving landscape of cryptocurrency price prediction, the integration of emerging technologies, and the significant role of industrial engineers in refining predictive models. By addressing current limitations and exploring future research directions, this chapter aims to advance the development of more accurate and robust prediction systems, supporting better-informed investment decisions and more stable market behavior.
comment: 63 pages, 6 figures
☆ How Do Training Methods Influence the Utilization of Vision Models? NeurIPS 2024
Not all learnable parameters (e.g., weights) contribute equally to a neural network's decision function. In fact, entire layers' parameters can sometimes be reset to random values with little to no impact on the model's decisions. We revisit earlier studies that examined how architecture and task complexity influence this phenomenon and ask: is this phenomenon also affected by how we train the model? We conducted experimental evaluations on a diverse set of ImageNet-1k classification models to explore this, keeping the architecture and training data constant but varying the training pipeline. Our findings reveal that the training method strongly influences which layers become critical to the decision function for a given task. For example, improved training regimes and self-supervised training increase the importance of early layers while significantly under-utilizing deeper layers. In contrast, methods such as adversarial training display an opposite trend. Our preliminary results extend previous findings, offering a more nuanced understanding of the inner mechanics of neural networks. Code: https://github.com/paulgavrikov/layer_criticality
comment: Accepted at the Interpretable AI: Past, Present and Future Workshop at NeurIPS 2024
☆ Flow-based Sampling for Entanglement Entropy and the Machine Learning of Defects
We introduce a novel technique to numerically calculate R\'enyi entanglement entropies in lattice quantum field theory using generative models. We describe how flow-based approaches can be combined with the replica trick using a custom neural-network architecture around a lattice defect connecting two replicas. Numerical tests for the $\phi^4$ scalar field theory in two and three dimensions demonstrate that our technique outperforms state-of-the-art Monte Carlo calculations, and exhibit a promising scaling with the defect size.
comment: 10 pages, 9 figures
☆ Electrocardiogram-Language Model for Few-Shot Question Answering with Meta Learning
Electrocardiogram (ECG) interpretation requires specialized expertise, often involving synthesizing insights from ECG signals with complex clinical queries posed in natural language. The scarcity of labeled ECG data coupled with the diverse nature of clinical inquiries presents a significant challenge for developing robust and adaptable ECG diagnostic systems. This work introduces a novel multimodal meta-learning method for few-shot ECG question answering, addressing the challenge of limited labeled data while leveraging the rich knowledge encoded within large language models (LLMs). Our LLM-agnostic approach integrates a pre-trained ECG encoder with a frozen LLM (e.g., LLaMA and Gemma) via a trainable fusion module, enabling the language model to reason about ECG data and generate clinically meaningful answers. Extensive experiments demonstrate superior generalization to unseen diagnostic tasks compared to supervised baselines, achieving notable performance even with limited ECG leads. For instance, in a 5-way 5-shot setting, our method using LLaMA-3.1-8B achieves accuracy of 84.6%, 77.3%, and 69.6% on single verify, choose and query question types, respectively. These results highlight the potential of our method to enhance clinical ECG interpretation by combining signal processing with the nuanced language understanding capabilities of LLMs, particularly in data-constrained scenarios.
☆ The Propensity for Density in Feed-forward Models
Does the process of training a neural network to solve a task tend to use all of the available weights even when the task could be solved with fewer weights? To address this question we study the effects of pruning fully connected, convolutional and residual models while varying their widths. We find that the proportion of weights that can be pruned without degrading performance is largely invariant to model size. Increasing the width of a model has little effect on the density of the pruned model relative to the increase in absolute size of the pruned network. In particular, we find substantial prunability across a large range of model sizes, where our biggest model is 50 times as wide as our smallest model. We explore three hypotheses that could explain these findings.
☆ Learning to refine domain knowledge for biological network inference
Perturbation experiments allow biologists to discover causal relationships between variables of interest, but the sparsity and high dimensionality of these data pose significant challenges for causal structure learning algorithms. Biological knowledge graphs can bootstrap the inference of causal structures in these situations, but since they compile vastly diverse information, they can bias predictions towards well-studied systems. Alternatively, amortized causal structure learning algorithms encode inductive biases through data simulation and train supervised models to recapitulate these synthetic graphs. However, realistically simulating biology is arguably even harder than understanding a specific system. In this work, we take inspiration from both strategies and propose an amortized algorithm for refining domain knowledge, based on data observations. On real and synthetic datasets, we show that our approach outperforms baselines in recovering ground truth causal graphs and identifying errors in the prior knowledge with limited interventional data.
☆ A Bioinformatic Approach Validated Utilizing Machine Learning Algorithms to Identify Relevant Biomarkers and Crucial Pathways in Gallbladder Cancer
Gallbladder cancer (GBC) is the most frequent cause of disease among biliary tract neoplasms. Identifying the molecular mechanisms and biomarkers linked to GBC progression has been a significant challenge in scientific research. Few recent studies have explored the roles of biomarkers in GBC. Our study aimed to identify biomarkers in GBC using machine learning (ML) and bioinformatics techniques. We compared GBC tumor samples with normal samples to identify differentially expressed genes (DEGs) from two microarray datasets (GSE100363, GSE139682) obtained from the NCBI GEO database. A total of 146 DEGs were found, with 39 up-regulated and 107 down-regulated genes. Functional enrichment analysis of these DEGs was performed using Gene Ontology (GO) terms and REACTOME pathways through DAVID. The protein-protein interaction network was constructed using the STRING database. To identify hub genes, we applied three ranking algorithms: Degree, MNC, and Closeness Centrality. The intersection of hub genes from these algorithms yielded 11 hub genes. Simultaneously, two feature selection methods (Pearson correlation and recursive feature elimination) were used to identify significant gene subsets. We then developed ML models using SVM and RF on the GSE100363 dataset, with validation on GSE139682, to determine the gene subset that best distinguishes GBC samples. The hub genes outperformed the other gene subsets. Finally, NTRK2, COL14A1, SCN4B, ATP1A2, SLC17A7, SLIT3, COL7A1, CLDN4, CLEC3B, ADCYAP1R1, and MFAP4 were identified as crucial genes, with SLIT3, COL7A1, and CLDN4 being strongly linked to GBC development and prediction.
☆ FashionR2R: Texture-preserving Rendered-to-Real Image Translation with Diffusion Models NeurIPS 2024
Modeling and producing lifelike clothed human images has attracted researchers' attention from different areas for decades, with the complexity from highly articulated and structured content. Rendering algorithms decompose and simulate the imaging process of a camera, while are limited by the accuracy of modeled variables and the efficiency of computation. Generative models can produce impressively vivid human images, however still lacking in controllability and editability. This paper studies photorealism enhancement of rendered images, leveraging generative power from diffusion models on the controlled basis of rendering. We introduce a novel framework to translate rendered images into their realistic counterparts, which consists of two stages: Domain Knowledge Injection (DKI) and Realistic Image Generation (RIG). In DKI, we adopt positive (real) domain finetuning and negative (rendered) domain embedding to inject knowledge into a pretrained Text-to-image (T2I) diffusion model. In RIG, we generate the realistic image corresponding to the input rendered image, with a Texture-preserving Attention Control (TAC) to preserve fine-grained clothing textures, exploiting the decoupled features encoded in the UNet structure. Additionally, we introduce SynFashion dataset, featuring high-quality digital clothing images with diverse textures. Extensive experimental results demonstrate the superiority and effectiveness of our method in rendered-to-real image translation.
comment: Accepted by NeurIPS 2024
☆ Predicting time-varying flux and balance in metabolic systems using structured neural-ODE processes
We develop a novel data-driven framework as an alternative to dynamic flux balance analysis, bypassing the demand for deep domain knowledge and manual efforts to formulate the optimization problem. The proposed framework is end-to-end, which trains a structured neural ODE process (SNODEP) model to estimate flux and balance samples using gene-expression time-series data. SNODEP is designed to circumvent the limitations of the standard neural ODE process model, including restricting the latent and decoder sampling distributions to be normal and lacking structure between context points for calculating the latent, thus more suitable for modeling the underlying dynamics of a metabolic system. Through comprehensive experiments ($156$ in total), we demonstrate that SNODEP not only predicts the unseen time points of real-world gene-expression data and the flux and balance estimates well but can even generalize to more challenging unseen knockout configurations and irregular data sampling scenarios, all essential for metabolic pathway analysis. We hope our work can serve as a catalyst for building more scalable and powerful models for genome-scale metabolic analysis. Our code is available at: \url{https://github.com/TrustMLRG/SNODEP}.
☆ Integrating Deep Learning with Fundus and Optical Coherence Tomography for Cardiovascular Disease Prediction
Early identification of patients at risk of cardiovascular diseases (CVD) is crucial for effective preventive care, reducing healthcare burden, and improving patients' quality of life. This study demonstrates the potential of retinal optical coherence tomography (OCT) imaging combined with fundus photographs for identifying future adverse cardiac events. We used data from 977 patients who experienced CVD within a 5-year interval post-image acquisition, alongside 1,877 control participants without CVD, totaling 2,854 subjects. We propose a novel binary classification network based on a Multi-channel Variational Autoencoder (MCVAE), which learns a latent embedding of patients' fundus and OCT images to classify individuals into two groups: those likely to develop CVD in the future and those who are not. Our model, trained on both imaging modalities, achieved promising results (AUROC 0.78 +/- 0.02, accuracy 0.68 +/- 0.002, precision 0.74 +/- 0.02, sensitivity 0.73 +/- 0.02, and specificity 0.68 +/- 0.01), demonstrating its efficacy in identifying patients at risk of future CVD events based on their retinal images. This study highlights the potential of retinal OCT imaging and fundus photographs as cost-effective, non-invasive alternatives for predicting cardiovascular disease risk. The widespread availability of these imaging techniques in optometry practices and hospitals further enhances their potential for large-scale CVD risk screening. Our findings contribute to the development of standardized, accessible methods for early CVD risk identification, potentially improving preventive care strategies and patient outcomes.
comment: Part of the book series: Lecture Notes in Computer Science ((LNCS,volume 15155))
☆ Asymptotic non-linear shrinkage formulas for weighted sample covariance
We compute asymptotic non-linear shrinkage formulas for covariance and precision matrix estimators for weighted sample covariances, in the spirit of Ledoit and P\'ech\'e. We detail explicitly the formulas for exponentially-weighted sample covariances. Those new tools pave a way for applying non-linear shrinkage methods on weighted sample covariance. We show experimentally the performance of the asymptotic shrinkage formulas. Finally, we test the robustness of the theory to a heavy-tailed distributions.
☆ An explainable machine learning approach for energy forecasting at the household level
Electricity forecasting has been a recurring research topic, as it is key to finding the right balance between production and consumption. While most papers are focused on the national or regional scale, few are interested in the household level. Desegregated forecast is a common topic in Machine Learning (ML) literature but lacks explainability that household energy forecasts require. This paper specifically targets the challenges of forecasting electricity use at the household level. This paper confronts common Machine Learning algorithms to electricity household forecasts, weighing the pros and cons, including accuracy and explainability with well-known key metrics. Furthermore, we also confront them in this paper with the business challenges specific to this sector such as explainability or outliers resistance. We introduce a custom decision tree, aiming at providing a fair estimate of the energy consumption, while being explainable and consistent with human intuition. We show that this novel method allows greater explainability without sacrificing much accuracy. The custom tree methodology can be used in various business use cases but is subject to limitations, such as a lack of resilience with outliers.
☆ WeSpeR: Population spectrum retrieval and spectral density estimation of weighted sample covariance
The spectrum of the weighted sample covariance shows a asymptotic non random behavior when the dimension grows with the number of samples. In this setting, we prove that the asymptotic spectral distribution $F$ of the weighted sample covariance has a continuous density on $\mathbb{R}^*$. We address then the practical problem of numerically finding this density. We propose a procedure to compute it, to determine the support of $F$ and define an efficient grid on it. We use this procedure to design the $\textit{WeSpeR}$ algorithm, which estimates the spectral density and retrieves the true spectral covariance spectrum. Empirical tests confirm the good properties of the $\textit{WeSpeR}$ algorithm.
☆ SNAC: Multi-Scale Neural Audio Codec
Neural audio codecs have recently gained popularity because they can represent audio signals with high fidelity at very low bitrates, making it feasible to use language modeling approaches for audio generation and understanding. Residual Vector Quantization (RVQ) has become the standard technique for neural audio compression using a cascade of VQ codebooks. This paper proposes the Multi-Scale Neural Audio Codec, a simple extension of RVQ where the quantizers can operate at different temporal resolutions. By applying a hierarchy of quantizers at variable frame rates, the codec adapts to the audio structure across multiple timescales. This leads to more efficient compression, as demonstrated by extensive objective and subjective evaluations. The code and model weights are open-sourced at https://github.com/hubertsiuzdak/snac.
☆ Debug Smarter, Not Harder: AI Agents for Error Resolution in Computational Notebooks EMNLP 2024
Computational notebooks became indispensable tools for research-related development, offering unprecedented interactivity and flexibility in the development process. However, these benefits come at the cost of reproducibility and an increased potential for bugs. With the rise of code-fluent Large Language Models empowered with agentic techniques, smart bug-fixing tools with a high level of autonomy have emerged. However, those tools are tuned for classical script programming and still struggle with non-linear computational notebooks. In this paper, we present an AI agent designed specifically for error resolution in a computational notebook. We have developed an agentic system capable of exploring a notebook environment by interacting with it -- similar to how a user would -- and integrated the system into the JetBrains service for collaborative data science called Datalore. We evaluate our approach against the pre-existing single-action solution by comparing costs and conducting a user study. Users rate the error resolution capabilities of the agentic system higher but experience difficulties with UI. We share the results of the study and consider them valuable for further improving user-agent collaboration.
comment: Accepted to EMNLP 2024 System Demonstrations
☆ Personalizing Low-Rank Bayesian Neural Networks Via Federated Learning
To support real-world decision-making, it is crucial for models to be well-calibrated, i.e., to assign reliable confidence estimates to their predictions. Uncertainty quantification is particularly important in personalized federated learning (PFL), as participating clients typically have small local datasets, making it difficult to unambiguously determine optimal model parameters. Bayesian PFL (BPFL) methods can potentially enhance calibration, but they often come with considerable computational and memory requirements due to the need to track the variances of all the individual model parameters. Furthermore, different clients may exhibit heterogeneous uncertainty levels owing to varying local dataset sizes and distributions. To address these challenges, we propose LR-BPFL, a novel BPFL method that learns a global deterministic model along with personalized low-rank Bayesian corrections. To tailor the local model to each client's inherent uncertainty level, LR-BPFL incorporates an adaptive rank selection mechanism. We evaluate LR-BPFL across a variety of datasets, demonstrating its advantages in terms of calibration, accuracy, as well as computational and memory requirements.
☆ SurgeryV2: Bridging the Gap Between Model Merging and Multi-Task Learning with Deep Representation Surgery ICML 2024
Model merging-based multitask learning (MTL) offers a promising approach for performing MTL by merging multiple expert models without requiring access to raw training data. However, in this paper, we examine the merged model's representation distribution and uncover a critical issue of "representation bias". This bias arises from a significant distribution gap between the representations of the merged and expert models, leading to the suboptimal performance of the merged MTL model. To address this challenge, we first propose a representation surgery solution called Surgery. Surgery is a lightweight, task-specific module that aligns the final layer representations of the merged model with those of the expert models, effectively alleviating bias and improving the merged model's performance. Despite these improvements, a performance gap remains compared to the traditional MTL method. Further analysis reveals that representation bias phenomena exist at each layer of the merged model, and aligning representations only in the last layer is insufficient for fully reducing systemic bias because biases introduced at each layer can accumulate and interact in complex ways. To tackle this, we then propose a more comprehensive solution, deep representation surgery (also called SurgeryV2), which mitigates representation bias across all layers, and thus bridges the performance gap between model merging-based MTL and traditional MTL. Finally, we design an unsupervised optimization objective to optimize both the Surgery and SurgeryV2 modules. Our experimental results show that incorporating these modules into state-of-the-art (SOTA) model merging schemes leads to significant performance gains. Notably, our SurgeryV2 scheme reaches almost the same level as individual expert models or the traditional MTL model. The code is available at \url{https://github.com/EnnengYang/SurgeryV2}.
comment: This paper is an extended version of our previous work [arXiv:2402.02705] presented at ICML 2024
☆ Unscrambling disease progression at scale: fast inference of event permutations with optimal transport NeurIPS 2024
Disease progression models infer group-level temporal trajectories of change in patients' features as a chronic degenerative condition plays out. They provide unique insight into disease biology and staging systems with individual-level clinical utility. Discrete models consider disease progression as a latent permutation of events, where each event corresponds to a feature becoming measurably abnormal. However, permutation inference using traditional maximum likelihood approaches becomes prohibitive due to combinatoric explosion, severely limiting model dimensionality and utility. Here we leverage ideas from optimal transport to model disease progression as a latent permutation matrix of events belonging to the Birkhoff polytope, facilitating fast inference via optimisation of the variational lower bound. This enables a factor of 1000 times faster inference than the current state of the art and, correspondingly, supports models with several orders of magnitude more features than the current state of the art can consider. Experiments demonstrate the increase in speed, accuracy and robustness to noise in simulation. Further experiments with real-world imaging data from two separate datasets, one from Alzheimer's disease patients, the other age-related macular degeneration, showcase, for the first time, pixel-level disease progression events in the brain and eye, respectively. Our method is low compute, interpretable and applicable to any progressive condition and data modality, giving it broad potential clinical utility.
comment: Pre-print of version accepted to NeurIPS 2024
☆ Investigating the Capabilities of Deep Learning for Processing and Interpreting One-Shot Multi-offset GPR Data: A Numerical Case Study for Lunar and Martian Environments
Ground-penetrating radar (GPR) is a mature geophysical method that has gained increasing popularity in planetary science over the past decade. GPR has been utilised both for Lunar and Martian missions providing pivotal information regarding the near surface geology of Terrestrial planets. Within that context, numerous processing pipelines have been suggested to address the unique challenges present in planetary setups. These processing pipelines often require manual tuning resulting to ambiguous outputs open to non-unique interpretations. These pitfalls combined with the large number of planetary GPR data (kilometers in magnitude), highlight the necessity for automatic, objective and advanced processing and interpretation schemes. The current paper investigates the potential of deep learning for interpreting and processing GPR data. The one-shot multi-offset configuration is investigated via a coherent numerical case study, showcasing the potential of deep learning for A) reconstructing the dielectric distribution of the the near surface of Terrestrial planets, and B) filling missing or bad-quality traces. Special care was taken for the numerical data to be both realistic and challenging. Moreover, the generated synthetic data are properly labelled and made publicly available for training future data-driven pipelines and contributing towards developing pre-trained foundation models for GPR.
☆ Dual-Label LearningWith Irregularly Present Labels
In multi-task learning, we often encounter the case when the presence of labels across samples exhibits irregular patterns: samples can be fully labeled, partially labeled or unlabeled. Taking drug analysis as an example, multiple toxicity properties of a drug molecule may not be concurrently available due to experimental limitations. It triggers a demand for a new training and inference mechanism that could accommodate irregularly present labels and maximize the utility of any available label information. In this work, we focus on the two-label learning task, and propose a novel training and inference framework, Dual-Label Learning (DLL). The DLL framework formulates the problem into a dual-function system, in which the two functions should simultaneously satisfy standard supervision, structural duality and probabilistic duality. DLL features a dual-tower model architecture that explicitly captures the information exchange between labels, aimed at maximizing the utility of partially available labels in understanding label correlation. During training, label imputation for missing labels is conducted as part of the forward propagation process, while during inference, labels are regarded as unknowns of a bivariate system of equations and are solved jointly. Theoretical analysis guarantees the feasibility of DLL, and extensive experiments are conducted to verify that by explicitly modeling label correlation and maximizing the utility of available labels, our method makes consistently better predictions than baseline approaches by up to a 10% gain in F1-score or MAPE. Remarkably, our method provided with data at a label missing rate as high as 60% can achieve similar or even better results than baseline approaches at a label missing rate of only 10%.
☆ Fine-Tuning Pre-trained Language Models for Robust Causal Representation Learning
The fine-tuning of pre-trained language models (PLMs) has been shown to be effective across various domains. By using domain-specific supervised data, the general-purpose representation derived from PLMs can be transformed into a domain-specific representation. However, these methods often fail to generalize to out-of-domain (OOD) data due to their reliance on non-causal representations, often described as spurious features. Existing methods either make use of adjustments with strong assumptions about lack of hidden common causes, or mitigate the effect of spurious features using multi-domain data. In this work, we investigate how fine-tuned pre-trained language models aid generalizability from single-domain scenarios under mild assumptions, targeting more general and practical real-world scenarios. We show that a robust representation can be derived through a so-called causal front-door adjustment, based on a decomposition assumption, using fine-tuned representations as a source of data augmentation. Comprehensive experiments in both synthetic and real-world settings demonstrate the superior generalizability of the proposed method compared to existing approaches. Our work thus sheds light on the domain generalization problem by introducing links between fine-tuning and causal mechanisms into representation learning.
☆ A Scientific Machine Learning Approach for Predicting and Forecasting Battery Degradation in Electric Vehicles
Carbon emissions are rising at an alarming rate, posing a significant threat to global efforts to mitigate climate change. Electric vehicles have emerged as a promising solution, but their reliance on lithium-ion batteries introduces the critical challenge of battery degradation. Accurate prediction and forecasting of battery degradation over both short and long time spans are essential for optimizing performance, extending battery life, and ensuring effective long-term energy management. This directly influences the reliability, safety, and sustainability of EVs, supporting their widespread adoption and aligning with key UN SDGs. In this paper, we present a novel approach to the prediction and long-term forecasting of battery degradation using Scientific Machine Learning framework which integrates domain knowledge with neural networks, offering more interpretable and scientifically grounded solutions for both predicting short-term battery health and forecasting degradation over extended periods. This hybrid approach captures both known and unknown degradation dynamics, improving predictive accuracy while reducing data requirements. We incorporate ground-truth data to inform our models, ensuring that both the predictions and forecasts reflect practical conditions. The model achieved MSE of 9.90 with the UDE and 11.55 with the NeuralODE, in experimental data, a loss of 1.6986 with the UDE, and a MSE of 2.49 in the NeuralODE, demonstrating the enhanced precision of our approach. This integration of data-driven insights with SciML's strengths in interpretability and scalability allows for robust battery management. By enhancing battery longevity and minimizing waste, our approach contributes to the sustainability of energy systems and accelerates the global transition toward cleaner, more responsible energy solutions, aligning with the UN's SDG agenda.
☆ Evaluating the evaluators: Towards human-aligned metrics for missing markers reconstruction
Animation data is often obtained through optical motion capture systems, which utilize a multitude of cameras to establish the position of optical markers. However, system errors or occlusions can result in missing markers, the manual cleaning of which can be time-consuming. This has sparked interest in machine learning-based solutions for missing marker reconstruction in the academic community. Most academic papers utilize a simplistic mean square error as the main metric. In this paper, we show that this metric does not correlate with subjective perception of the fill quality. We introduce and evaluate a set of better-correlated metrics that can drive progress in the field.
☆ Fast proxy centers for Jeffreys centroids: The Jeffreys-Fisher-Rao and the inductive Gauss-Bregman centers
The symmetric Kullback-Leibler centroid also called the Jeffreys centroid of a set of mutually absolutely continuous probability distributions on a measure space provides a notion of centrality which has proven useful in many tasks including information retrieval, information fusion, and clustering in image, video and sound processing. However, the Jeffreys centroid is not available in closed-form for sets of categorical or normal distributions, two widely used statistical models, and thus need to be approximated numerically in practice. In this paper, we first propose the new Jeffreys-Fisher-Rao center defined as the Fisher-Rao midpoint of the sided Kullback-Leibler centroids as a plug-in replacement of the Jeffreys centroid. This Jeffreys-Fisher-Rao center admits a generic formula for uni-parameter exponential family distributions, and closed-form formula for categorical and normal distributions, matches exactly the Jeffreys centroid for same-mean normal distributions, and is experimentally observed in practice to be close to the Jeffreys centroid. Second, we define a new type of inductive centers generalizing the principle of Gauss arithmetic-geometric double sequence mean for pairs of densities of any given exponential family. This center is shown experimentally to approximate very well the Jeffreys centroid and is suggested to use when the Jeffreys-Fisher-Rao center is not available in closed form. Moreover, this Gauss-Bregman inductive center always converges and matches the Jeffreys centroid for sets of same-mean normal distributions. We report on our experiments demonstrating the use of the Jeffreys-Fisher-Rao and Gauss-Bregman centers instead of the Jeffreys centroid. Finally, we conclude this work by reinterpreting these fast proxy centers of Jeffreys centroids under the lens of dually flat spaces in information geometry.
comment: 35 pages, 10 figures
☆ Debiasing Mini-Batch Quadratics for Applications in Deep Learning
Quadratic approximations form a fundamental building block of machine learning methods. E.g., second-order optimizers try to find the Newton step into the minimum of a local quadratic proxy to the objective function; and the second-order approximation of a network's loss function can be used to quantify the uncertainty of its outputs via the Laplace approximation. When computations on the entire training set are intractable - typical for deep learning - the relevant quantities are computed on mini-batches. This, however, distorts and biases the shape of the associated stochastic quadratic approximations in an intricate way with detrimental effects on applications. In this paper, we (i) show that this bias introduces a systematic error, (ii) provide a theoretical explanation for it, (iii) explain its relevance for second-order optimization and uncertainty quantification via the Laplace approximation in deep learning, and (iv) develop and evaluate debiasing strategies.
comment: Main text (including references): 13 pages, 6 figures; Supplements: 25 pages, 13 figures
☆ Optimizing importance weighting in the presence of sub-population shifts
A distribution shift between the training and test data can severely harm performance of machine learning models. Importance weighting addresses this issue by assigning different weights to data points during training. We argue that existing heuristics for determining the weights are suboptimal, as they neglect the increase of the variance of the estimated model due to the finite sample size of the training data. We interpret the optimal weights in terms of a bias-variance trade-off, and propose a bi-level optimization procedure in which the weights and model parameters are optimized simultaneously. We apply this optimization to existing importance weighting techniques for last-layer retraining of deep neural networks in the presence of sub-population shifts and show empirically that optimizing weights significantly improves generalization performance.
comment: Preprint. Currently under review
☆ PTR: A Pre-trained Language Model for Trajectory Recovery
Spatiotemporal trajectory data is vital for web-of-things services and is extensively collected and analyzed by web-based hardware and platforms. However, issues such as service interruptions and network instability often lead to sparsely recorded trajectories, resulting in a loss of detailed movement data. As a result, recovering these trajectories to restore missing information becomes essential. Despite progress, several challenges remain unresolved. First, the lack of large-scale dense trajectory data hampers the performance of existing deep learning methods, which rely heavily on abundant data for supervised training. Second, current methods struggle to generalize across sparse trajectories with varying sampling intervals, necessitating separate re-training for each interval and increasing computational costs. Third, external factors crucial for the recovery of missing points are not fully incorporated. To address these challenges, we propose a framework called PTR. This framework mitigates the issue of limited dense trajectory data by leveraging the capabilities of pre-trained language models (PLMs). PTR incorporates an explicit trajectory prompt and is trained on datasets with multiple sampling intervals, enabling it to generalize effectively across different intervals in sparse trajectories. To capture external factors, we introduce an implicit trajectory prompt that models road conditions, providing richer information for recovering missing points. Additionally, we present a trajectory embedder that encodes trajectory points and transforms the embeddings of both observed and missing points into a format comprehensible to PLMs. Experimental results on two public trajectory datasets with three sampling intervals demonstrate the efficacy and scalability of PTR.
☆ Stochastic Quasi-Newton Optimization in Large Dimensions Including Deep Network Training
Our proposal is on a new stochastic optimizer for non-convex and possibly non-smooth objective functions typically defined over large dimensional design spaces. Towards this, we have tried to bridge noise-assisted global search and faster local convergence, the latter being the characteristic feature of a Newton-like search. Our specific scheme -- acronymed FINDER (Filtering Informed Newton-like and Derivative-free Evolutionary Recursion), exploits the nonlinear stochastic filtering equations to arrive at a derivative-free update that has resemblance with the Newton search employing the inverse Hessian of the objective function. Following certain simplifications of the update to enable a linear scaling with dimension and a few other enhancements, we apply FINDER to a range of problems, starting with some IEEE benchmark objective functions to a couple of archetypal data-driven problems in deep networks to certain cases of physics-informed deep networks. The performance of the new method vis-\'a-vis the well-known Adam and a few others bears evidence to its promise and potentialities for large dimensional optimization problems of practical interest.
comment: 19 pages, 12 figures, 3 tables
☆ On time series clustering with k-means
There is a long history of research into time series clustering using distance-based partitional clustering. Many of the most popular algorithms adapt k-means (also known as Lloyd's algorithm) to exploit time dependencies in the data by specifying a time series distance function. However, these algorithms are often presented with k-means configured in various ways, altering key parameters such as the initialisation strategy. This variability makes it difficult to compare studies because k-means is known to be highly sensitive to its configuration. To address this, we propose a standard Lloyd's-based model for TSCL that adopts an end-to-end approach, incorporating a specialised distance function not only in the assignment step but also in the initialisation and stopping criteria. By doing so, we create a unified structure for comparing seven popular Lloyd's-based TSCL algorithms. This common framework enables us to more easily attribute differences in clustering performance to the distance function itself, rather than variations in the k-means configuration.
☆ MoDification: Mixture of Depths Made Easy
Long-context efficiency has recently become a trending topic in serving large language models (LLMs). And mixture of depths (MoD) is proposed as a perfect fit to bring down both latency and memory. In this paper, however, we discover that MoD can barely transform existing LLMs without costly training over an extensive number of tokens. To enable the transformations from any LLMs to MoD ones, we showcase top-k operator in MoD should be promoted to threshold-p operator, and refinement to architecture and data should also be crafted along. All these designs form our method termed MoDification. Through a comprehensive set of experiments covering model scales from 3B to 70B, we exhibit MoDification strikes an excellent balance between efficiency and effectiveness. MoDification can achieve up to ~1.2x speedup in latency and ~1.8x reduction in memory compared to original LLMs especially in long-context applications.
comment: 12 pages, 9 figures, 5 tables, work in progress
☆ Revisiting SLO and Goodput Metrics in LLM Serving
Large language models (LLMs) have achieved remarkable performance and are widely deployed in various applications, while the serving of LLM inference has raised concerns about user experience and serving throughput. Accordingly, service level objectives (SLOs) and goodput-the number of requests that meet SLOs per second-are introduced to evaluate the performance of LLM serving. However, existing metrics fail to capture the nature of user experience. We observe two ridiculous phenomena in existing metrics: 1) delaying token delivery can smooth the tail time between tokens (tail TBT) of a request and 2) dropping the request that fails to meet the SLOs midway can improve goodput. In this paper, we revisit SLO and goodput metrics in LLM serving and propose a unified metric framework smooth goodput including SLOs and goodput to reflect the nature of user experience in LLM serving. The framework can adapt to specific goals of different tasks by setting parameters. We re-evaluate the performance of different LLM serving systems under multiple workloads based on this unified framework and provide possible directions for future optimization of existing strategies. We hope that this framework can provide a unified standard for evaluating LLM serving and foster researches in the field of LLM serving optimization to move in a cohesive direction.
☆ RAZOR: Refining Accuracy by Zeroing Out Redundancies
In many application domains, the proliferation of sensors and devices is generating vast volumes of data, imposing significant pressure on existing data analysis and data mining techniques. Nevertheless, an increase in data volume does not inherently imply an increase in informational content, as a substantial portion may be redundant or represent noise. This challenge is particularly evident in the deep learning domain, where the utility of additional data is contingent on its informativeness. In the absence of such, larger datasets merely exacerbate the computational cost and complexity of the learning process. To address these challenges, we propose RAZOR, a novel instance selection technique designed to extract a significantly smaller yet sufficiently informative subset from a larger set of instances without compromising the learning process. RAZOR has been specifically engineered to be robust, efficient, and scalable, making it suitable for large-scale datasets. Unlike many techniques in the literature, RAZOR is capable of operating in both supervised and unsupervised settings. Experimental results demonstrate that RAZOR outperforms recent state-of-the-art techniques in terms of both effectiveness and efficiency.
comment: 17 pages, 3 figures
☆ Pseudo-label Refinement for Improving Self-Supervised Learning Systems
Self-supervised learning systems have gained significant attention in recent years by leveraging clustering-based pseudo-labels to provide supervision without the need for human annotations. However, the noise in these pseudo-labels caused by the clustering methods poses a challenge to the learning process leading to degraded performance. In this work, we propose a pseudo-label refinement (SLR) algorithm to address this issue. The cluster labels from the previous epoch are projected to the current epoch cluster-labels space and a linear combination of the new label and the projected label is computed as a soft refined label containing the information from the previous epoch clusters as well as from the current epoch. In contrast to the common practice of using the maximum value as a cluster/class indicator, we employ hierarchical clustering on these soft pseudo-labels to generate refined hard-labels. This approach better utilizes the information embedded in the soft labels, outperforming the simple maximum value approach for hard label generation. The effectiveness of the proposed SLR algorithm is evaluated in the context of person re-identification (Re-ID) using unsupervised domain adaptation (UDA). Experimental results demonstrate that the modified Re-ID baseline, incorporating the SLR algorithm, achieves significantly improved mean Average Precision (mAP) performance in various UDA tasks, including real-to-synthetic, synthetic-to-real, and different real-to-real scenarios. These findings highlight the efficacy of the SLR algorithm in enhancing the performance of self-supervised learning systems.
☆ Almost-Linear RNNs Yield Highly Interpretable Symbolic Codes in Dynamical Systems Reconstruction NeurIPS 2024
Dynamical systems (DS) theory is fundamental for many areas of science and engineering. It can provide deep insights into the behavior of systems evolving in time, as typically described by differential or recursive equations. A common approach to facilitate mathematical tractability and interpretability of DS models involves decomposing nonlinear DS into multiple linear DS separated by switching manifolds, i.e. piecewise linear (PWL) systems. PWL models are popular in engineering and a frequent choice in mathematics for analyzing the topological properties of DS. However, hand-crafting such models is tedious and only possible for very low-dimensional scenarios, while inferring them from data usually gives rise to unnecessarily complex representations with very many linear subregions. Here we introduce Almost-Linear Recurrent Neural Networks (AL-RNNs) which automatically and robustly produce most parsimonious PWL representations of DS from time series data, using as few PWL nonlinearities as possible. AL-RNNs can be efficiently trained with any SOTA algorithm for dynamical systems reconstruction (DSR), and naturally give rise to a symbolic encoding of the underlying DS that provably preserves important topological properties. We show that for the Lorenz and R\"ossler systems, AL-RNNs discover, in a purely data-driven way, the known topologically minimal PWL representations of the corresponding chaotic attractors. We further illustrate on two challenging empirical datasets that interpretable symbolic encodings of the dynamics can be achieved, tremendously facilitating mathematical and computational analysis of the underlying systems.
comment: 38th Conference on Neural Information Processing Systems (NeurIPS 2024)
☆ Unified Convergence Analysis for Score-Based Diffusion Models with Deterministic Samplers
Score-based diffusion models have emerged as powerful techniques for generating samples from high-dimensional data distributions. These models involve a two-phase process: first, injecting noise to transform the data distribution into a known prior distribution, and second, sampling to recover the original data distribution from noise. Among the various sampling methods, deterministic samplers stand out for their enhanced efficiency. However, analyzing these deterministic samplers presents unique challenges, as they preclude the use of established techniques such as Girsanov's theorem, which are only applicable to stochastic samplers. Furthermore, existing analysis for deterministic samplers usually focuses on specific examples, lacking a generalized approach for general forward processes and various deterministic samplers. Our paper addresses these limitations by introducing a unified convergence analysis framework. To demonstrate the power of our framework, we analyze the variance-preserving (VP) forward process with the exponential integrator (EI) scheme, achieving iteration complexity of $\tilde O(d^2/\epsilon)$. Additionally, we provide a detailed analysis of Denoising Diffusion Implicit Models (DDIM)-type samplers, which have been underexplored in previous research, achieving polynomial iteration complexity.
comment: 68 pages
☆ G-NeuroDAVIS: A Neural Network model for generalized embedding, data visualization and sample generation
Visualizing high-dimensional datasets through a generalized embedding has been a challenge for a long time. Several methods have shown up for the same, but still, they have not been able to generate a generalized embedding, which not only can reveal the hidden patterns present in the data but also generate realistic high-dimensional samples from it. Motivated by this aspect, in this study, a novel generative model, called G-NeuroDAVIS, has been developed, which is capable of visualizing high-dimensional data through a generalized embedding, and thereby generating new samples. The model leverages advanced generative techniques to produce high-quality embedding that captures the underlying structure of the data more effectively than existing methods. G-NeuroDAVIS can be trained in both supervised and unsupervised settings. We rigorously evaluated our model through a series of experiments, demonstrating superior performance in classification tasks, which highlights the robustness of the learned representations. Furthermore, the conditional sample generation capability of the model has been described through qualitative assessments, revealing a marked improvement in generating realistic and diverse samples. G-NeuroDAVIS has outperformed the Variational Autoencoder (VAE) significantly in multiple key aspects, including embedding quality, classification performance, and sample generation capability. These results underscore the potential of our generative model to serve as a powerful tool in various applications requiring high-quality data generation and representation learning.
comment: 15 pages, 8 figures
☆ Formal Explanations for Neuro-Symbolic AI
Despite the practical success of Artificial Intelligence (AI), current neural AI algorithms face two significant issues. First, the decisions made by neural architectures are often prone to bias and brittleness. Second, when a chain of reasoning is required, neural systems often perform poorly. Neuro-symbolic artificial intelligence is a promising approach that tackles these (and other) weaknesses by combining the power of neural perception and symbolic reasoning. Meanwhile, the success of AI has made it critical to understand its behaviour, leading to the development of explainable artificial intelligence (XAI). While neuro-symbolic AI systems have important advantages over purely neural AI, we still need to explain their actions, which are obscured by the interactions of the neural and symbolic components. To address the issue, this paper proposes a formal approach to explaining the decisions of neuro-symbolic systems. The approach hinges on the use of formal abductive explanations and on solving the neuro-symbolic explainability problem hierarchically. Namely, it first computes a formal explanation for the symbolic component of the system, which serves to identify a subset of the individual parts of neural information that needs to be explained. This is followed by explaining only those individual neural inputs, independently of each other, which facilitates succinctness of hierarchical formal explanations and helps to increase the overall performance of the approach. Experimental results for a few complex reasoning tasks demonstrate practical efficiency of the proposed approach, in comparison to purely neural systems, from the perspective of explanation size, explanation time, training time, model sizes, and the quality of explanations reported.
☆ Comparative Evaluation of Clustered Federated Learning Method
Over recent years, Federated Learning (FL) has proven to be one of the most promising methods of distributed learning which preserves data privacy. As the method evolved and was confronted to various real-world scenarios, new challenges have emerged. One such challenge is the presence of highly heterogeneous (often referred as non-IID) data distributions among participants of the FL protocol. A popular solution to this hurdle is Clustered Federated Learning (CFL), which aims to partition clients into groups where the distribution are homogeneous. In the literature, state-of-the-art CFL algorithms are often tested using a few cases of data heterogeneities, without systematically justifying the choices. Further, the taxonomy used for differentiating the different heterogeneity scenarios is not always straightforward. In this paper, we explore the performance of two state-of-theart CFL algorithms with respect to a proposed taxonomy of data heterogeneities in federated learning (FL). We work with three image classification datasets and analyze the resulting clusters against the heterogeneity classes using extrinsic clustering metrics. Our objective is to provide a clearer understanding of the relationship between CFL performances and data heterogeneity scenarios.
☆ Montessori-Instruct: Generate Influential Training Data Tailored for Student Learning
Synthetic data has been widely used to train large language models, but their generative nature inevitably introduces noisy, non-informative, and misleading learning signals. In this paper, we propose Montessori-Instruct, a novel data synthesis framework that tailors the data synthesis ability of the teacher language model toward the student language model's learning process. Specifically, we utilize local data influence of synthetic training data points on students to characterize students' learning preferences. Then, we train the teacher model with Direct Preference Optimization (DPO) to generate synthetic data tailored toward student learning preferences. Experiments with Llama3-8B-Instruct (teacher) and Llama3-8B (student) on Alpaca Eval and MT-Bench demonstrate that Montessori-Instruct significantly outperforms standard synthesis methods by 18.35\% and 46.24\% relatively. Our method also beats data synthesized by a stronger teacher model, GPT-4o. Further analysis confirms the benefits of teacher's learning to generate more influential training data in the student's improved learning, the advantages of local data influence in accurately measuring student preferences, and the robustness of Montessori-Instruct across different student models. Our code and data are open-sourced at https://github.com/cxcscmu/Montessori-Instruct.
comment: Codes and data are open-sourced at https://github.com/cxcscmu/Montessori-Instruct
☆ Flexi-Fuzz least squares SVM for Alzheimer's diagnosis: Tackling noise, outliers, and class imbalance
Alzheimer's disease (AD) is a leading neurodegenerative condition and the primary cause of dementia, characterized by progressive cognitive decline and memory loss. Its progression, marked by shrinkage in the cerebral cortex, is irreversible. Numerous machine learning algorithms have been proposed for the early diagnosis of AD. However, they often struggle with the issues of noise, outliers, and class imbalance. To tackle the aforementioned limitations, in this article, we introduce a novel, robust, and flexible membership scheme called Flexi-Fuzz. This scheme integrates a novel flexible weighting mechanism, class probability, and imbalance ratio. The proposed flexible weighting mechanism assigns the maximum weight to samples within a specific proximity to the center, with a gradual decrease in weight beyond a certain threshold. This approach ensures that samples near the class boundary still receive significant weight, maintaining their influence in the classification process. Class probability is used to mitigate the impact of noisy samples, while the imbalance ratio addresses class imbalance. Leveraging this, we incorporate the proposed Flexi-Fuzz membership scheme into the least squares support vector machines (LSSVM) framework, resulting in a robust and flexible model termed Flexi-Fuzz-LSSVM. We determine the class-center using two methods: the conventional mean approach and an innovative median approach, leading to two model variants, Flexi-Fuzz-LSSVM-I and Flexi-Fuzz-LSSVM-II. To validate the effectiveness of the proposed Flexi-Fuzz-LSSVM models, we evaluated them on benchmark UCI and KEEL datasets, both with and without label noise. Additionally, we tested the models on the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset for AD diagnosis. Experimental results demonstrate the superiority of the Flexi-Fuzz-LSSVM models over baseline models.
☆ xPerT: Extended Persistence Transformer
A persistence diagram provides a compact summary of persistent homology, which captures the topological features of a space at different scales. However, due to its nature as a set, incorporating it as a feature into a machine learning framework is challenging. Several methods have been proposed to use persistence diagrams as input for machine learning models, but they often require complex preprocessing steps and extensive hyperparameter tuning. In this paper, we propose a novel transformer architecture called the \textit{Extended Persistence Transformer (xPerT)}, which is highly scalable than the compared to Persformer, an existing transformer for persistence diagrams. xPerT reduces GPU memory usage by over 90\% and improves accuracy on multiple datasets. Additionally, xPerT does not require complex preprocessing steps or extensive hyperparameter tuning, making it easy to use in practice. Our code is available at https://github.com/sehunfromdaegu/ECG_JEPA.
☆ Combining Hough Transform and Deep Learning Approaches to Reconstruct ECG Signals From Printouts
This work presents our team's (SignalSavants) winning contribution to the 2024 George B. Moody PhysioNet Challenge. The Challenge had two goals: reconstruct ECG signals from printouts and classify them for cardiac diseases. Our focus was the first task. Despite many ECGs being digitally recorded today, paper ECGs remain common throughout the world. Digitising them could help build more diverse datasets and enable automated analyses. However, the presence of varying recording standards and poor image quality requires a data-centric approach for developing robust models that can generalise effectively. Our approach combines the creation of a diverse training set, Hough transform to rotate images, a U-Net based segmentation model to identify individual signals, and mask vectorisation to reconstruct the signals. We assessed the performance of our models using the 10-fold stratified cross-validation (CV) split of 21,799 recordings proposed by the PTB-XL dataset. On the digitisation task, our model achieved an average CV signal-to-noise ratio of 17.02 and an official Challenge score of 12.15 on the hidden set, securing first place in the competition. Our study shows the challenges of building robust, generalisable, digitisation approaches. Such models require large amounts of resources (data, time, and computational power) but have great potential in diversifying the data available.
☆ Provable In-context Learning for Mixture of Linear Regressions using Transformers
We theoretically investigate the in-context learning capabilities of transformers in the context of learning mixtures of linear regression models. For the case of two mixtures, we demonstrate the existence of transformers that can achieve an accuracy, relative to the oracle predictor, of order $\mathcal{\tilde{O}}((d/n)^{1/4})$ in the low signal-to-noise ratio (SNR) regime and $\mathcal{\tilde{O}}(\sqrt{d/n})$ in the high SNR regime, where $n$ is the length of the prompt, and $d$ is the dimension of the problem. Additionally, we derive in-context excess risk bounds of order $\mathcal{O}(L/\sqrt{B})$, where $B$ denotes the number of (training) prompts, and $L$ represents the number of attention layers. The order of $L$ depends on whether the SNR is low or high. In the high SNR regime, we extend the results to $K$-component mixture models for finite $K$. Extensive simulations also highlight the advantages of transformers for this task, outperforming other baselines such as the Expectation-Maximization algorithm.
☆ LabSafety Bench: Benchmarking LLMs on Safety Issues in Scientific Labs
Laboratory accidents pose significant risks to human life and property, underscoring the importance of robust safety protocols. Despite advancements in safety training, laboratory personnel may still unknowingly engage in unsafe practices. With the increasing reliance on large language models (LLMs) for guidance in various fields, including laboratory settings, there is a growing concern about their reliability in critical safety-related decision-making. Unlike trained human researchers, LLMs lack formal lab safety education, raising questions about their ability to provide safe and accurate guidance. Existing research on LLM trustworthiness primarily focuses on issues such as ethical compliance, truthfulness, and fairness but fails to fully cover safety-critical real-world applications, like lab safety. To address this gap, we propose the Laboratory Safety Benchmark (LabSafety Bench), a comprehensive evaluation framework based on a new taxonomy aligned with Occupational Safety and Health Administration (OSHA) protocols. This benchmark includes 765 multiple-choice questions verified by human experts, assessing LLMs and vision language models (VLMs) performance in lab safety contexts. Our evaluations demonstrate that while GPT-4o outperforms human participants, it is still prone to critical errors, highlighting the risks of relying on LLMs in safety-critical environments. Our findings emphasize the need for specialized benchmarks to accurately assess the trustworthiness of LLMs in real-world safety applications.
comment: 50 pages, 19 figures
☆ Auto Detecting Cognitive Events Using Machine Learning on Pupillary Data
Assessing cognitive workload is crucial for human performance as it affects information processing, decision making, and task execution. Pupil size is a valuable indicator of cognitive workload, reflecting changes in attention and arousal governed by the autonomic nervous system. Cognitive events are closely linked to cognitive workload as they activate mental processes and trigger cognitive responses. This study explores the potential of using machine learning to automatically detect cognitive events experienced using individuals. We framed the problem as a binary classification task, focusing on detecting stimulus onset across four cognitive tasks using CNN models and 1-second pupillary data. The results, measured by Matthew's correlation coefficient, ranged from 0.47 to 0.80, depending on the cognitive task. This paper discusses the trade-offs between generalization and specialization, model behavior when encountering unseen stimulus onset times, structural variances among cognitive tasks, factors influencing model predictions, and real-time simulation. These findings highlight the potential of machine learning techniques in detecting cognitive events based on pupil and eye movement responses, contributing to advancements in personalized learning and optimizing neurocognitive workload management.
comment: 10 pages, 7 figures
☆ Heavy-Tailed Diffusion Models
Diffusion models achieve state-of-the-art generation quality across many applications, but their ability to capture rare or extreme events in heavy-tailed distributions remains unclear. In this work, we show that traditional diffusion and flow-matching models with standard Gaussian priors fail to capture heavy-tailed behavior. We address this by repurposing the diffusion framework for heavy-tail estimation using multivariate Student-t distributions. We develop a tailored perturbation kernel and derive the denoising posterior based on the conditional Student-t distribution for the backward process. Inspired by $\gamma$-divergence for heavy-tailed distributions, we derive a training objective for heavy-tailed denoisers. The resulting framework introduces controllable tail generation using only a single scalar hyperparameter, making it easily tunable for diverse real-world distributions. As specific instantiations of our framework, we introduce t-EDM and t-Flow, extensions of existing diffusion and flow models that employ a Student-t prior. Remarkably, our approach is readily compatible with standard Gaussian diffusion models and requires only minimal code changes. Empirically, we show that our t-EDM and t-Flow outperform standard diffusion models in heavy-tail estimation on high-resolution weather datasets in which generating rare and extreme events is crucial.
comment: 51 pages, Contains GIF animations and is best viewed with a dedicated pdf reader
☆ Assessing Open-world Forgetting in Generative Image Model Customization
Recent advances in diffusion models have significantly enhanced image generation capabilities. However, customizing these models with new classes often leads to unintended consequences that compromise their reliability. We introduce the concept of open-world forgetting to emphasize the vast scope of these unintended alterations, contrasting it with the well-studied closed-world forgetting, which is measurable by evaluating performance on a limited set of classes or skills. Our research presents the first comprehensive investigation into open-world forgetting in diffusion models, focusing on semantic and appearance drift of representations. We utilize zero-shot classification to analyze semantic drift, revealing that even minor model adaptations lead to unpredictable shifts affecting areas far beyond newly introduced concepts, with dramatic drops in zero-shot classification of up to 60%. Additionally, we observe significant changes in texture and color of generated content when analyzing appearance drift. To address these issues, we propose a mitigation strategy based on functional regularization, designed to preserve original capabilities while accommodating new concepts. Our study aims to raise awareness of unintended changes due to model customization and advocates for the analysis of open-world forgetting in future research on model customization and finetuning methods. Furthermore, we provide insights for developing more robust adaptation methodologies.
comment: Project page: https://hecoding.github.io/open-world-forgetting/
☆ A Mirror Descent Perspective of Smoothed Sign Descent
Recent work by Woodworth et al. (2020) shows that the optimization dynamics of gradient descent for overparameterized problems can be viewed as low-dimensional dual dynamics induced by a mirror map, explaining the implicit regularization phenomenon from the mirror descent perspective. However, the methodology does not apply to algorithms where update directions deviate from true gradients, such as ADAM. We use the mirror descent framework to study the dynamics of smoothed sign descent with a stability constant $\varepsilon$ for regression problems. We propose a mirror map that establishes equivalence to dual dynamics under some assumptions. By studying dual dynamics, we characterize the convergent solution as an approximate KKT point of minimizing a Bregman divergence style function, and show the benefit of tuning the stability constant $\varepsilon$ to reduce the KKT error.
☆ Beyond Autoregression: Discrete Diffusion for Complex Reasoning and Planning
Autoregressive language models, despite their impressive capabilities, struggle with complex reasoning and long-term planning tasks. We introduce discrete diffusion models as a novel solution to these challenges. Through the lens of subgoal imbalance, we demonstrate how diffusion models effectively learn difficult subgoals that elude autoregressive approaches. We propose Multi-granularity Diffusion Modeling (MDM), which prioritizes subgoals based on difficulty during learning. On complex tasks like Countdown, Sudoku, and Boolean Satisfiability Problems, MDM significantly outperforms autoregressive models without using search techniques. For instance, MDM achieves 91.5\% and 100\% accuracy on Countdown and Sudoku, respectively, compared to 45.8\% and 20.7\% for autoregressive models. Our work highlights the potential of diffusion-based approaches in advancing AI capabilities for sophisticated language understanding and problem-solving tasks.
☆ Wireless Human-Machine Collaboration in Industry 5.0
Wireless Human-Machine Collaboration (WHMC) represents a critical advancement for Industry 5.0, enabling seamless interaction between humans and machines across geographically distributed systems. As the WHMC systems become increasingly important for achieving complex collaborative control tasks, ensuring their stability is essential for practical deployment and long-term operation. Stability analysis certifies how the closed-loop system will behave under model randomness, which is essential for systems operating with wireless communications. However, the fundamental stability analysis of the WHMC systems remains an unexplored challenge due to the intricate interplay between the stochastic nature of wireless communications, dynamic human operations, and the inherent complexities of control system dynamics. This paper establishes a fundamental WHMC model incorporating dual wireless loops for machine and human control. Our framework accounts for practical factors such as short-packet transmissions, fading channels, and advanced HARQ schemes. We model human control lag as a Markov process, which is crucial for capturing the stochastic nature of human interactions. Building on this model, we propose a stochastic cycle-cost-based approach to derive a stability condition for the WHMC system, expressed in terms of wireless channel statistics, human dynamics, and control parameters. Our findings are validated through extensive numerical simulations and a proof-of-concept experiment, where we developed and tested a novel wireless collaborative cart-pole control system. The results confirm the effectiveness of our approach and provide a robust framework for future research on WHMC systems in more complex environments.
comment: Paper accepted by IEEE Transactions on Automatic Control
☆ CausalChat: Interactive Causal Model Development and Refinement Using Large Language Models
Causal networks are widely used in many fields to model the complex relationships between variables. A recent approach has sought to construct causal networks by leveraging the wisdom of crowds through the collective participation of humans. While this can yield detailed causal networks that model the underlying phenomena quite well, it requires a large number of individuals with domain understanding. We adopt a different approach: leveraging the causal knowledge that large language models, such as OpenAI's GPT-4, have learned by ingesting massive amounts of literature. Within a dedicated visual analytics interface, called CausalChat, users explore single variables or variable pairs recursively to identify causal relations, latent variables, confounders, and mediators, constructing detailed causal networks through conversation. Each probing interaction is translated into a tailored GPT-4 prompt and the response is conveyed through visual representations which are linked to the generated text for explanations. We demonstrate the functionality of CausalChat across diverse data contexts and conduct user studies involving both domain experts and laypersons.
☆ Preview-based Category Contrastive Learning for Knowledge Distillation
Knowledge distillation is a mainstream algorithm in model compression by transferring knowledge from the larger model (teacher) to the smaller model (student) to improve the performance of student. Despite many efforts, existing methods mainly investigate the consistency between instance-level feature representation or prediction, which neglects the category-level information and the difficulty of each sample, leading to undesirable performance. To address these issues, we propose a novel preview-based category contrastive learning method for knowledge distillation (PCKD). It first distills the structural knowledge of both instance-level feature correspondence and the relation between instance features and category centers in a contrastive learning fashion, which can explicitly optimize the category representation and explore the distinct correlation between representations of instances and categories, contributing to discriminative category centers and better classification results. Besides, we introduce a novel preview strategy to dynamically determine how much the student should learn from each sample according to their difficulty. Different from existing methods that treat all samples equally and curriculum learning that simply filters out hard samples, our method assigns a small weight for hard instances as a preview to better guide the student training. Extensive experiments on several challenging datasets, including CIFAR-100 and ImageNet, demonstrate the superiority over state-of-the-art methods.
comment: 14 pages, 8 figures, Journal
☆ Hierarchical Conditional Multi-Task Learning for Streamflow Modeling
Streamflow, vital for water resource management, is governed by complex hydrological systems involving intermediate processes driven by meteorological forces. While deep learning models have achieved state-of-the-art results of streamflow prediction, their end-to-end single-task learning approach often fails to capture the causal relationships within these systems. To address this, we propose Hierarchical Conditional Multi-Task Learning (HCMTL), a hierarchical approach that jointly models soil water and snowpack processes based on their causal connections to streamflow. HCMTL utilizes task embeddings to connect network modules, enhancing flexibility and expressiveness while capturing unobserved processes beyond soil water and snowpack. It also incorporates the Conditional Mini-Batch strategy to improve long time series modeling. We compare HCMTL with five baselines on a global dataset. HCMTL's superior performance across hundreds of drainage basins over extended periods shows that integrating domain-specific causal knowledge into deep learning enhances both prediction accuracy and interpretability. This is essential for advancing our understanding of complex hydrological systems and supporting efficient water resource management to mitigate natural disasters like droughts and floods.
☆ Inverse Reinforcement Learning from Non-Stationary Learning Agents
In this paper, we study an inverse reinforcement learning problem that involves learning the reward function of a learning agent using trajectory data collected while this agent is learning its optimal policy. To address this problem, we propose an inverse reinforcement learning method that allows us to estimate the policy parameters of the learning agent which can then be used to estimate its reward function. Our method relies on a new variant of the behavior cloning algorithm, which we call bundle behavior cloning, and uses a small number of trajectories generated by the learning agent's policy at different points in time to learn a set of policies that match the distribution of actions observed in the sampled trajectories. We then use the cloned policies to train a neural network model that estimates the reward function of the learning agent. We provide a theoretical analysis to show a complexity result on bound guarantees for our method that beats standard behavior cloning as well as numerical experiments for a reinforcement learning problem that validate the proposed method.
☆ Estimating the Causal Effects of T Cell Receptors
A central question in human immunology is how a patient's repertoire of T cells impacts disease. Here, we introduce a method to infer the causal effects of T cell receptor (TCR) sequences on patient outcomes using observational TCR repertoire sequencing data and clinical outcomes data. Our approach corrects for unobserved confounders, such as a patient's environment and life history, by using the patient's immature, pre-selection TCR repertoire. The pre-selection repertoire can be estimated from nonproductive TCR data, which is widely available. It is generated by a randomized mutational process, V(D)J recombination, which provides a natural experiment. We show formally how to use the pre-selection repertoire to draw causal inferences, and develop a scalable neural-network estimator for our identification formula. Our method produces an estimate of the effect of interventions that add a specific TCR sequence to patient repertoires. As a demonstration, we use it to analyze the effects of TCRs on COVID-19 severity, uncovering potentially therapeutic TCRs that are (1) observed in patients, (2) bind SARS-CoV-2 antigens in vitro and (3) have strong positive effects on clinical outcomes.
☆ Towards Robust Transcription: Exploring Noise Injection Strategies for Training Data Augmentation
Recent advancements in Automatic Piano Transcription (APT) have significantly improved system performance, but the impact of noisy environments on the system performance remains largely unexplored. This study investigates the impact of white noise at various Signal-to-Noise Ratio (SNR) levels on state-of-the-art APT models and evaluates the performance of the Onsets and Frames model when trained on noise-augmented data. We hope this research provides valuable insights as preliminary work toward developing transcription models that maintain consistent performance across a range of acoustic conditions.
comment: Accepted to the Late-Breaking Demo Session of the 25th International Society for Music Information Retrieval (ISMIR) Conference, 2024
☆ FedMSE: Federated learning for IoT network intrusion detection
This paper proposes a novel federated learning approach for improving IoT network intrusion detection. The rise of IoT has expanded the cyber attack surface, making traditional centralized machine learning methods insufficient due to concerns about data availability, computational resources, transfer costs, and especially privacy preservation. A semi-supervised federated learning model was developed to overcome these issues, combining the Shrink Autoencoder and Centroid one-class classifier (SAE-CEN). This approach enhances the performance of intrusion detection by effectively representing normal network data and accurately identifying anomalies in the decentralized strategy. Additionally, a mean square error-based aggregation algorithm (MSEAvg) was introduced to improve global model performance by prioritizing more accurate local models. The results obtained in our experimental setup, which uses various settings relying on the N-BaIoT dataset and Dirichlet distribution, demonstrate significant improvements in real-world heterogeneous IoT networks in detection accuracy from 93.98$\pm$2.90 to 97.30$\pm$0.49, reduced learning costs when requiring only 50\% of gateways participating in the training process, and robustness in large-scale networks.
☆ Skill Generalization with Verbs IROS 2023
It is imperative that robots can understand natural language commands issued by humans. Such commands typically contain verbs that signify what action should be performed on a given object and that are applicable to many objects. We propose a method for generalizing manipulation skills to novel objects using verbs. Our method learns a probabilistic classifier that determines whether a given object trajectory can be described by a specific verb. We show that this classifier accurately generalizes to novel object categories with an average accuracy of 76.69% across 13 object categories and 14 verbs. We then perform policy search over the object kinematics to find an object trajectory that maximizes classifier prediction for a given verb. Our method allows a robot to generate a trajectory for a novel object based on a verb, which can then be used as input to a motion planner. We show that our model can generate trajectories that are usable for executing five verb commands applied to novel instances of two different object categories on a real robot.
comment: 7 pages + 2 pages (references), 6 figures. Accepted at IROS 2023. Code, dataset info and demo videos can be found at: https://rachelma80000.github.io/SkillGenVerbs/
☆ A Communication and Computation Efficient Fully First-order Method for Decentralized Bilevel Optimization
Bilevel optimization, crucial for hyperparameter tuning, meta-learning and reinforcement learning, remains less explored in the decentralized learning paradigm, such as decentralized federated learning (DFL). Typically, decentralized bilevel methods rely on both gradients and Hessian matrices to approximate hypergradients of upper-level models. However, acquiring and sharing the second-order oracle is compute and communication intensive. % and sharing this information incurs heavy communication overhead. To overcome these challenges, this paper introduces a fully first-order decentralized method for decentralized Bilevel optimization, $\text{C}^2$DFB which is both compute- and communicate-efficient. In $\text{C}^2$DFB, each learning node optimizes a min-min-max problem to approximate hypergradient by exclusively using gradients information. To reduce the traffic load at the inner-loop of solving the lower-level problem, $\text{C}^2$DFB incorporates a lightweight communication protocol for efficiently transmitting compressed residuals of local parameters. % during the inner loops. Rigorous theoretical analysis ensures its convergence % of the algorithm, indicating a first-order oracle calls of $\tilde{\mathcal{O}}(\epsilon^{-4})$. Experiments on hyperparameter tuning and hyper-representation tasks validate the superiority of $\text{C}^2$DFB across various typologies and heterogeneous data distributions.
comment: 19 Pages
☆ Improving Graph Neural Networks by Learning Continuous Edge Directions
Graph Neural Networks (GNNs) traditionally employ a message-passing mechanism that resembles diffusion over undirected graphs, which often leads to homogenization of node features and reduced discriminative power in tasks such as node classification. Our key insight for addressing this limitation is to assign fuzzy edge directions -- that can vary continuously from node $i$ pointing to node $j$ to vice versa -- to the edges of a graph so that features can preferentially flow in one direction between nodes to enable long-range information transmission across the graph. We also introduce a novel complex-valued Laplacian for directed graphs with fuzzy edges where the real and imaginary parts represent information flow in opposite directions. Using this Laplacian, we propose a general framework, called Continuous Edge Direction (CoED) GNN, for learning on graphs with fuzzy edges and prove its expressivity limits using a generalization of the Weisfeiler-Leman (WL) graph isomorphism test for directed graphs with fuzzy edges. Our architecture aggregates neighbor features scaled by the learned edge directions and processes the aggregated messages from in-neighbors and out-neighbors separately alongside the self-features of the nodes. Since continuous edge directions are differentiable, they can be learned jointly with the GNN weights via gradient-based optimization. CoED GNN is particularly well-suited for graph ensemble data where the graph structure remains fixed but multiple realizations of node features are available, such as in gene regulatory networks, web connectivity graphs, and power grids. We demonstrate through extensive experiments on both synthetic and real datasets that learning continuous edge directions significantly improves performance both for undirected and directed graphs compared with existing methods.
♻ ☆ Locate-then-edit for Multi-hop Factual Recall under Knowledge Editing
The locate-then-edit paradigm has shown significant promise for knowledge editing (KE) in Large Language Models (LLMs). While previous methods perform well on single-hop fact recall tasks, they consistently struggle with multi-hop factual recall tasks involving newly edited knowledge. In this paper, leveraging tools in mechanistic interpretability, we first identify that in multi-hop tasks, LLMs tend to retrieve implicit subject knowledge from deeper MLP layers, unlike single-hop tasks, which rely on earlier layers. This distinction explains the poor performance of current methods in multi-hop queries, as they primarily focus on editing shallow layers, leaving deeper layers unchanged. To address this, we propose IFMET, a novel locate-then-edit KE approach designed to edit both shallow and deep MLP layers. IFMET employs multi-hop editing prompts and supplementary sets to locate and modify knowledge across different reasoning stages. Experimental results demonstrate that IFMET significantly improves performance on multi-hop factual recall tasks, effectively overcoming the limitations of previous locate-then-edit methods.
comment: 21 pages
♻ ☆ A Distance-based Anomaly Detection Framework for Deep Reinforcement Learning
In deep reinforcement learning (RL) systems, abnormal states pose significant risks by potentially triggering unpredictable behaviors and unsafe actions, thus impeding the deployment of RL systems in real-world scenarios. It is crucial for reliable decision-making systems to have the capability to cast an alert whenever they encounter unfamiliar observations that they are not equipped to handle. In this paper, we propose a novel Mahalanobis distance-based (MD) anomaly detection framework, called \textit{MDX}, for deep RL algorithms. MDX simultaneously addresses random, adversarial, and out-of-distribution (OOD) state outliers in both offline and online settings. It utilizes Mahalanobis distance within class-conditional distributions for each action and operates within a statistical hypothesis testing framework under the Gaussian assumption. We further extend it to robust and distribution-free versions by incorporating Robust MD and conformal inference techniques. Through extensive experiments on classical control environments, Atari games, and autonomous driving scenarios, we demonstrate the effectiveness of our MD-based detection framework. MDX offers a simple, unified, and practical anomaly detection tool for enhancing the safety and reliability of RL systems in real-world applications.
comment: 19 pages, 21 figures
♻ ☆ Liger Kernel: Efficient Triton Kernels for LLM Training
Training Large Language Models (LLMs) efficiently at scale presents a formidable challenge, driven by their ever-increasing computational demands and the need for enhanced performance. In this work, we introduce Liger-Kernel, an open-sourced set of Triton kernels developed specifically for LLM training. With kernel optimization techniques like kernel operation fusing and input chunking, our kernels achieve on average a 20% increase in training throughput and a 60% reduction in GPU memory usage for popular LLMs compared to HuggingFace implementations. In addition, Liger-Kernel is designed with modularity, accessibility, and adaptability in mind, catering to both casual and expert users. Comprehensive benchmarks and integration tests are built in to ensure compatibility, performance, correctness, and convergence across diverse computing environments and model architectures. The source code is available under a permissive license at: github.com/linkedin/Liger-Kernel.
comment: 17 pages, 12 figures
♻ ☆ Learning Linear Attention in Polynomial Time
Previous research has explored the computational expressivity of Transformer models in simulating Boolean circuits or Turing machines. However, the learnability of these simulators from observational data has remained an open question. Our study addresses this gap by providing the first polynomial-time learnability results (specifically strong, agnostic PAC learning) for single-layer Transformers with linear attention. We show that linear attention may be viewed as a linear predictor in a suitably defined RKHS. As a consequence, the problem of learning any linear transformer may be converted into the problem of learning an ordinary linear predictor in an expanded feature space, and any such predictor may be converted back into a multiheaded linear transformer. Moving to generalization, we show how to efficiently identify training datasets for which every empirical risk minimizer is equivalent (up to trivial symmetries) to the linear Transformer that generated the data, thereby guaranteeing the learned model will correctly generalize across all inputs. Finally, we provide examples of computations expressible via linear attention and therefore polynomial-time learnable, including associative memories, finite automata, and a class of Universal Turing Machine (UTMs) with polynomially bounded computation histories. We empirically validate our theoretical findings on three tasks: learning random linear attention networks, key--value associations, and learning to execute finite automata. Our findings bridge a critical gap between theoretical expressivity and learnability of Transformers, and show that flexible and general models of computation are efficiently learnable.
♻ ☆ One size doesn't fit all: Predicting the Number of Examples for In-Context Learning
In-context learning (ICL) refers to the process of adding a small number of localized examples (ones that are semantically similar to the input) from a training set of labelled data to an LLM's prompt with an objective to effectively control the generative process seeking to improve the downstream task performance. Existing ICL approaches use an identical number of examples (a pre-configured hyper-parameter) for each data instance. Our work alleviates the limitations of this 'one fits all' approach by dynamically predicting the number of examples for each data instance to be used in few-shot inference with LLMs. In particular, we employ a multi-label classifier, the parameters of which are fitted using a training set, where the label for each instance in the training set indicates if using a specific value of k (number of most similar examples from 0 up to a maximum value) leads to correct k-shot downstream predictions. Our experiments on a number of text classification benchmarks show that AICL substantially outperforms standard ICL by up to 17%.
♻ ☆ Modular Boundaries in Recurrent Neural Networks
Recent theoretical and experimental work in neuroscience has focused on the representational and dynamical character of neural manifolds --subspaces in neural activity space wherein many neurons coactivate. Importantly, neural populations studied under this "neural manifold hypothesis" are continuous and not cleanly divided into separate neural populations. This perspective clashes with the "modular hypothesis" of brain organization, wherein neural elements maintain an "all-or-nothing" affiliation with modules. In line with this modular hypothesis, recent research on recurrent neural networks suggests that multi-task networks become modular across training, such that different modules specialize for task-general dynamical motifs. If the modular hypothesis is true, then it would be important to use a dimensionality reduction technique that captures modular structure. Here, we investigate the features of such a method. We leverage RNNs as a model system to study the character of modular neural populations, using a community detection method from network science known as modularity maximization to partition neurons into distinct modules. These partitions allow us to ask the following question: do these modular boundaries matter to the system? ...
♻ ☆ TGB 2.0: A Benchmark for Learning on Temporal Knowledge Graphs and Heterogeneous Graphs NeurIPS 2024
Multi-relational temporal graphs are powerful tools for modeling real-world data, capturing the evolving and interconnected nature of entities over time. Recently, many novel models are proposed for ML on such graphs intensifying the need for robust evaluation and standardized benchmark datasets. However, the availability of such resources remains scarce and evaluation faces added complexity due to reproducibility issues in experimental protocols. To address these challenges, we introduce Temporal Graph Benchmark 2.0 (TGB 2.0), a novel benchmarking framework tailored for evaluating methods for predicting future links on Temporal Knowledge Graphs and Temporal Heterogeneous Graphs with a focus on large-scale datasets, extending the Temporal Graph Benchmark. TGB 2.0 facilitates comprehensive evaluations by presenting eight novel datasets spanning five domains with up to 53 million edges. TGB 2.0 datasets are significantly larger than existing datasets in terms of number of nodes, edges, or timestamps. In addition, TGB 2.0 provides a reproducible and realistic evaluation pipeline for multi-relational temporal graphs. Through extensive experimentation, we observe that 1) leveraging edge-type information is crucial to obtain high performance, 2) simple heuristic baselines are often competitive with more complex methods, 3) most methods fail to run on our largest datasets, highlighting the need for research on more scalable methods.
comment: 29 pages, 8 figures, 11 tables, accepted at NeurIPS 2024 Track on Datasets and Benchmarks
♻ ☆ Scalable Drift Monitoring in Medical Imaging AI
The integration of artificial intelligence (AI) into medical imaging has advanced clinical diagnostics but poses challenges in managing model drift and ensuring long-term reliability. To address these challenges, we develop MMC+, an enhanced framework for scalable drift monitoring, building upon the CheXstray framework that introduced real-time drift detection for medical imaging AI models using multi-modal data concordance. This work extends the original framework's methodologies, providing a more scalable and adaptable solution for real-world healthcare settings and offers a reliable and cost-effective alternative to continuous performance monitoring addressing limitations of both continuous and periodic monitoring methods. MMC+ introduces critical improvements to the original framework, including more robust handling of diverse data streams, improved scalability with the integration of foundation models like MedImageInsight for high-dimensional image embeddings without site-specific training, and the introduction of uncertainty bounds to better capture drift in dynamic clinical environments. Validated with real-world data from Massachusetts General Hospital during the COVID-19 pandemic, MMC+ effectively detects significant data shifts and correlates them with model performance changes. While not directly predicting performance degradation, MMC+ serves as an early warning system, indicating when AI systems may deviate from acceptable performance bounds and enabling timely interventions. By emphasizing the importance of monitoring diverse data streams and evaluating data shifts alongside model performance, this work contributes to the broader adoption and integration of AI solutions in clinical settings.
♻ ☆ Learning diffusion at lightspeed NeurIPS 2024
Diffusion regulates numerous natural processes and the dynamics of many successful generative models. Existing models to learn the diffusion terms from observational data rely on complex bilevel optimization problems and model only the drift of the system. We propose a new simple model, JKOnet*, which bypasses the complexity of existing architectures while presenting significantly enhanced representational capabilities: JKOnet* recovers the potential, interaction, and internal energy components of the underlying diffusion process. JKOnet* minimizes a simple quadratic loss and outperforms other baselines in terms of sample efficiency, computational complexity, and accuracy. Additionally, JKOnet* provides a closed-form optimal solution for linearly parametrized functionals, and, when applied to predict the evolution of cellular processes from real-world data, it achieves state-of-the-art accuracy at a fraction of the computational cost of all existing methods. Our methodology is based on the interpretation of diffusion processes as energy-minimizing trajectories in the probability space via the so-called JKO scheme, which we study via its first-order optimality conditions.
comment: Accepted for presentation at, and publication in the proceedings of, the 38th Conference on Neural Information Processing Systems (NeurIPS 2024, oral)
♻ ☆ English offensive text detection using CNN based Bi-GRU model
Over the years, the number of users of social media has increased drastically. People frequently share their thoughts through social platforms, and this leads to an increase in hate content. In this virtual community, individuals share their views, express their feelings, and post photos, videos, blogs, and more. Social networking sites like Facebook and Twitter provide platforms to share vast amounts of content with a single click. However, these platforms do not impose restrictions on the uploaded content, which may include abusive language and explicit images unsuitable for social media. To resolve this issue, a new idea must be implemented to divide the inappropriate content. Numerous studies have been done to automate the process. In this paper, we propose a new Bi-GRU-CNN model to classify whether the text is offensive or not. The combination of the Bi-GRU and CNN models outperforms the existing model.
comment: 5 pages and 6 figures
♻ ☆ Retraining with Predicted Hard Labels Provably Increases Model Accuracy
The performance of a model trained with \textit{noisy labels} is often improved by simply \textit{retraining} the model with its own predicted \textit{hard} labels (i.e., $1$/$0$ labels). Yet, a detailed theoretical characterization of this phenomenon is lacking. In this paper, we theoretically analyze retraining in a linearly separable setting with randomly corrupted labels given to us and prove that retraining can improve the population accuracy obtained by initially training with the given (noisy) labels. To the best of our knowledge, this is the first such theoretical result. Retraining finds application in improving training with local label differential privacy (DP) which involves training with noisy labels. We empirically show that retraining selectively on the samples for which the predicted label matches the given label significantly improves label DP training at \textit{no extra privacy cost}; we call this \textit{consensus-based retraining}. As an example, when training ResNet-18 on CIFAR-100 with $\epsilon=3$ label DP, we obtain $6.4\%$ improvement in accuracy with consensus-based retraining.
♻ ☆ Clustering of timed sequences -- Application to the analysis of care pathways
Improving the future of healthcare starts by better understanding the current actual practices in hospital settings. This motivates the objective of discovering typical care pathways from patient data. Revealing typical care pathways can be achieved through clustering. The difficulty in clustering care pathways, represented by sequences of timestamped events, lies in defining a semantically appropriate metric and clustering algorithms. In this article, we adapt two methods developed for time series to the clustering of timed sequences: the drop-DTW metric and the DBA approach for the construction of averaged time sequences. These methods are then applied in clustering algorithms to propose original and sound clustering algorithms for timed sequences. This approach is experimented with and evaluated on synthetic and real-world data.
♻ ☆ On Debiasing Text Embeddings Through Context Injection
Current advances in Natural Language Processing (NLP) have made it increasingly feasible to build applications leveraging textual data. Generally, the core of these applications rely on having a good semantic representation of text into vectors, via embedding models. However, it has been shown that these embeddings capture and perpetuate biases already present in text. While a few techniques have been proposed to debias embeddings, they do not take advantage of the recent advances in context understanding of modern embedding models. In this paper, we fill this gap by conducting a review of 19 embedding models by quantifying their biases and how well they respond to context injection as a mean of debiasing. We show that higher performing models are more prone to capturing biases, but are also better at incorporating context. Surprisingly, we find that while models can easily embed affirmative semantics, they fail at embedding neutral semantics. Finally, in a retrieval task, we show that biases in embeddings can lead to non-desirable outcomes. We use our new-found insights to design a simple algorithm for top $k$ retrieval, where $k$ is dynamically selected. We show that our algorithm is able to retrieve all relevant gendered and neutral chunks.
♻ ☆ Inferring Change Points in High-Dimensional Regression via Approximate Message Passing ICML 2024
We consider the problem of localizing change points in a generalized linear model (GLM), a model that covers many widely studied problems in statistical learning including linear, logistic, and rectified linear regression. We propose a novel and computationally efficient Approximate Message Passing (AMP) algorithm for estimating both the signals and the change point locations, and rigorously characterize its performance in the high-dimensional limit where the number of parameters $p$ is proportional to the number of samples $n$. This characterization is in terms of a state evolution recursion, which allows us to precisely compute performance measures such as the asymptotic Hausdorff error of our change point estimates, and allows us to tailor the algorithm to take advantage of any prior structural information on the signals and change points. Moreover, we show how our AMP iterates can be used to efficiently compute a Bayesian posterior distribution over the change point locations in the high-dimensional limit. We validate our theory via numerical experiments, and demonstrate the favorable performance of our estimators on both synthetic and real data in the settings of linear, logistic, and rectified linear regression.
comment: 43 pages, 9 figures. A preliminary version of this paper appeared in ICML 2024
♻ ☆ Kernel Density Estimators in Large Dimensions
This paper studies Kernel Density Estimation for a high-dimensional distribution $\rho(x)$. Traditional approaches have focused on the limit of large number of data points $n$ and fixed dimension $d$. We analyze instead the regime where both the number $n$ of data points $y_i$ and their dimensionality $d$ grow with a fixed ratio $\alpha=(\log n)/d$. Our study reveals three distinct statistical regimes for the kernel-based estimate of the density $\hat \rho_h^{\mathcal {D}}(x)=\frac{1}{n h^d}\sum_{i=1}^n K\left(\frac{x-y_i}{h}\right)$, depending on the bandwidth $h$: a classical regime for large bandwidth where the Central Limit Theorem (CLT) holds, which is akin to the one found in traditional approaches. Below a certain value of the bandwidth, $h_{CLT}(\alpha)$, we find that the CLT breaks down. The statistics of $\hat\rho_h^{\mathcal {D}}(x)$ for a fixed $x$ drawn from $\rho(x)$ is given by a heavy-tailed distribution (an alpha-stable distribution). In particular below a value $h_G(\alpha)$, we find that $\hat\rho_h^{\mathcal {D}}(x)$ is governed by extreme value statistics: only a few points in the database matter and give the dominant contribution to the density estimator. We provide a detailed analysis for high-dimensional multivariate Gaussian data. We show that the optimal bandwidth threshold based on Kullback-Leibler divergence lies in the new statistical regime identified in this paper. As known by practitioners, when decreasing the bandwidth a Kernel-estimated estimated changes from a smooth curve to a collections of peaks centred on the data points. Our findings reveal that this general phenomenon is related to sharp transitions between phases characterized by different statistical properties, and offer new insights for Kernel density estimation in high-dimensional settings.
♻ ☆ Deep Implicit Optimization for Robust and Flexible Image Registration
Deep Learning in Image Registration (DLIR) methods have been tremendously successful in image registration due to their speed and ability to incorporate weak label supervision at training time. However, DLIR methods forego many of the benefits of classical optimization-based methods. The functional nature of deep networks do not guarantee that the predicted transformation is a local minima of the registration objective, the representation of the transformation (displacement/velocity field/affine) is fixed, and the networks are not robust to domain shift. Our method aims to bridge this gap between classical and learning methods by incorporating optimization as a layer in a deep network. A deep network is trained to predict multi-scale dense feature images that are registered using a black box iterative optimization solver. This optimal warp is then used to minimize image and label alignment errors. By implicitly differentiating end-to-end through an iterative optimization solver, our learned features are registration and label-aware, and the warp functions are guaranteed to be local minima of the registration objective in the feature space. Our framework shows excellent performance on in-domain datasets, and is agnostic to domain shift such as anisotropy and varying intensity profiles. For the first time, our method allows switching between arbitrary transformation representations (free-form to diffeomorphic) at test time with zero retraining. End-to-end feature learning also facilitates interpretability of features, and out-of-the-box promptability using additional label-fidelity terms at inference.
♻ ☆ Overcoming Slow Decision Frequencies in Continuous Control: Model-Based Sequence Reinforcement Learning for Model-Free Control
Reinforcement learning (RL) is rapidly reaching and surpassing human-level control capabilities. However, state-of-the-art RL algorithms often require timesteps and reaction times significantly faster than human capabilities, which is impractical in real-world settings and typically necessitates specialized hardware. Such speeds are difficult to achieve in the real world and often requires specialized hardware. We introduce Sequence Reinforcement Learning (SRL), an RL algorithm designed to produce a sequence of actions for a given input state, enabling effective control at lower decision frequencies. SRL addresses the challenges of learning action sequences by employing both a model and an actor-critic architecture operating at different temporal scales. We propose a "temporal recall" mechanism, where the critic uses the model to estimate intermediate states between primitive actions, providing a learning signal for each individual action within the sequence. Once training is complete, the actor can generate action sequences independently of the model, achieving model-free control at a slower frequency. We evaluate SRL on a suite of continuous control tasks, demonstrating that it achieves performance comparable to state-of-the-art algorithms while significantly reducing actor sample complexity. To better assess performance across varying decision frequencies, we introduce the Frequency-Averaged Score (FAS) metric. Our results show that SRL significantly outperforms traditional RL algorithms in terms of FAS, making it particularly suitable for applications requiring variable decision frequencies. Additionally, we compare SRL with model-based online planning, showing that SRL achieves superior FAS while leveraging the same model during training that online planners use for planning.
♻ ☆ Sample Compression Scheme Reductions
We present novel reductions from sample compression schemes in multiclass classification, regression, and adversarially robust learning settings to binary sample compression schemes. Assuming we have a compression scheme for binary classes of size $f(d_\mathrm{VC})$, where $d_\mathrm{VC}$ is the VC dimension, then we have the following results: (1) If the binary compression scheme is a majority-vote or a stable compression scheme, then there exists a multiclass compression scheme of size $O(f(d_\mathrm{G}))$, where $d_\mathrm{G}$ is the graph dimension. Moreover, for general binary compression schemes, we obtain a compression of size $O(f(d_\mathrm{G})\log|Y|)$, where $Y$ is the label space. (2) If the binary compression scheme is a majority-vote or a stable compression scheme, then there exists an $\epsilon$-approximate compression scheme for regression over $[0,1]$-valued functions of size $O(f(d_\mathrm{P}))$, where $d_\mathrm{P}$ is the pseudo-dimension. For general binary compression schemes, we obtain a compression of size $O(f(d_\mathrm{P})\log(1/\epsilon))$. These results would have significant implications if the sample compression conjecture, which posits that any binary concept class with a finite VC dimension admits a binary compression scheme of size $O(d_\mathrm{VC})$, is resolved (Littlestone and Warmuth, 1986; Floyd and Warmuth, 1995; Warmuth, 2003). Our results would then extend the proof of the conjecture immediately to other settings. We establish similar results for adversarially robust learning and also provide an example of a concept class that is robustly learnable but has no bounded-size compression scheme, demonstrating that learnability is not equivalent to having a compression scheme independent of the sample size, unlike in binary classification, where compression of size $2^{O(d_\mathrm{VC})}$ is attainable (Moran and Yehudayoff, 2016).
♻ ☆ BlackDAN: A Black-Box Multi-Objective Approach for Effective and Contextual Jailbreaking of Large Language Models
While large language models (LLMs) exhibit remarkable capabilities across various tasks, they encounter potential security risks such as jailbreak attacks, which exploit vulnerabilities to bypass security measures and generate harmful outputs. Existing jailbreak strategies mainly focus on maximizing attack success rate (ASR), frequently neglecting other critical factors, including the relevance of the jailbreak response to the query and the level of stealthiness. This narrow focus on single objectives can result in ineffective attacks that either lack contextual relevance or are easily recognizable. In this work, we introduce BlackDAN, an innovative black-box attack framework with multi-objective optimization, aiming to generate high-quality prompts that effectively facilitate jailbreaking while maintaining contextual relevance and minimizing detectability. BlackDAN leverages Multiobjective Evolutionary Algorithms (MOEAs), specifically the NSGA-II algorithm, to optimize jailbreaks across multiple objectives including ASR, stealthiness, and semantic relevance. By integrating mechanisms like mutation, crossover, and Pareto-dominance, BlackDAN provides a transparent and interpretable process for generating jailbreaks. Furthermore, the framework allows customization based on user preferences, enabling the selection of prompts that balance harmfulness, relevance, and other factors. Experimental results demonstrate that BlackDAN outperforms traditional single-objective methods, yielding higher success rates and improved robustness across various LLMs and multimodal LLMs, while ensuring jailbreak responses are both relevant and less detectable.
♻ ☆ IncidentResponseGPT: Generating Traffic Incident Response Plans with Generative Artificial Intelligence
The proposed IncidentResponseGPT framework - a novel system that applies generative artificial intelligence (AI) to potentially enhance the efficiency and effectiveness of traffic incident response. This model allows for synthesis of region-specific incident response guidelines and generates incident response plans adapted to specific area, aiming to expedite decision-making for traffic management authorities. This approach aims to accelerate incident resolution times by suggesting various recommendations (e.g. optimal rerouting strategies, estimating resource needs) to minimize the overall impact on the urban traffic network. The system suggests specific actions, including dynamic lane closures, optimized rerouting and dispatching appropriate emergency resources. We utilize the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) to rank generated response plans based on criteria like impact minimization and resource efficiency based on their proximity to an human-proposed solution.
♻ ☆ Model Internals-based Answer Attribution for Trustworthy Retrieval-Augmented Generation EMNLP 2024
Ensuring the verifiability of model answers is a fundamental challenge for retrieval-augmented generation (RAG) in the question answering (QA) domain. Recently, self-citation prompting was proposed to make large language models (LLMs) generate citations to supporting documents along with their answers. However, self-citing LLMs often struggle to match the required format, refer to non-existent sources, and fail to faithfully reflect LLMs' context usage throughout the generation. In this work, we present MIRAGE --Model Internals-based RAG Explanations -- a plug-and-play approach using model internals for faithful answer attribution in RAG applications. MIRAGE detects context-sensitive answer tokens and pairs them with retrieved documents contributing to their prediction via saliency methods. We evaluate our proposed approach on a multilingual extractive QA dataset, finding high agreement with human answer attribution. On open-ended QA, MIRAGE achieves citation quality and efficiency comparable to self-citation while also allowing for a finer-grained control of attribution parameters. Our qualitative evaluation highlights the faithfulness of MIRAGE's attributions and underscores the promising application of model internals for RAG answer attribution.
comment: Accepted by EMNLP 2024 Main Conference. Code and data released at https://github.com/Betswish/MIRAGE
♻ ☆ Hip Fracture Patient Pathways and Agent-based Modelling
Increased healthcare demand is significantly straining European services. Digital solutions including advanced modelling techniques offer a promising solution to optimising patient flow without impacting day-to-day healthcare provision. In this work we outline an ongoing project that aims to optimise healthcare resources using agent-based simulations.
comment: 6 pages, 2 figures
♻ ☆ Are High-Degree Representations Really Unnecessary in Equivariant Graph Neural Networks?
Equivariant Graph Neural Networks (GNNs) that incorporate E(3) symmetry have achieved significant success in various scientific applications. As one of the most successful models, EGNN leverages a simple scalarization technique to perform equivariant message passing over only Cartesian vectors (i.e., 1st-degree steerable vectors), enjoying greater efficiency and efficacy compared to equivariant GNNs using higher-degree steerable vectors. This success suggests that higher-degree representations might be unnecessary. In this paper, we disprove this hypothesis by exploring the expressivity of equivariant GNNs on symmetric structures, including $k$-fold rotations and regular polyhedra. We theoretically demonstrate that equivariant GNNs will always degenerate to a zero function if the degree of the output representations is fixed to 1 or other specific values. Based on this theoretical insight, we propose HEGNN, a high-degree version of EGNN to increase the expressivity by incorporating high-degree steerable vectors while maintaining EGNN's efficiency through the scalarization trick. Our extensive experiments demonstrate that HEGNN not only aligns with our theoretical analyses on toy datasets consisting of symmetric structures, but also shows substantial improvements on more complicated datasets such as $N$-body and MD17. Our theoretical findings and empirical results potentially open up new possibilities for the research of equivariant GNNs.
♻ ☆ A Novel Cartography-Based Curriculum Learning Method Applied on RoNLI: The First Romanian Natural Language Inference Corpus ACL 2024
Natural language inference (NLI), the task of recognizing the entailment relationship in sentence pairs, is an actively studied topic serving as a proxy for natural language understanding. Despite the relevance of the task in building conversational agents and improving text classification, machine translation and other NLP tasks, to the best of our knowledge, there is no publicly available NLI corpus for the Romanian language. To this end, we introduce the first Romanian NLI corpus (RoNLI) comprising 58K training sentence pairs, which are obtained via distant supervision, and 6K validation and test sentence pairs, which are manually annotated with the correct labels. We conduct experiments with multiple machine learning methods based on distant learning, ranging from shallow models based on word embeddings to transformer-based neural networks, to establish a set of competitive baselines. Furthermore, we improve on the best model by employing a new curriculum learning strategy based on data cartography. Our dataset and code to reproduce the baselines are available at https://github.com/Eduard6421/RONLI.
comment: Accepted at ACL 2024 (Main)
♻ ☆ Plug-and-Play Posterior Sampling under Mismatched Measurement and Prior Models
Posterior sampling has been shown to be a powerful Bayesian approach for solving imaging inverse problems. The recent plug-and-play unadjusted Langevin algorithm (PnP-ULA) has emerged as a promising method for Monte Carlo sampling and minimum mean squared error (MMSE) estimation by combining physical measurement models with deep-learning priors specified using image denoisers. However, the intricate relationship between the sampling distribution of PnP-ULA and the mismatched data-fidelity and denoiser has not been theoretically analyzed. We address this gap by proposing a posterior-L2 pseudometric and using it to quantify an explicit error bound for PnP-ULA under mismatched posterior distribution. We numerically validate our theory on several inverse problems such as sampling from Gaussian mixture models and image deblurring. Our results suggest that the sensitivity of the sampling distribution of PnP-ULA to a mismatch in the measurement model and the denoiser can be precisely characterized.
♻ ☆ Multi-LLM QA with Embodied Exploration
Large language models (LLMs) have grown in popularity due to their natural language interface and pre trained knowledge, leading to rapidly increasing success in question-answering (QA) tasks. More recently, multi-agent systems with LLM-based agents (Multi-LLM) have been utilized increasingly more for QA. In these scenarios, the models may each answer the question and reach a consensus or each model is specialized to answer different domain questions. However, most prior work dealing with Multi-LLM QA has focused on scenarios where the models are asked in a zero-shot manner or are given information sources to extract the answer. For question answering of an unknown environment, embodied exploration of the environment is first needed to answer the question. This skill is necessary for personalizing embodied AI to environments such as households. There is a lack of insight into whether a Multi-LLM system can handle question-answering based on observations from embodied exploration. In this work, we address this gap by investigating the use of Multi-Embodied LLM Explorers (MELE) for QA in an unknown environment. Multiple LLM-based agents independently explore and then answer queries about a household environment. We analyze different aggregation methods to generate a single, final answer for each query: debating, majority voting, and training a central answer module (CAM). Using CAM, we observe a $46\%$ higher accuracy compared against the other non-learning-based aggregation methods. We provide code and the query dataset for further research.
comment: 16 pages, 9 Figures, 5 Tables
♻ ☆ Learning Social Cost Functions for Human-Aware Path Planning
Achieving social acceptance is one of the main goals of Social Robotic Navigation. Despite this topic has received increasing interest in recent years, most of the research has focused on driving the robotic agent along obstacle-free trajectories, planning around estimates of future human motion to respect personal distances and optimize navigation. However, social interactions in everyday life are also dictated by norms that do not strictly depend on movement, such as when standing at the end of a queue rather than cutting it. In this paper, we propose a novel method to recognize common social scenarios and modify a traditional planner's cost function to adapt to them. This solution enables the robot to carry out different social navigation behaviors that would not arise otherwise, maintaining the robustness of traditional navigation. Our approach allows the robot to learn different social norms with a single learned model, rather than having different modules for each task. As a proof of concept, we consider the tasks of queuing and respect interaction spaces of groups of people talking to one another, but the method can be extended to other human activities that do not involve motion.
♻ ☆ An algorithm for clustering with confidence-based must-link and cannot-link constraints
We study here the semi-supervised $k$-clustering problem where information is available on whether pairs of objects are in the same or in different clusters. This information is either available with certainty or with a limited level of confidence. We introduce the PCCC (Pairwise-Confidence-Constraints-Clustering) algorithm, which iteratively assigns objects to clusters while accounting for the information provided on the pairs of objects. Our algorithm uses integer programming for the assignment of objects which allows to include relationships as hard constraints that are guaranteed to be satisfied or as soft constraints that can be violated subject to a penalty. This flexibility distinguishes our algorithm from the state-of-the-art in which all pairwise constraints are either considered hard, or all are considered soft. We developed an enhanced multi-start approach and a model-size reduction technique for the integer program that contributes to the effectiveness and the efficiency of the algorithm. Unlike existing algorithms, our algorithm scales to large-scale instances with up to 60,000 objects, 100 clusters, and millions of cannot-link constraints (which are the most challenging constraints to incorporate). We compare the PCCC algorithm with state-of-the-art approaches in an extensive computational study. Even though the PCCC algorithm is more general than the state-of-the-art approaches in its applicability, it outperforms the state-of-the-art approaches on instances with all hard or all soft constraints both in terms of runtime and various metrics of solution quality. The code of the PCCC algorithm is publicly available on GitHub.
comment: To appear in INFORMS Journal on Computing
♻ ☆ Explaining Modern Gated-Linear RNNs via a Unified Implicit Attention Formulation
Recent advances in efficient sequence modeling have led to attention-free layers, such as Mamba, RWKV, and various gated RNNs, all featuring sub-quadratic complexity in sequence length and excellent scaling properties, enabling the construction of a new type of foundation models. In this paper, we present a unified view of these models, formulating such layers as implicit causal self-attention layers. The formulation includes most of their sub-components and is not limited to a specific part of the architecture. The framework compares the underlying mechanisms on similar grounds for different layers and provides a direct means for applying explainability methods. Our experiments show that our attention matrices and attribution method outperform an alternative and a more limited formulation that was recently proposed for Mamba. For the other architectures for which our method is the first to provide such a view, our method is effective and competitive in the relevant metrics compared to the results obtained by state-of-the-art Transformer explainability methods. Our code is publicly available.
♻ ☆ MolecularGPT: Open Large Language Model (LLM) for Few-Shot Molecular Property Prediction
Molecular property prediction (MPP) is a fundamental and crucial task in drug discovery. However, prior methods are limited by the requirement for a large number of labeled molecules and their restricted ability to generalize for unseen and new tasks, both of which are essential for real-world applications. To address these challenges, we present MolecularGPT for few-shot MPP. From a perspective on instruction tuning, we fine-tune large language models (LLMs) based on curated molecular instructions spanning over 1000 property prediction tasks. This enables building a versatile and specialized LLM that can be adapted to novel MPP tasks without any fine-tuning through zero- and few-shot in-context learning (ICL). MolecularGPT exhibits competitive in-context reasoning capabilities across 10 downstream evaluation datasets, setting new benchmarks for few-shot molecular prediction tasks. More importantly, with just two-shot examples, MolecularGPT can outperform standard supervised graph neural network methods on 4 out of 7 datasets. It also excels state-of-the-art LLM baselines by up to 15.7% increase on classification accuracy and decrease of 17.9 on regression metrics (e.g., RMSE) under zero-shot. This study demonstrates the potential of LLMs as effective few-shot molecular property predictors. The code is available at https://github.com/NYUSHCS/MolecularGPT.
♻ ☆ Timeseria: an object-oriented time series processing library
Timeseria is an object-oriented time series processing library implemented in Python, which aims at making it easier to manipulate time series data and to build statistical and machine learning models on top of it. Unlike common data analysis frameworks, it builds up from well defined and reusable logical units (objects), which can be easily combined together in order to ensure a high level of consistency. Thanks to this approach, Timeseria can address by design several non-trivial issues often underestimated, such as handling data losses, non-uniform sampling rates, differences between aggregated data and punctual observations, time zones, daylight saving times, and more. Timeseria comes with a comprehensive set of base data structures, common data manipulation operations, and extensible models for data reconstruction, forecasting and anomaly detection. It also integrates a powerful plotting engine capable of handling even millions of data points.
♻ ☆ Spectral and Rhythm Features for Audio Classification with Deep Convolutional Neural Networks
Convolutional neural networks (CNNs) are widely used in computer vision. They can be used not only for conventional digital image material to recognize patterns, but also for feature extraction from digital imagery representing spectral and rhythm features extracted from time-domain digital audio signals for the acoustic classification of sounds. Different spectral and rhythm feature representations like mel-scaled spectrograms, mel-frequency cepstral coefficients (MFCCs), cyclic tempograms, short-time Fourier transform (STFT) chromagrams, constant-Q transform (CQT) chromagrams and chroma energy normalized statistics (CENS) chromagrams are investigated in terms of the audio classification performance using a deep convolutional neural network. It can be clearly shown that the mel-scaled spectrograms and the mel-frequency cepstral coefficients (MFCCs) perform significantly better than the other spectral and rhythm features investigated in this research for audio classification tasks using deep CNNs. The experiments were carried out with the aid of the ESC-50 dataset with 2,000 labeled environmental audio recordings.
♻ ☆ Distributionally and Adversarially Robust Logistic Regression via Intersecting Wasserstein Balls
Adversarially robust optimization (ARO) has become the de facto standard for training models to defend against adversarial attacks during testing. However, despite their robustness, these models often suffer from severe overfitting. To mitigate this issue, several successful approaches have been proposed, including replacing the empirical distribution in training with: (i) a worst-case distribution within an ambiguity set, leading to a distributionally robust (DR) counterpart of ARO; or (ii) a mixture of the empirical distribution with one derived from an auxiliary dataset (e.g., synthetic, external, or out-of-domain). Building on the first approach, we explore the Wasserstein DR counterpart of ARO for logistic regression and show it admits a tractable convex optimization reformulation. Adopting the second approach, we enhance the DR framework by intersecting its ambiguity set with one constructed from an auxiliary dataset, which yields significant improvements when the Wasserstein distance between the data-generating and auxiliary distributions can be estimated. We analyze the resulting optimization problem, develop efficient solutions, and show that our method outperforms benchmark approaches on standard datasets.
comment: 33 pages, 3 color figures, under review at a conference
♻ ☆ Predicting Accurate Lagrangian Multipliers for Mixed Integer Linear Programs
Lagrangian relaxation stands among the most efficient approaches for solving a Mixed Integer Linear Programs (MILP) with difficult constraints. Given any duals for these constraints, called Lagrangian Multipliers (LMs), it returns a bound on the optimal value of the MILP, and Lagrangian methods seek the LMs giving the best such bound. But these methods generally rely on iterative algorithms resembling gradient descent to maximize the concave piecewise linear dual function: the computational burden grows quickly with the number of relaxed constraints. We introduce a deep learning approach that bypasses the descent, effectively amortizing the local, per instance, optimization. A probabilistic encoder based on a graph convolutional network computes high-dimensional representations of relaxed constraints in MILP instances. A decoder then turns these representations into LMs. We train the encoder and decoder jointly by directly optimizing the bound obtained from the predicted multipliers. Numerical experiments show that our approach closes up to 85~\% of the gap between the continuous relaxation and the best Lagrangian bound, and provides a high quality warm-start for descent based Lagrangian methods.
♻ ☆ 3-D Magnetotelluric Deep Learning Inversion Guided by Pseudo-Physical Information
Magnetotelluric deep learning (DL) inversion methods based on joint data-driven and physics-driven have become a hot topic in recent years. When mapping observation data (or forward modeling data) to the resistivity model using neural networks (NNs), incorporating the error (loss) term of the inversion resistivity's forward modeling response--which introduces physical information about electromagnetic field propagation--can significantly enhance the inversion accuracy. To efficiently achieve data-physical dual-driven MT deep learning inversion for large-scale 3-D MT data, we propose using DL forward modeling networks to compute this portion of the loss. This approach introduces pseudo-physical information through the forward modeling of NN simulation, further guiding the inversion network fitting. Specifically, we first pre-train the forward modeling networks as fixed forward modeling operators, then transfer and integrate them into the inversion network training, and finally optimize the inversion network by minimizing the multinomial loss. Theoretical experimental results indicate that despite some simulation errors in DL forward modeling, the introduced pseudo-physical information still enhances inversion accuracy and significantly mitigates the overfitting problem during training. Additionally, we propose a new input mode that involves masking and adding noise to the data, simulating the field data environment of 3-D MT inversion, thereby making the method more flexible and effective for practical applications.
♻ ☆ A Survey of Mamba
As one of the most representative DL techniques, Transformer architecture has empowered numerous advanced models, especially the large language models (LLMs) that comprise billions of parameters, becoming a cornerstone in deep learning. Despite the impressive achievements, Transformers still face inherent limitations, particularly the time-consuming inference resulting from the quadratic computation complexity of attention calculation. Recently, a novel architecture named Mamba, drawing inspiration from classical state space models (SSMs), has emerged as a promising alternative for building foundation models, delivering comparable modeling abilities to Transformers while preserving near-linear scalability concerning sequence length. This has sparked an increasing number of studies actively exploring Mamba's potential to achieve impressive performance across diverse domains. Given such rapid evolution, there is a critical need for a systematic review that consolidates existing Mamba-empowered models, offering a comprehensive understanding of this emerging model architecture. In this survey, we therefore conduct an in-depth investigation of recent Mamba-associated studies, covering three main aspects: the advancements of Mamba-based models, the techniques of adapting Mamba to diverse data, and the applications where Mamba can excel. Specifically, we first review the foundational knowledge of various representative deep learning models and the details of Mamba-1&2 as preliminaries. Then, to showcase the significance of Mamba for AI, we comprehensively review the related studies focusing on Mamba models' architecture design, data adaptability, and applications. Finally, we present a discussion of current limitations and explore various promising research directions to provide deeper insights for future investigations.
♻ ☆ Optimization Dynamics of Equivariant and Augmented Neural Networks
We investigate the optimization of neural networks on symmetric data, and compare the strategy of constraining the architecture to be equivariant to that of using data augmentation. Our analysis reveals that that the relative geometry of the admissible and the equivariant layers, respectively, plays a key role. Under natural assumptions on the data, network, loss, and group of symmetries, we show that compatibility of the spaces of admissible layers and equivariant layers, in the sense that the corresponding orthogonal projections commute, implies that the sets of equivariant stationary points are identical for the two strategies. If the linear layers of the network also are given a unitary parametrization, the set of equivariant layers is even invariant under the gradient flow for augmented models. Our analysis however also reveals that even in the latter situation, stationary points may be unstable for augmented training although they are stable for the manifestly equivariant models.
comment: v4: Some discussions added, along with an updated experiment section. v3: Completely revised manuscript: New framework for neural nets, new main result (involving compability condition), new experiments, new author. v2: Revised manuscript. Mostly small edits, apart from new experiments (see Appendix E)
♻ ☆ Entity Matching using Large Language Models EDBT
Entity matching is the task of deciding whether two entity descriptions refer to the same real-world entity. Entity matching is a central step in most data integration pipelines. Many state-of-the-art entity matching methods rely on pre-trained language models (PLMs) such as BERT or RoBERTa. Two major drawbacks of these models for entity matching are that (i) the models require significant amounts of task-specific training data and (ii) the fine-tuned models are not robust concerning out-of-distribution entities. This paper investigates using generative large language models (LLMs) as a less task-specific training data-dependent and more robust alternative to PLM-based matchers. The study covers hosted and open-source LLMs which can be run locally. We evaluate these models in a zero-shot scenario and a scenario where task-specific training data is available. We compare different prompt designs and the prompt sensitivity of the models. We show that there is no single best prompt but that the prompt needs to be tuned for each model/dataset combination. We further investigate (i) the selection of in-context demonstrations, (ii) the generation of matching rules, as well as (iii) fine-tuning LLMs using the same pool of training data. Our experiments show that the best LLMs require no or only a few training examples to perform comparably to PLMs that were fine-tuned using thousands of examples. LLM-based matchers further exhibit higher robustness to unseen entities. We show that GPT4 can generate structured explanations for matching decisions and can automatically identify potential causes of matching errors by analyzing explanations of wrong decisions. We demonstrate that the model can generate meaningful textual descriptions of the identified error classes, which can help data engineers to improve entity matching pipelines.
comment: Published in Proceedings of the 28th International Conference on Extending Database Technology (EDBT), 25th March-28th March, 2025, ISBN 978-3-89318-098-1 on OpenProceedings.org
♻ ☆ A Survey of Multi-Agent Deep Reinforcement Learning with Communication
Communication is an effective mechanism for coordinating the behaviors of multiple agents, broadening their views of the environment, and to support their collaborations. In the field of multi-agent deep reinforcement learning (MADRL), agents can improve the overall learning performance and achieve their objectives by communication. Agents can communicate various types of messages, either to all agents or to specific agent groups, or conditioned on specific constraints. With the growing body of research work in MADRL with communication (Comm-MADRL), there is a lack of a systematic and structural approach to distinguish and classify existing Comm-MADRL approaches. In this paper, we survey recent works in the Comm-MADRL field and consider various aspects of communication that can play a role in designing and developing multi-agent reinforcement learning systems. With these aspects in mind, we propose 9 dimensions along which Comm-MADRL approaches can be analyzed, developed, and compared. By projecting existing works into the multi-dimensional space, we discover interesting trends. We also propose some novel directions for designing future Comm-MADRL systems through exploring possible combinations of the dimensions.
comment: 34 pages, 5 figures, 13 tables; published on Autonomous Agents and Multi-Agent Systems
♻ ☆ Understanding Likelihood Over-optimisation in Direct Alignment Algorithms
Direct Alignment Algorithms (DAAs), such as Direct Preference Optimisation (DPO) and Identity Preference Optimisation (IPO), have emerged as alternatives to online Reinforcement Learning from Human Feedback (RLHF) algorithms such as Proximal Policy Optimisation (PPO) for aligning language models to human preferences, without the need for explicit reward modelling. These methods generally aim to increase the likelihood of generating better (preferred) completions while discouraging worse (non-preferred) ones, while staying close to the original model's behaviour. In this work, we explore the relationship between completion likelihood and model performance in state-of-the-art DAAs, and identify a critical issue of likelihood over-optimisation. Contrary to expectations, we find that higher likelihood of better completions and larger margins between better and worse completion likelihoods do not necessarily lead to better performance, and may even degrade it. Our analysis reveals that while higher likelihood correlates with better memorisation of factual knowledge patterns, a slightly lower completion likelihood tends to improve output diversity, thus leading to better generalisation to unseen scenarios. Moreover, we identify two key indicators that signal when over-optimised output diversity begins to harm performance: Decreasing Entropy over Top-k Tokens and Diminishing Top-k Probability Mass. Our experimental results validate that these indicators are reliable signs of declining performance under different regularisations, helping prevent over-optimisation and improve alignment with human preferences.
comment: Preprint Version
♻ ☆ GLANCE: Global Actions in a Nutshell for Counterfactual Explainability
The widespread deployment of machine learning systems in critical real-world decision-making applications has highlighted the urgent need for counterfactual explainability methods that operate effectively. Global counterfactual explanations, expressed as actions to offer recourse, aim to provide succinct explanations and insights applicable to large population subgroups. Effectiveness is measured by the fraction of the population that is provided recourse, ensuring that the actions benefit as many individuals as possible. Keeping the cost of actions low ensures the proposed recourse actions remain practical and actionable. Limiting the number of actions that provide global counterfactuals is essential to maximize interpretability. The primary challenge, therefore, is balancing these trade-offs, i.e., maximizing effectiveness, minimizing cost, while maintaining a small number of actions. We introduce GLANCE, a versatile and adaptive framework, comprising two algorithms, that allows the careful balancing of the trade-offs among the three key objectives, with the size objective functioning as a tunable parameter to keep the actions few and easy to interpret. C-GLANCE employs a clustering approach that considers both the feature space and the space of counterfactual actions, thereby accounting for the distribution of points in a way that aligns with the structure of the model. T-GLANCE provides additional features to enhance flexibility. It employs a tree-based approach, that allows users to specify split features, to build a decision tree with a single counterfactual action at each node that can be used as a subgroup policy. Our extensive experimental evaluation demonstrates that our method consistently shows greater robustness and performance compared to existing methods across various datasets and models.
♻ ☆ Dating ancient manuscripts using radiocarbon and AI-based writing style analysis
Determining the chronology of ancient handwritten manuscripts is essential for reconstructing the evolution of ideas. For the Dead Sea Scrolls, this is particularly important. However, there is an almost complete lack of date-bearing manuscripts evenly distributed across the timeline and written in similar scripts available for palaeographic comparison. Here, we present Enoch, a state-of-the-art AI-based date-prediction model, trained on the basis of new radiocarbon-dated samples of the scrolls. Enoch uses established handwriting-style descriptors and applies Bayesian ridge regression. The challenge of this study is that the number of radiocarbon-dated manuscripts is small, while current machine learning requires an abundance of training data. We show that by using combined angular and allographic writing style feature vectors and applying Bayesian ridge regression, Enoch could predict the radiocarbon-based dates from style, supported by leave-one-out validation, with varied MAEs of 27.9 to 30.7 years relative to the radiocarbon dating. Enoch was then used to estimate the dates of 135 unseen manuscripts, revealing that 79 per cent of the samples were considered 'realistic' upon palaeographic post-hoc evaluation. We present a new chronology of the scrolls. The radiocarbon ranges and Enoch's style-based predictions are often older than the traditionally assumed palaeographic estimates. In the range of 300-50 BCE, Enoch's date prediction provides an improved granularity. The study is in line with current developments in multimodal machine-learning techniques, and the methods can be used for date prediction in other partially-dated manuscript collections. This research shows how Enoch's quantitative, probability-based approach can be a tool for palaeographers and historians, re-dating ancient Jewish key texts and contributing to current debates on Jewish and Christian origins.
comment: 16 pages of main article, 103 pages of supplementary materials; the first version of this article is originally prepared in July 2023 after the completion of all the experiments
♻ ☆ MixEval-X: Any-to-Any Evaluations from Real-World Data Mixtures
Perceiving and generating diverse modalities are crucial for AI models to effectively learn from and engage with real-world signals, necessitating reliable evaluations for their development. We identify two major issues in current evaluations: (1) inconsistent standards, shaped by different communities with varying protocols and maturity levels; and (2) significant query, grading, and generalization biases. To address these, we introduce MixEval-X, the first any-to-any, real-world benchmark designed to optimize and standardize evaluations across diverse input and output modalities. We propose multi-modal benchmark mixture and adaptation-rectification pipelines to reconstruct real-world task distributions, ensuring evaluations generalize effectively to real-world use cases. Extensive meta-evaluations show our approach effectively aligns benchmark samples with real-world task distributions. Meanwhile, MixEval-X's model rankings correlate strongly with that of crowd-sourced real-world evaluations (up to 0.98) while being much more efficient. We provide comprehensive leaderboards to rerank existing models and organizations and offer insights to enhance understanding of multi-modal evaluations and inform future research.
♻ ☆ FLEdge: Benchmarking Federated Machine Learning Applications in Edge Computing Systems
Federated Learning (FL) has become a viable technique for realizing privacy-enhancing distributed deep learning on the network edge. Heterogeneous hardware, unreliable client devices, and energy constraints often characterize edge computing systems. In this paper, we propose FLEdge, which complements existing FL benchmarks by enabling a systematic evaluation of client capabilities. We focus on computational and communication bottlenecks, client behavior, and data security implications. Our experiments with models varying from 14K to 80M trainable parameters are carried out on dedicated hardware with emulated network characteristics and client behavior. We find that state-of-the-art embedded hardware has significant memory bottlenecks, leading to 4x longer processing times than on modern data center GPUs.
comment: Paper accepted for publication at the ACM/IFIP Middleware Conference 2024. Please cite the published version via https://doi.org/10.1145/3652892.3700751
♻ ☆ Context-Enhanced Multi-View Trajectory Representation Learning: Bridging the Gap through Self-Supervised Models
Modeling trajectory data with generic-purpose dense representations has become a prevalent paradigm for various downstream applications, such as trajectory classification, travel time estimation and similarity computation. However, existing methods typically rely on trajectories from a single spatial view, limiting their ability to capture the rich contextual information that is crucial for gaining deeper insights into movement patterns across different geospatial contexts. To this end, we propose MVTraj, a novel multi-view modeling method for trajectory representation learning. MVTraj integrates diverse contextual knowledge, from GPS to road network and points-of-interest to provide a more comprehensive understanding of trajectory data. To align the learning process across multiple views, we utilize GPS trajectories as a bridge and employ self-supervised pretext tasks to capture and distinguish movement patterns across different spatial views. Following this, we treat trajectories from different views as distinct modalities and apply a hierarchical cross-modal interaction module to fuse the representations, thereby enriching the knowledge derived from multiple sources. Extensive experiments on real-world datasets demonstrate that MVTraj significantly outperforms existing baselines in tasks associated with various spatial views, validating its effectiveness and practical utility in spatio-temporal modeling.
♻ ☆ The Model Openness Framework: Promoting Completeness and Openness for Reproducibility, Transparency, and Usability in Artificial Intelligence
Generative artificial intelligence (AI) offers numerous opportunities for research and innovation, but its commercialization has raised concerns about the transparency and safety of frontier AI models. Most models lack the necessary components for full understanding, auditing, and reproducibility, and some model producers use restrictive licenses whilst claiming that their models are "open source". To address these concerns, we introduce the Model Openness Framework (MOF), a three-tiered ranked classification system that rates machine learning models based on their completeness and openness, following open science principles. For each MOF class, we specify code, data, and documentation components of the model development lifecycle that must be released and under which open licenses. In addition, the Model Openness Tool (MOT) provides a user-friendly reference implementation to evaluate the openness and completeness of models against the MOF classification system. Together, the MOF and MOT provide timely practical guidance for (i) model producers to enhance the openness and completeness of their publicly-released models, and (ii) model consumers to identify open models and their constituent components that can be permissively used, studied, modified, and redistributed. Through the MOF, we seek to establish completeness and openness as core tenets of responsible AI research and development, and to promote best practices in the burgeoning open AI ecosystem.
comment: 28 pages, 4 figures, 2 tables
♻ ☆ TotalVibeSegmentator: Full Body MRI Segmentation for the NAKO and UK Biobank
Objectives: To present a publicly available torso segmentation network for large epidemiology datasets on volumetric interpolated breath-hold examination (VIBE) images. Materials & Methods: We extracted preliminary segmentations from TotalSegmentator, spine, and body composition networks for VIBE images, then improved them iteratively and retrained a nnUNet network. Using subsets of NAKO (85 subjects) and UK Biobank (16 subjects), we evaluated with Dice-score on a holdout set (12 subjects) and existing organ segmentation approach (1000 subjects), generating 71 semantic segmentation types for VIBE images. We provide an additional network for the vertebra segments 22 individual vertebra types. Results: We achieved an average Dice score of 0.89 +- 0.07 overall 71 segmentation labels. We scored > 0.90 Dice-score on the abdominal organs except for the pancreas with a Dice of 0.70. Conclusion: Our work offers a detailed and refined publicly available full torso segmentation on VIBE images.
comment: https://github.com/robert-graf/TotalVibeSegmentator
♻ ☆ Boosting Graph Pooling with Persistent Homology NeurIPS 2024
Recently, there has been an emerging trend to integrate persistent homology (PH) into graph neural networks (GNNs) to enrich expressive power. However, naively plugging PH features into GNN layers always results in marginal improvement with low interpretability. In this paper, we investigate a novel mechanism for injecting global topological invariance into pooling layers using PH, motivated by the observation that filtration operation in PH naturally aligns graph pooling in a cut-off manner. In this fashion, message passing in the coarsened graph acts along persistent pooled topology, leading to improved performance. Experimentally, we apply our mechanism to a collection of graph pooling methods and observe consistent and substantial performance gain over several popular datasets, demonstrating its wide applicability and flexibility.
comment: Published at NeurIPS 2024
♻ ☆ Communication-Efficient Distributed Deep Learning via Federated Dynamic Averaging EDBT 2025
Driven by the ever-growing volume and decentralized nature of data, coupled with the need to harness this data and generate knowledge from it, has led to the extensive use of distributed deep learning (DDL) techniques for training. These techniques rely on local training that is performed at the distributed nodes based on locally collected data, followed by a periodic synchronization process that combines these models to create a global model. However, frequent synchronization of DL models, encompassing millions to many billions of parameters, creates a communication bottleneck, severely hindering scalability. Worse yet, DDL algorithms typically waste valuable bandwidth, and make themselves less practical in bandwidth-constrained federated settings, by relying on overly simplistic, periodic, and rigid synchronization schedules. These drawbacks also have a direct impact on the time required for the training process, necessitating excessive time for data communication. To address these shortcomings, we propose Federated Dynamic Averaging (FDA), a communication-efficient DDL strategy that dynamically triggers synchronization based on the value of the model variance. In essence, the costly synchronization step is triggered only if the local models, which are initialized from a common global model after each synchronization, have significantly diverged. This decision is facilitated by the communication of a small local state from each distributed node/worker. Through extensive experiments across a wide range of learning tasks we demonstrate that FDA reduces communication cost by orders of magnitude, compared to both traditional and cutting-edge communication-efficient algorithms. Additionally, we show that FDA maintains robust performance across diverse data heterogeneity settings.
comment: Accepted as research paper at EDBT 2025
♻ ☆ FedECA: A Federated External Control Arm Method for Causal Inference with Time-To-Event Data in Distributed Settings
External control arms (ECA) can inform the early clinical development of experimental drugs and provide efficacy evidence for regulatory approval. However, the main challenge in implementing ECA lies in accessing real-world or historical clinical trials data. Indeed, regulations protecting patients' rights by strictly controlling data processing make pooling data from multiple sources in a central server often difficult. To address these limitations, we develop a new method, 'FedECA' that leverages federated learning (FL) to enable inverse probability of treatment weighting (IPTW) for time-to-event outcomes on separate cohorts without needing to pool data. To showcase the potential of FedECA, we apply it in different settings of increasing complexity culminating with a real-world use-case in which FedECA provides evidence for a differential effect between two drugs that would have otherwise gone unnoticed. By sharing our code, we hope FedECA will foster the creation of federated research networks and thus accelerate drug development.
comment: code available at: https://github.com/owkin/fedeca, bug in SMD computation present in v1 and v2 has been fixed, many experiments on real data have been added + fix in YODA experiments using imputed data instead of raw data as well as typos and affiliations fix
♻ ☆ Simple Opinion Dynamics for No-Regret Learning
We study a cooperative multi-agent bandit setting in the distributed GOSSIP model: in every round, each of $n$ agents chooses an action from a common set, observes the action's corresponding reward, and subsequently exchanges information with a single randomly chosen neighbor, which may inform its choice in the next round. We introduce and analyze families of memoryless and time-independent protocols for this setting, inspired by opinion dynamics that are well-studied for other algorithmic tasks in the GOSSIP model. For stationary reward settings, we prove for the first time that these simple protocols exhibit best-of-both-worlds behavior, simultaneously obtaining constant cumulative regret scaling like $R(T)/T = \widetilde O(1/T)$, and also reaching consensus on the highest-mean action within $\widetilde O(\sqrt{n})$ rounds. We obtain these results by showing a new connection between the global evolution of these decentralized protocols and a class of zero-sum multiplicative weights update} processes. Using this connection, we establish a general framework for analyzing the population-level regret and other properties of our protocols. Finally, we show our protocols are also surprisingly robust to adversarial rewards, and in this regime we obtain sublinear regret scaling like $R(T)/T = \widetilde O(1/\sqrt{T})$ as long as the number of rounds does not grow too fast as a function of $n$.
♻ ☆ Integrating spoken instructions into flight trajectory prediction to optimize automation in air traffic control
The booming air transportation industry inevitably burdens air traffic controllers' workload, causing unexpected human factor-related incidents. Current air traffic control systems fail to consider spoken instructions for traffic prediction, bringing significant challenges in detecting human errors during real-time traffic operations. Here, we present an automation paradigm integrating controlling intent into the information processing loop through the spoken instruction-aware flight trajectory prediction framework. A 3-stage progressive multi-modal learning paradigm is proposed to address the modality gap between the trajectory and spoken instructions, as well as minimize the data requirements. Experiments on a real-world dataset show the proposed framework achieves flight trajectory prediction with high predictability and timeliness, obtaining over 20% relative reduction in mean deviation error. Moreover, the generalizability of the proposed framework is also confirmed by various model architectures. The proposed framework can formulate full-automated information processing in real-world air traffic applications, supporting human error detection and enhancing aviation safety.
comment: This paper has been accepted in principle by Nature Communications
♻ ☆ Large Language Model Based Generative Error Correction: A Challenge and Baselines for Speech Recognition, Speaker Tagging, and Emotion Recognition
Given recent advances in generative AI technology, a key question is how large language models (LLMs) can enhance acoustic modeling tasks using text decoding results from a frozen, pretrained automatic speech recognition (ASR) model. To explore new capabilities in language modeling for speech processing, we introduce the generative speech transcription error correction (GenSEC) challenge. This challenge comprises three post-ASR language modeling tasks: (i) post-ASR transcription correction, (ii) speaker tagging, and (iii) emotion recognition. These tasks aim to emulate future LLM-based agents handling voice-based interfaces while remaining accessible to a broad audience by utilizing open pretrained language models or agent-based APIs. We also discuss insights from baseline evaluations, as well as lessons learned for designing future evaluations.
comment: IEEE SLT 2024. The initial draft version has been done in December 2023. Post-ASR Text Processing and Understanding Community and LlaMA-7B pre-training correction model: https://huggingface.co/GenSEC-LLM/SLT-Task1-Llama2-7b-HyPo-baseline
♻ ☆ WaterMax: breaking the LLM watermark detectability-robustness-quality trade-off
Watermarking is a technical means to dissuade malfeasant usage of Large Language Models. This paper proposes a novel watermarking scheme, so-called WaterMax, that enjoys high detectability while sustaining the quality of the generated text of the original LLM. Its new design leaves the LLM untouched (no modification of the weights, logits, temperature, or sampling technique). WaterMax balances robustness and complexity contrary to the watermarking techniques of the literature inherently provoking a trade-off between quality and robustness. Its performance is both theoretically proven and experimentally validated. It outperforms all the SotA techniques under the most complete benchmark suite. Code available at https://github.com/eva-giboulot/WaterMax.
♻ ☆ Towards Satellite Non-IID Imagery: A Spectral Clustering-Assisted Federated Learning Approach
Low Earth orbit (LEO) satellites are capable of gathering abundant Earth observation data (EOD) to enable different Internet of Things (IoT) applications. However, to accomplish an effective EOD processing mechanism, it is imperative to investigate: 1) the challenge of processing the observed data without transmitting those large-size data to the ground because the connection between the satellites and the ground stations is intermittent, and 2) the challenge of processing the non-independent and identically distributed (non-IID) satellite data. In this paper, to cope with those challenges, we propose an orbit-based spectral clustering-assisted clustered federated self-knowledge distillation (OSC-FSKD) approach for each orbit of an LEO satellite constellation, which retains the advantage of FL that the observed data does not need to be sent to the ground. Specifically, we introduce normalized Laplacian-based spectral clustering (NLSC) into federated learning (FL) to create clustered FL in each round to address the challenge resulting from non-IID data. Particularly, NLSC is adopted to dynamically group clients into several clusters based on cosine similarities calculated by model updates. In addition, self-knowledge distillation is utilized to construct each local client, where the most recent updated local model is used to guide current local model training. Experiments demonstrate that the observation accuracy obtained by the proposed method is separately 1.01x, 2.15x, 1.10x, and 1.03x higher than that of pFedSD, FedProx, FedAU, and FedALA approaches using the SAT4 dataset. The proposed method also shows superiority when using other datasets.
comment: 10 pages, 5 figures
♻ ☆ Node Identifiers: Compact, Discrete Representations for Efficient Graph Learning
We present a novel end-to-end framework that generates highly compact (typically 6-15 dimensions), discrete (int4 type), and interpretable node representations, termed node identifiers (node IDs), to tackle inference challenges on large-scale graphs. By employing vector quantization, we compress continuous node embeddings from multiple layers of a Graph Neural Network (GNN) into discrete codes, applicable under both self-supervised and supervised learning paradigms. These node IDs capture high-level abstractions of graph data and offer interpretability that traditional GNN embeddings lack. Extensive experiments on 34 datasets, encompassing node classification, graph classification, link prediction, and attributed graph clustering tasks, demonstrate that the generated node IDs significantly enhance speed and memory efficiency while achieving competitive performance compared to current state-of-the-art methods.
♻ ☆ ScoreFusion: fusing score-based generative models via Kullback-Leibler barycenters
We introduce ScoreFusion, a theoretically grounded method for fusing multiple pre-trained diffusion models that are assumed to generate from auxiliary populations. ScoreFusion is particularly useful for enhancing the generative modeling of a target population with limited observed data. Our starting point considers the family of KL barycenters of the auxiliary populations, which is proven to be an optimal parametric class in the KL sense, but difficult to learn. Nevertheless, by recasting the learning problem as score matching in denoising diffusion, we obtain a tractable way of computing the optimal KL barycenter weights. We prove a dimension-free sample complexity bound in total variation distance, provided that the auxiliary models are well fitted for their own task and the auxiliary tasks combined capture the target well. We also explain a connection of the practice of checkpoint merging in AI art creation to an approximation of our KL-barycenter-based fusion approach. However, our fusion method differs in key aspects, allowing generation of new populations, as we illustrate in experiments.
comment: 53 pages, 15 figures
♻ ☆ FedSN: A Federated Learning Framework over Heterogeneous LEO Satellite Networks
Recently, a large number of Low Earth Orbit (LEO) satellites have been launched and deployed successfully in space by commercial companies, such as SpaceX. Due to multimodal sensors equipped by the LEO satellites, they serve not only for communication but also for various machine learning applications, such as space modulation recognition, remote sensing image classification, etc. However, the ground station (GS) may be incapable of downloading such a large volume of raw sensing data for centralized model training due to the limited contact time with LEO satellites (e.g. 5 minutes). Therefore, federated learning (FL) has emerged as the promising solution to address this problem via on-device training. Unfortunately, to enable FL on LEO satellites, we still face three critical challenges that are i) heterogeneous computing and memory capabilities, ii) limited uplink rate, and iii) model staleness. To this end, we propose FedSN as a general FL framework to tackle the above challenges, and fully explore data diversity on LEO satellites. Specifically, we first present a novel sub-structure scheme to enable heterogeneous local model training considering different computing, memory, and communication constraints on LEO satellites. Additionally, we propose a pseudo-synchronous model aggregation strategy to dynamically schedule model aggregation for compensating model staleness. To further demonstrate the effectiveness of the FedSN, we evaluate it using space modulation recognition and remote sensing image classification tasks by leveraging the data from real-world satellite networks. Extensive experimental results demonstrate that FedSN framework achieves higher accuracy, lower computing, and communication overhead than the state-of-the-art benchmarks and the effectiveness of each components in FedSN.
comment: 15 pages, 17 figures
♻ ☆ Social Dynamics of Consumer Response: A Unified Framework Integrating Statistical Physics and Marketing Dynamics
Understanding how consumers react to advertising inputs is essential for marketers aiming to optimize advertising strategies and improve campaign effectiveness. This study examines the complex nature of consumer behaviour by applying theoretical frameworks derived from physics and social psychology. We present an innovative equation that captures the relation between spending on advertising and consumer response, using concepts such as symmetries, scaling laws, and phase transitions. By validating our equation against well-known models such as the Michaelis-Menten and Hill equations, we prove its effectiveness in accurately representing the complexity of consumer response dynamics. The analysis emphasizes the importance of key model parameters, such as marketing effectiveness, response sensitivity, and behavioural sensitivity, in influencing consumer behaviour. The work explores the practical implications for advertisers and marketers, as well as discussing the limitations and future research directions. In summary, this study provides a thorough framework for comprehending and forecasting consumer reactions to advertising, which has implications for optimizing advertising strategies and allocating resources.
♻ ☆ Theories of synaptic memory consolidation and intelligent plasticity for continual learning
Humans and animals learn throughout life. Such continual learning is crucial for intelligence. In this chapter, we examine the pivotal role plasticity mechanisms with complex internal synaptic dynamics could play in enabling this ability in neural networks. By surveying theoretical research, we highlight two fundamental enablers for continual learning. First, synaptic plasticity mechanisms must maintain and evolve an internal state over several behaviorally relevant timescales. Second, plasticity algorithms must leverage the internal state to intelligently regulate plasticity at individual synapses to facilitate the seamless integration of new memories while avoiding detrimental interference with existing ones. Our chapter covers successful applications of these principles to deep neural networks and underscores the significance of synaptic metaplasticity in sustaining continual learning capabilities. Finally, we outline avenues for further research to understand the brain's superb continual learning abilities and harness similar mechanisms for artificial intelligence systems.
comment: An introductory-level book chapter. 35 pages, 14 figures
♻ ☆ Fisher Information-based Efficient Curriculum Federated Learning with Large Language Models EMNLP 2024
As a promising paradigm to collaboratively train models with decentralized data, Federated Learning (FL) can be exploited to fine-tune Large Language Models (LLMs). While LLMs correspond to huge size, the scale of the training data significantly increases, which leads to tremendous amounts of computation and communication costs. The training data is generally non-Independent and Identically Distributed (non-IID), which requires adaptive data processing within each device. Although Low Rank Adaptation (LoRA) can significantly reduce the scale of parameters to update in the fine-tuning process, it still takes unaffordable time to transfer the low-rank parameters of all the layers in LLMs. In this paper, we propose a Fisher Information-based Efficient Curriculum Federated Learning framework (FibecFed) with two novel methods, i.e., adaptive federated curriculum learning and efficient sparse parameter update. First, we propose a fisher information-based method to adaptively sample data within each device to improve the effectiveness of the FL fine-tuning process. Second, we dynamically select the proper layers for global aggregation and sparse parameters for local update with LoRA so as to improve the efficiency of the FL fine-tuning process. Extensive experimental results based on 10 datasets demonstrate that FibecFed yields excellent performance (up to 45.35% in terms of accuracy) and superb fine-tuning speed (up to 98.61% faster) compared with 17 baseline approaches).
comment: 27 pages, 8 figures, 14 tables, to appear in EMNLP 2024
♻ ☆ QUIS: Question-guided Insights Generation for Automated Exploratory Data Analysis
Discovering meaningful insights from a large dataset, known as Exploratory Data Analysis (EDA), is a challenging task that requires thorough exploration and analysis of the data. Automated Data Exploration (ADE) systems use goal-oriented methods with Large Language Models and Reinforcement Learning towards full automation. However, these methods require human involvement to anticipate goals that may limit insight extraction, while fully automated systems demand significant computational resources and retraining for new datasets. We introduce QUIS, a fully automated EDA system that operates in two stages: insight generation (ISGen) driven by question generation (QUGen). The QUGen module generates questions in iterations, refining them from previous iterations to enhance coverage without human intervention or manually curated examples. The ISGen module analyzes data to produce multiple relevant insights in response to each question, requiring no prior training and enabling QUIS to adapt to new datasets.
comment: Accepted for ENLP 2024 Industry Track
♻ ☆ SciLitLLM: How to Adapt LLMs for Scientific Literature Understanding
Scientific literature understanding is crucial for extracting targeted information and garnering insights, thereby significantly advancing scientific discovery. Despite the remarkable success of Large Language Models (LLMs), they face challenges in scientific literature understanding, primarily due to (1) a lack of scientific knowledge and (2) unfamiliarity with specialized scientific tasks. To develop an LLM specialized in scientific literature understanding, we propose a hybrid strategy that integrates continual pre-training (CPT) and supervised fine-tuning (SFT), to simultaneously infuse scientific domain knowledge and enhance instruction-following capabilities for domain-specific tasks.cIn this process, we identify two key challenges: (1) constructing high-quality CPT corpora, and (2) generating diverse SFT instructions. We address these challenges through a meticulous pipeline, including PDF text extraction, parsing content error correction, quality filtering, and synthetic instruction creation. Applying this strategy, we present a suite of LLMs: SciLitLLM, specialized in scientific literature understanding. These models demonstrate promising performance on scientific literature understanding benchmarks. Our contributions are threefold: (1) We present an effective framework that integrates CPT and SFT to adapt LLMs to scientific literature understanding, which can also be easily adapted to other domains. (2) We propose an LLM-based synthesis method to generate diverse and high-quality scientific instructions, resulting in a new instruction set -- SciLitIns -- for supervised fine-tuning in less-represented scientific domains. (3) SciLitLLM achieves promising performance improvements on scientific literature understanding benchmarks.
♻ ☆ Explainable Graph Neural Networks Under Fire
Predictions made by graph neural networks (GNNs) usually lack interpretability due to their complex computational behavior and the abstract nature of graphs. In an attempt to tackle this, many GNN explanation methods have emerged. Their goal is to explain a model's predictions and thereby obtain trust when GNN models are deployed in decision critical applications. Most GNN explanation methods work in a post-hoc manner and provide explanations in the form of a small subset of important edges and/or nodes. In this paper we demonstrate that these explanations can unfortunately not be trusted, as common GNN explanation methods turn out to be highly susceptible to adversarial perturbations. That is, even small perturbations of the original graph structure that preserve the model's predictions may yield drastically different explanations. This calls into question the trustworthiness and practical utility of post-hoc explanation methods for GNNs. To be able to attack GNN explanation models, we devise a novel attack method dubbed \textit{GXAttack}, the first \textit{optimization-based} adversarial white-box attack method for post-hoc GNN explanations under such settings. Due to the devastating effectiveness of our attack, we call for an adversarial evaluation of future GNN explainers to demonstrate their robustness. For reproducibility, our code is available via GitHub.
♻ ☆ LatentExplainer: Explaining Latent Representations in Deep Generative Models with Multi-modal Foundation Models
Deep generative models like VAEs and diffusion models have advanced various generation tasks by leveraging latent variables to learn data distributions and generate high-quality samples. Despite the field of explainable AI making strides in interpreting machine learning models, understanding latent variables in generative models remains challenging. This paper introduces \textit{LatentExplainer}, a framework for automatically generating semantically meaningful explanations of latent variables in deep generative models. \textit{LatentExplainer} tackles three main challenges: inferring the meaning of latent variables, aligning explanations with inductive biases, and handling varying degrees of explainability. Our approach perturbs latent variables, interpreting changes in generated data, and uses multi-modal large language models (MLLMs) to produce human-understandable explanations. We evaluate our proposed method on several real-world and synthetic datasets, and the results demonstrate superior performance in generating high-quality explanations for latent variables. The results highlight the effectiveness of incorporating inductive biases and uncertainty quantification, significantly enhancing model interpretability.
♻ ☆ Supervised Fine-Tuning Achieve Rapid Task Adaption Via Alternating Attention Head Activation Patterns
LLMs' performance on complex tasks is still unsatisfactory. A key issue is that presently LLMs learn in a data-driven schema, while the instructions about these complex tasks are both scarce and hard to collect or construct. On the contrary, a prominent phenomenon is that LLMs can learn rather fast on simpler tasks with adequate prior knowledge captured during pretraining stage. Thus, if the prerequisite and mechanism of such rapid generalization could be elucidated, it could enhance the efficiency and effectiveness of the LLM's ability to learn complex tasks. Thus, in this paper, we employ a gradient-based method, to dissect the process that the SFT process adapts LLMs to downstream tasks via the perspective of attention patterns. We find that: (1) LLMs selectively activate task-specific attention heads during SFT; (2) activation patterns for complex tasks are combinations of basic task patterns; and (3) changes in a few parameters can significantly impact activation patterns after SFT on a small number of samples.Based on these insights, experiments are conducted to actually enhance the efficiency and effectiveness of SFT.
comment: in review
♻ ☆ From Introspection to Best Practices: Principled Analysis of Demonstrations in Multimodal In-Context Learning
Motivated by in-context learning (ICL) capabilities of Large Language models (LLMs), multimodal LLMs with additional visual modality are also exhibited with similar ICL abilities when multiple image-text pairs are provided as demonstrations. However, relatively less work has been done to investigate the principles behind how and why multimodal ICL works. We conduct a systematic and principled evaluation of multimodal ICL for models of different scales on a broad spectrum of new yet critical tasks. Through perturbations over different modality information, we show that modalities matter differently across tasks in multimodal ICL. Guided by task-specific modality impact, we recommend modality-driven demonstration strategies to boost ICL performance. We also find that models may follow inductive biases from multimodal ICL even if they are rarely seen in or contradict semantic priors from pretraining data. Our principled analysis provides a comprehensive way of understanding the role of demonstrations in multimodal in-context learning, and sheds light on effectively improving multimodal ICL on a wide range of tasks.
♻ ☆ Tight bounds on Pauli channel learning without entanglement
Quantum entanglement is a crucial resource for learning properties from nature, but a precise characterization of its advantage can be challenging. In this work, we consider learning algorithms without entanglement to be those that only utilize states, measurements, and operations that are separable between the main system of interest and an ancillary system. Interestingly, we show that these algorithms are equivalent to those that apply quantum circuits on the main system interleaved with mid-circuit measurements and classical feedforward. Within this setting, we prove a tight lower bound for Pauli channel learning without entanglement that closes the gap between the best-known upper and lower bound. In particular, we show that $\Theta(2^n\varepsilon^{-2})$ rounds of measurements are required to estimate each eigenvalue of an $n$-qubit Pauli channel to $\varepsilon$ error with high probability when learning without entanglement. In contrast, a learning algorithm with entanglement only needs $\Theta(\varepsilon^{-2})$ copies of the Pauli channel. The tight lower bound strengthens the foundation for an experimental demonstration of entanglement-enhanced advantages for Pauli noise characterization.
comment: 20 pages, 2 figure; v2: add Fig.2 comparing our lower bound with noisy-entanglement-assisted upper bound, close to accepted version; v3: correct a typo in Notes added
♻ ☆ Emergence in non-neural models: grokking modular arithmetic via average gradient outer product
Neural networks trained to solve modular arithmetic tasks exhibit grokking, a phenomenon where the test accuracy starts improving long after the model achieves 100% training accuracy in the training process. It is often taken as an example of "emergence", where model ability manifests sharply through a phase transition. In this work, we show that the phenomenon of grokking is not specific to neural networks nor to gradient descent-based optimization. Specifically, we show that this phenomenon occurs when learning modular arithmetic with Recursive Feature Machines (RFM), an iterative algorithm that uses the Average Gradient Outer Product (AGOP) to enable task-specific feature learning with general machine learning models. When used in conjunction with kernel machines, iterating RFM results in a fast transition from random, near zero, test accuracy to perfect test accuracy. This transition cannot be predicted from the training loss, which is identically zero, nor from the test loss, which remains constant in initial iterations. Instead, as we show, the transition is completely determined by feature learning: RFM gradually learns block-circulant features to solve modular arithmetic. Paralleling the results for RFM, we show that neural networks that solve modular arithmetic also learn block-circulant features. Furthermore, we present theoretical evidence that RFM uses such block-circulant features to implement the Fourier Multiplication Algorithm, which prior work posited as the generalizing solution neural networks learn on these tasks. Our results demonstrate that emergence can result purely from learning task-relevant features and is not specific to neural architectures nor gradient descent-based optimization methods. Furthermore, our work provides more evidence for AGOP as a key mechanism for feature learning in neural networks.
♻ ☆ LayerSkip: Enabling Early Exit Inference and Self-Speculative Decoding ACL 2024
We present LayerSkip, an end-to-end solution to speed-up inference of large language models (LLMs). First, during training we apply layer dropout, with low dropout rates for earlier layers and higher dropout rates for later layers, and an early exit loss where all transformer layers share the same exit. Second, during inference, we show that this training recipe increases the accuracy of early exit at earlier layers, without adding any auxiliary layers or modules to the model. Third, we present a novel self-speculative decoding solution where we exit at early layers and verify and correct with remaining layers of the model. Our proposed self-speculative decoding approach has less memory footprint than other speculative decoding approaches and benefits from shared compute and activations of the draft and verification stages. We run experiments on different Llama model sizes on different types of training: pretraining from scratch, continual pretraining, finetuning on specific data domain, and finetuning on specific task. We implement our inference solution and show speedups of up to 2.16x on summarization for CNN/DM documents, 1.82x on coding, and 2.0x on TOPv2 semantic parsing task. We open source our code and checkpoints at https://github.com/facebookresearch/LayerSkip.
comment: ACL 2024
♻ ☆ LeanAgent: Lifelong Learning for Formal Theorem Proving
Large Language Models (LLMs) have been successful in mathematical reasoning tasks such as formal theorem proving when integrated with interactive proof assistants like Lean. Existing approaches involve training or fine-tuning an LLM on a specific dataset to perform well on particular domains, such as undergraduate-level mathematics. These methods struggle with generalizability to advanced mathematics. A fundamental limitation is that these approaches operate on static domains, failing to capture how mathematicians often work across multiple domains and projects simultaneously or cyclically. We present LeanAgent, a novel lifelong learning framework for theorem proving that continuously generalizes to and improves on ever-expanding mathematical knowledge without forgetting previously learned knowledge. LeanAgent introduces several key innovations, including a curriculum learning strategy that optimizes the learning trajectory in terms of mathematical difficulty, a dynamic database for efficient management of evolving mathematical knowledge, and progressive training to balance stability and plasticity. LeanAgent successfully proves 162 theorems previously unproved by humans across 23 diverse Lean repositories, many from advanced mathematics. It performs significantly better than the static LLM baseline, proving challenging theorems in domains like abstract algebra and algebraic topology while showcasing a clear progression of learning from basic concepts to advanced topics. In addition, we analyze LeanAgent's superior performance on key lifelong learning metrics. LeanAgent achieves exceptional scores in stability and backward transfer, where learning new tasks improves performance on previously learned tasks. This emphasizes LeanAgent's continuous generalizability and improvement, explaining its superior theorem-proving performance.
♻ ☆ A Tighter Complexity Analysis of SparseGPT
In this work, we improved the analysis of the running time of SparseGPT [Frantar, Alistarh ICML 2023] from $O(d^{3})$ to $O(d^{\omega} + d^{2+a+o(1)} + d^{1+\omega(1,1,a)-a})$ for any $a \in [0, 1]$, where $\omega$ is the exponent of matrix multiplication. In particular, for the current $\omega \approx 2.371$ [Alman, Duan, Williams, Xu, Xu, Zhou 2024], our running time boils down to $O(d^{2.53})$. This running time is due to the analysis of the lazy update behavior in iterative maintenance problems such as [Deng, Song, Weinstein 2022; Brand, Song, Zhou ICML 2024].
♻ ☆ Granger Causality in Extremes
We introduce a rigorous mathematical framework for Granger causality in extremes, designed to identify causal links from extreme events in time series. Granger causality plays a pivotal role in uncovering directional relationships among time-varying variables. While this notion gains heightened importance during extreme and highly volatile periods, state-of-the-art methods primarily focus on causality within the body of the distribution, often overlooking causal mechanisms that manifest only during extreme events. Our framework is designed to infer causality mainly from extreme events by leveraging the causal tail coefficient. We establish equivalences between causality in extremes and other causal concepts, including (classical) Granger causality, Sims causality, and structural causality. We prove other key properties of Granger causality in extremes and show that the framework is especially helpful under the presence of hidden confounders. We also propose a novel inference method for detecting the presence of Granger causality in extremes from data. Our method is model-free, can handle non-linear and high-dimensional time series, outperforms current state-of-the-art methods in all considered setups, both in performance and speed, and was found to uncover coherent effects when applied to financial and extreme weather observations.
♻ ☆ Timer: Generative Pre-trained Transformers Are Large Time Series Models
Deep learning has contributed remarkably to the advancement of time series analysis. Still, deep models can encounter performance bottlenecks in real-world data-scarce scenarios, which can be concealed due to the performance saturation with small models on current benchmarks. Meanwhile, large models have demonstrated great powers in these scenarios through large-scale pre-training. Continuous progress has been achieved with the emergence of large language models, exhibiting unprecedented abilities such as few-shot generalization, scalability, and task generality, which are however absent in small deep models. To change the status quo of training scenario-specific small models from scratch, this paper aims at the early development of large time series models (LTSM). During pre-training, we curate large-scale datasets with up to 1 billion time points, unify heterogeneous time series into single-series sequence (S3) format, and develop the GPT-style architecture toward LTSMs. To meet diverse application needs, we convert forecasting, imputation, and anomaly detection of time series into a unified generative task. The outcome of this study is a Time Series Transformer (Timer), which is generative pre-trained by next token prediction and adapted to various downstream tasks with promising capabilities as an LTSM. Code and datasets are available at: https://github.com/thuml/Large-Time-Series-Model.
♻ ☆ Experimenting on Markov Decision Processes with Local Treatments
Utilizing randomized experiments to evaluate the effect of short-term treatments on the short-term outcomes has been well understood and become the golden standard in industrial practice. However, as service systems become increasingly dynamical and personalized, much focus is shifting toward maximizing long-term cumulative outcomes, such as customer lifetime value, through lifetime exposure to interventions. To bridge this gap, we investigate the randomized experiments within dynamical systems modeled as Markov Decision Processes (MDPs). Our goal is to assess the impact of treatment and control policies on long-term cumulative rewards from relatively short-term observations. We first develop optimal inference techniques for assessing the effects of general treatment patterns. Furthermore, recognizing that many real-world treatments tend to be fine-grained and localized for practical efficiency and operational convenience, we then propose methods to harness this localized structure by sharing information on the non-targeted states. Our new estimator effectively overcomes the variance lower bound for general treatments while matching the more stringent lower bound incorporating the local treatment structure. Furthermore, our estimator can optimally achieve a linear reduction with the number of test arms for a major part of the variance. Finally, we explore scenarios with perfect knowledge of the control arm and design estimators that further improve inference efficiency.
♻ ☆ $\textbf{Only-IF}$:Revealing the Decisive Effect of Instruction Diversity on Generalization
Understanding and accurately following instructions is critical for large language models (LLMs) to be effective across diverse tasks. In this work, we rigorously examine the key factors that enable models to generalize to unseen instructions, providing insights to guide the collection of data for instruction-tuning. Through controlled experiments, inspired by the Turing-complete Markov algorithm, we demonstrate that such generalization $\textbf{only emerges}$ when training data is diversified enough across semantic domains. Our findings also reveal that merely diversifying within limited domains fails to ensure robust generalization. In contrast, cross-domain data diversification, even under constrained data budgets, significantly enhances a model's adaptability. We further extend our analysis to real-world scenarios, including fine-tuning of $\textit{$\textbf{specialist}$}$ and $\textit{$\textbf{generalist}$}$ models. In both cases, we demonstrate that 1) better performance can be achieved by increasing the diversity of an established dataset while keeping the data size constant, and 2) when scaling up the data, diversifying the semantics of instructions is more effective than simply increasing the quantity of similar data. Our research provides important insights for dataset collation, particularly when optimizing model performance by expanding training data for both specialist and generalist scenarios. We show that careful consideration of data diversification is key: training specialist models with data extending beyond their core domain leads to significant performance improvements, while generalist models benefit from diverse data mixtures that enhance their overall instruction-following capabilities across a wide range of applications. Our results highlight the critical role of strategic diversification and offer clear guidelines for improving data quality.
comment: Fix formatting issues
♻ ☆ A Data-Adaptive Prior for Bayesian Learning of Kernels in Operators
Kernels are efficient in representing nonlocal dependence and they are widely used to design operators between function spaces. Thus, learning kernels in operators from data is an inverse problem of general interest. Due to the nonlocal dependence, the inverse problem can be severely ill-posed with a data-dependent singular inversion operator. The Bayesian approach overcomes the ill-posedness through a non-degenerate prior. However, a fixed non-degenerate prior leads to a divergent posterior mean when the observation noise becomes small, if the data induces a perturbation in the eigenspace of zero eigenvalues of the inversion operator. We introduce a data-adaptive prior to achieve a stable posterior whose mean always has a small noise limit. The data-adaptive prior's covariance is the inversion operator with a hyper-parameter selected adaptive to data by the L-curve method. Furthermore, we provide a detailed analysis on the computational practice of the data-adaptive prior, and demonstrate it on Toeplitz matrices and integral operators. Numerical tests show that a fixed prior can lead to a divergent posterior mean in the presence of any of the four types of errors: discretization error, model error, partial observation and wrong noise assumption. In contrast, the data-adaptive prior always attains posterior means with small noise limits.
♻ ☆ MoR: Mixture of Ranks for Low-Rank Adaptation Tuning
Low-Rank Adaptation (LoRA) drives research to align its performance with full fine-tuning. However, significant challenges remain: (1) Simply increasing the rank size of LoRA does not effectively capture high-rank information, which leads to a performance bottleneck.(2) MoE-style LoRA methods substantially increase parameters and inference latency, contradicting the goals of efficient fine-tuning and ease of application. To address these challenges, we introduce Mixture of Ranks (MoR), which learns rank-specific information for different tasks based on input and efficiently integrates multi-rank information. We firstly propose a new framework that equates the integration of multiple LoRAs to expanding the rank of LoRA. Moreover, we hypothesize that low-rank LoRA already captures sufficient intrinsic information, and MoR can derive high-rank information through mathematical transformations of the low-rank components. Thus, MoR can reduces the learning difficulty of LoRA and enhances its multi-task capabilities. MoR achieves impressive results, with MoR delivering a 1.31\% performance improvement while using only 93.93\% of the parameters compared to baseline methods.
comment: 11 pages, 7 figures
♻ ☆ UniAutoML: A Human-Centered Framework for Unified Discriminative and Generative AutoML with Large Language Models
Automated Machine Learning (AutoML) has simplified complex ML processes such as data pre-processing, model selection, and hyper-parameter searching. However, traditional AutoML frameworks focus solely on discriminative tasks, often falling short in tackling AutoML for generative models. Additionally, these frameworks lack interpretability and user engagement during the training process, primarily due to the absence of human-centered design. It leads to a lack of transparency in final decision-making and limited user control, potentially reducing trust and adoption of AutoML methods. To address these limitations, we introduce UniAutoML, a human-centered AutoML framework that leverages Large Language Models (LLMs) to unify AutoML for both discriminative (e.g., Transformers and CNNs for classification or regression tasks) and generative tasks (e.g., fine-tuning diffusion models or LLMs). The human-centered design of UniAutoML innovatively features a conversational user interface (CUI) that facilitates natural language interactions, providing users with real-time guidance, feedback, and progress updates for better interpretability. This design enhances transparency and user control throughout the AutoML training process, allowing users to seamlessly break down or modify the model being trained. To mitigate potential risks associated with LLM generated content, UniAutoML incorporates a safety guardline that filters inputs and censors outputs. We evaluated UniAutoML's performance and usability through experiments on eight diverse datasets and user studies involving 25 participants, demonstrating that UniAutoML not only enhances performance but also improves user control and trust. Our human-centered design bridges the gap between AutoML capabilities and user understanding, making ML more accessible to a broader audience.
♻ ☆ Contextual Bandits with Packing and Covering Constraints: A Modular Lagrangian Approach via Regression COLT 2023
We consider contextual bandits with linear constraints (CBwLC), a variant of contextual bandits in which the algorithm consumes multiple resources subject to linear constraints on total consumption. This problem generalizes contextual bandits with knapsacks (CBwK), allowing for packing and covering constraints, as well as positive and negative resource consumption. We provide the first algorithm for CBwLC (or CBwK) that is based on regression oracles. The algorithm is simple, computationally efficient, and statistically optimal under mild assumptions. Further, we provide the first vanishing-regret guarantees for CBwLC (or CBwK) that extend beyond the stochastic environment. We side-step strong impossibility results from prior work by identifying a weaker (and, arguably, fairer) benchmark to compare against. Our algorithm builds on LagrangeBwK (Immorlica et al., FOCS 2019), a Lagrangian-based technique for CBwK, and SquareCB (Foster and Rakhlin, ICML 2020), a regression-based technique for contextual bandits. Our analysis leverages the inherent modularity of both techniques.
comment: A preliminary version of this paper, authored by A. Slivkins, K.A. Sankararaman and D.J. Foster, has been published at COLT 2023. The present version (since Jun'24) features an important improvement, due to Xingyu Zhou. The most recent version fixes an inaccuracy in Section 6 when the analysis from Section 4 is invoked
♻ ☆ Disentangling Heterogeneous Knowledge Concept Embedding for Cognitive Diagnosis on Untested Knowledge
Cognitive diagnosis is a fundamental and critical task in learning assessment, which aims to infer students' proficiency on knowledge concepts from their response logs. Current works assume each knowledge concept will certainly be tested and covered by multiple exercises. However, whether online or offline courses, it's hardly feasible to completely cover all knowledge concepts in several exercises. Restricted tests lead to undiscovered knowledge deficits, especially untested knowledge concepts(UKCs). In this paper, we propose a novel framework for Cognitive Diagnosis called Disentangling Heterogeneous Knowledge Cognitive Diagnosis(DisKCD) on untested knowledge. Specifically, we leverage course grades, exercise questions, and learning resources to learn the potential representations of students, exercises, and knowledge concepts. In particular, knowledge concepts are disentangled into tested and untested based on the limiting actual exercises. We construct a heterogeneous relation graph network via students, exercises, tested knowledge concepts(TKCs), and UKCs. Then, through a hierarchical heterogeneous message-passing mechanism, the fine-grained relations are incorporated into the embeddings of the entities. Finally, the embeddings will be applied to multiple existing cognitive diagnosis models to infer students' proficiency on UKCs. Experimental results on real-world datasets show that the proposed model can effectively improve the performance of the task of diagnosing students' proficiency on UKCs. Our code is available at https://github.com/Hubuers/DisKCD.
♻ ☆ On Subjective Uncertainty Quantification and Calibration in Natural Language Generation
Applications of large language models often involve the generation of free-form responses, in which case uncertainty quantification becomes challenging. This is due to the need to identify task-specific uncertainties (e.g., about the semantics) which appears difficult to define in general cases. This work addresses these challenges from a perspective of Bayesian decision theory, starting from the assumption that our utility is characterized by a similarity measure that compares a generated response with a hypothetical true response. We discuss how this assumption enables principled quantification of the model's subjective uncertainty and its calibration. We further derive a measure for epistemic uncertainty, based on a missing data perspective and its characterization as an excess risk. The proposed methods can be applied to black-box language models. We illustrate the methods on question answering and machine translation tasks. Our experiments provide a principled evaluation of task-specific calibration, and demonstrate that epistemic uncertainty offers a promising deferral strategy for efficient data acquisition in in-context learning.
♻ ☆ An Evolved Universal Transformer Memory
Prior methods propose to offset the escalating costs of modern foundation models by dropping specific parts of their contexts with hand-designed rules, while attempting to preserve their original performance. We overcome this trade-off with Neural Attention Memory Models (NAMMs), introducing a learned network for memory management that improves both the performance and efficiency of transformers. We evolve NAMMs atop pre-trained transformers to provide different latent contexts focusing on the most relevant information for individual layers and attention heads. NAMMs are universally applicable to any model using self-attention as they condition exclusively on the values in the produced attention matrices. Learning NAMMs on a small set of problems, we achieve substantial performance improvements across multiple long-context benchmarks while cutting the model's input contexts up to a fraction of the original sizes. We show the generality of our conditioning enables zero-shot transfer of NAMMs trained only on language to entirely new transformer architectures even across input modalities, with their benefits carrying over to vision and reinforcement learning.
comment: 29 pages, 14 figures. Preprint, under submission. Source code is available at https://github.com/SakanaAI/evo-memory
♻ ☆ CPT: Competence-progressive Training Strategy for Few-shot Node Classification
Graph Neural Networks (GNNs) have made significant advancements in node classification, but their success relies on sufficient labeled nodes per class in the training data. Real-world graph data often exhibits a long-tail distribution with sparse labels, emphasizing the importance of GNNs' ability in few-shot node classification, which entails categorizing nodes with limited data. Traditional episodic meta-learning approaches have shown promise in this domain, but they face an inherent limitation: it might lead the model to converge to suboptimal solutions because of random and uniform task assignment, ignoring task difficulty levels. This could lead the meta-learner to face complex tasks too soon, hindering proper learning. Ideally, the meta-learner should start with simple concepts and advance to more complex ones, like human learning. So, we introduce CPT, a novel two-stage curriculum learning method that aligns task difficulty with the meta-learner's progressive competence, enhancing overall performance. Specifically, in CPT's initial stage, the focus is on simpler tasks, fostering foundational skills for engaging with complex tasks later. Importantly, the second stage dynamically adjusts task difficulty based on the meta-learner's growing competence, aiming for optimal knowledge acquisition. Extensive experiments on popular node classification datasets demonstrate significant improvements of our strategy over existing methods.
comment: arXiv admin note: text overlap with arXiv:2206.11972 by other authors
♻ ☆ Discrete Messages Improve Communication Efficiency among Isolated Intelligent Agents
Individuals, despite having varied life experiences and learning processes, can communicate effectively through languages. This study aims to explore the efficiency of language as a communication medium. We put forth two specific hypotheses: First, discrete messages are more effective than continuous ones when agents have diverse personal experiences. Second, communications using multiple discrete tokens are more advantageous than those using a single token. To valdate these hypotheses, we designed multi-agent machine learning experiments to assess communication efficiency using various information transmission methods between speakers and listeners. Our empirical findings indicate that, in scenarios where agents are exposed to different data, communicating through sentences composed of discrete tokens offers the best inter-agent communication efficiency. The limitations of our finding include lack of systematic advantages over other more sophisticated encoder-decoder model such as variational autoencoder and lack of evluation on non-image dataset, which we will leave for future studies.
♻ ☆ Residual-INR: Communication Efficient On-Device Learning Using Implicit Neural Representation
Edge computing is a distributed computing paradigm that collects and processes data at or near the source of data generation. The on-device learning at edge relies on device-to-device wireless communication to facilitate real-time data sharing and collaborative decision-making among multiple devices. This significantly improves the adaptability of the edge computing system to the changing environments. However, as the scale of the edge computing system is getting larger, communication among devices is becoming the bottleneck because of the limited bandwidth of wireless communication leads to large data transfer latency. To reduce the amount of device-to-device data transmission and accelerate on-device learning, in this paper, we propose Residual-INR, a fog computing-based communication-efficient on-device learning framework by utilizing implicit neural representation (INR) to compress images/videos into neural network weights. Residual-INR enhances data transfer efficiency by collecting JPEG images from edge devices, compressing them into INR format at the fog node, and redistributing them for on-device learning. By using a smaller INR for full image encoding and a separate object INR for high-quality object region reconstruction through residual encoding, our technique can reduce the encoding redundancy while maintaining the object quality. Residual-INR is a promising solution for edge on-device learning because it reduces data transmission by up to 5.16 x across a network of 10 edge devices. It also facilitates CPU-free accelerated on-device learning, achieving up to 2.9 x speedup without sacrificing accuracy. Our code is available at: https://github.com/sharclab/Residual-INR.
comment: This paper has been accepted by ICCAD 2024
♻ ☆ Contextual Linear Optimization with Bandit Feedback
Contextual linear optimization (CLO) uses predictive contextual features to reduce uncertainty in random cost coefficients and thereby improve average-cost performance. An example is the stochastic shortest path problem with random edge costs (e.g., traffic) and contextual features (e.g., lagged traffic, weather). Existing work on CLO assumes the data has fully observed cost coefficient vectors, but in many applications, we can only see the realized cost of a historical decision, that is, just one projection of the random cost coefficient vector, to which we refer as bandit feedback. We study a class of offline learning algorithms for CLO with bandit feedback, which we term induced empirical risk minimization (IERM), where we fit a predictive model to directly optimize the downstream performance of the policy it induces. We show a fast-rate regret bound for IERM that allows for misspecified model classes and flexible choices of the optimization estimate, and we develop computationally tractable surrogate losses. A byproduct of our theory of independent interest is fast-rate regret bound for IERM with full feedback and misspecified policy class. We compare the performance of different modeling choices numerically using a stochastic shortest path example and provide practical insights from the empirical results.
♻ ☆ Efficient Arbitrary Precision Acceleration for Large Language Models on GPU Tensor Cores SP
Large language models (LLMs) have been widely applied but face challenges in efficient inference. While quantization methods reduce computational demands, ultra-low bit quantization with arbitrary precision is hindered by limited GPU Tensor Core support and inefficient memory management, leading to suboptimal acceleration. To address these challenges, we propose a comprehensive acceleration scheme for arbitrary precision LLMs. At its core, we introduce a novel bipolar-INT data format that facilitates parallel computing and supports symmetric quantization, effectively reducing data redundancy. Building on this, we implement an arbitrary precision matrix multiplication scheme that decomposes and recovers matrices at the bit level, enabling flexible precision while maximizing GPU Tensor Core utilization. Furthermore, we develop an efficient matrix preprocessing method that optimizes data layout for subsequent computations. Finally, we design a data recovery-oriented memory management system that strategically utilizes fast shared memory, significantly enhancing kernel execution speed and minimizing memory access latency. Experimental results demonstrate our approach's effectiveness, with up to 2.4\times speedup in matrix multiplication compared to NVIDIA's CUTLASS. When integrated into LLMs, we achieve up to 6.7\times inference acceleration. These improvements significantly enhance LLM inference efficiency, enabling broader and more responsive applications of LLMs.
comment: This paper is accepted by ASP-DAC 2025
♻ ☆ Path-based Explanation for Knowledge Graph Completion KDD 2024
Graph Neural Networks (GNNs) have achieved great success in Knowledge Graph Completion (KGC) by modelling how entities and relations interact in recent years. However, the explanation of the predicted facts has not caught the necessary attention. Proper explanations for the results of GNN-based KGC models increase model transparency and help researchers develop more reliable models. Existing practices for explaining KGC tasks rely on instance/subgraph-based approaches, while in some scenarios, paths can provide more user-friendly and interpretable explanations. Nonetheless, the methods for generating path-based explanations for KGs have not been well-explored. To address this gap, we propose Power-Link, the first path-based KGC explainer that explores GNN-based models. We design a novel simplified graph-powering technique, which enables the generation of path-based explanations with a fully parallelisable and memory-efficient training scheme. We further introduce three new metrics for quantitative evaluation of the explanations, together with a qualitative human evaluation. Extensive experiments demonstrate that Power-Link outperforms the SOTA baselines in interpretability, efficiency, and scalability.
comment: Accepted by KDD 2024
♻ ☆ Polyhedral Complex Derivation from Piecewise Trilinear Networks NeurIPS 2024
Recent advancements in visualizing deep neural networks provide insights into their structures and mesh extraction from Continuous Piecewise Affine (CPWA) functions. Meanwhile, developments in neural surface representation learning incorporate non-linear positional encoding, addressing issues like spectral bias; however, this poses challenges in applying mesh extraction techniques based on CPWA functions. Focusing on trilinear interpolating methods as positional encoding, we present theoretical insights and an analytical mesh extraction, showing the transformation of hypersurfaces to flat planes within the trilinear region under the eikonal constraint. Moreover, we introduce a method for approximating intersecting points among three hypersurfaces contributing to broader applications. We empirically validate correctness and parsimony through chamfer distance and efficiency, and angular distance, while examining the correlation between the eikonal loss and the planarity of the hypersurfaces.
comment: Accepted at NeurIPS 2024. Updated with the camera-ready version
♻ ☆ MOS: Model Synergy for Test-Time Adaptation on LiDAR-Based 3D Object Detection
LiDAR-based 3D object detection is crucial for various applications but often experiences performance degradation in real-world deployments due to domain shifts. While most studies focus on cross-dataset shifts, such as changes in environments and object geometries, practical corruptions from sensor variations and weather conditions remain underexplored. In this work, we propose a novel online test-time adaptation framework for 3D detectors that effectively tackles these shifts, including a challenging cross-corruption scenario where cross-dataset shifts and corruptions co-occur. By leveraging long-term knowledge from previous test batches, our approach mitigates catastrophic forgetting and adapts effectively to diverse shifts. Specifically, we propose a Model Synergy (MOS) strategy that dynamically selects historical checkpoints with diverse knowledge and assembles them to best accommodate the current test batch. This assembly is directed by our proposed Synergy Weights (SW), which perform a weighted averaging of the selected checkpoints, minimizing redundancy in the composite model. The SWs are computed by evaluating the similarity of predicted bounding boxes on the test data and the independence of features between checkpoint pairs in the model bank. To maintain an efficient and informative model bank, we discard checkpoints with the lowest average SW scores, replacing them with newly updated models. Our method was rigorously tested against existing test-time adaptation strategies across three datasets and eight types of corruptions, demonstrating superior adaptability to dynamic scenes and conditions. Notably, it achieved a 67.3% improvement in a challenging cross-corruption scenario, offering a more comprehensive benchmark for adaptation. The source code will be made publicly available.
Multimedia 8
☆ Parallel Backpropagation for Inverse of a Convolution with Application to Normalizing Flows
Inverse of an invertible convolution is an important operation that comes up in Normalizing Flows, Image Deblurring, etc. The naive algorithm for backpropagation of this operation using Gaussian elimination has running time $O(n^3)$ where $n$ is the number of pixels in the image. We give a fast parallel backpropagation algorithm with running time $O(\sqrt{n})$ for a square image and provide a GPU implementation of the same. Inverse Convolutions are usually used in Normalizing Flows in the sampling pass, making them slow. We propose to use Inverse Convolutions in the forward (image to latent vector) pass of the Normalizing flow. Since the sampling pass is the inverse of the forward pass, it will use convolutions only, resulting in efficient sampling times. We use our parallel backpropagation algorithm for optimizing the inverse convolution layer resulting in fast training times also. We implement this approach in various Normalizing Flow backbones, resulting in our Inverse-Flow models. We benchmark Inverse-Flow on standard datasets and show significantly improved sampling times with similar bits per dimension compared to previous models.
comment: Preprint
☆ RA-BLIP: Multimodal Adaptive Retrieval-Augmented Bootstrapping Language-Image Pre-training
Multimodal Large Language Models (MLLMs) have recently received substantial interest, which shows their emerging potential as general-purpose models for various vision-language tasks. MLLMs involve significant external knowledge within their parameters; however, it is challenging to continually update these models with the latest knowledge, which involves huge computational costs and poor interpretability. Retrieval augmentation techniques have proven to be effective plugins for both LLMs and MLLMs. In this study, we propose multimodal adaptive Retrieval-Augmented Bootstrapping Language-Image Pre-training (RA-BLIP), a novel retrieval-augmented framework for various MLLMs. Considering the redundant information within vision modality, we first leverage the question to instruct the extraction of visual information through interactions with one set of learnable queries, minimizing irrelevant interference during retrieval and generation. Besides, we introduce a pre-trained multimodal adaptive fusion module to achieve question text-to-multimodal retrieval and integration of multimodal knowledge by projecting visual and language modalities into a unified semantic space. Furthermore, we present an Adaptive Selection Knowledge Generation (ASKG) strategy to train the generator to autonomously discern the relevance of retrieved knowledge, which realizes excellent denoising performance. Extensive experiments on open multimodal question-answering datasets demonstrate that RA-BLIP achieves significant performance and surpasses the state-of-the-art retrieval-augmented models.
comment: 10 pages, 6 figures, Journal
♻ ☆ Movie101v2: Improved Movie Narration Benchmark
Automatic movie narration aims to generate video-aligned plot descriptions to assist visually impaired audiences. Unlike standard video captioning, it involves not only describing key visual details but also inferring plots that unfold across multiple movie shots, presenting distinct and complex challenges. To advance this field, we introduce Movie101v2, a large-scale, bilingual dataset with enhanced data quality specifically designed for movie narration. Revisiting the task, we propose breaking down the ultimate goal of automatic movie narration into three progressive stages, offering a clear roadmap with corresponding evaluation metrics. Based on our new benchmark, we baseline a range of large vision-language models, including GPT-4V, and conduct an in-depth analysis of the challenges in narration generation. Our findings highlight that achieving applicable movie narration generation is a fascinating goal that requires significant research.
♻ ☆ Perceptual Quality Assessment of Octree-RAHT Encoded 3D Point Clouds
No-reference bitstream-layer point cloud quality assessment (PCQA) can be deployed without full decoding at any network node to achieve real-time quality monitoring. In this work, we focus on the PCQA problem dedicated to Octree-RAHT encoding mode. First, to address the issue that existing PCQA databases have a small scale and limited distortion levels, we establish the WPC5.0 database which is the first one dedicated to Octree-RAHT encoding mode with a scale of 400 distorted point clouds (PCs) including 4 geometric multiplied by 5 attitude distortion levels. Then, we propose the first PCQA model dedicated to Octree-RAHT encoding mode by parsing PC bitstreams without full decoding. The model introduces texture bitrate (TBPP) to predict texture complexity (TC) and further derives the texture distortion factor. In addition, the Geometric Quantization Parameter (PQS) is used to estimate the geometric distortion factor, which is then integrated into the model along with the texture distortion factor to obtain the proposed PCQA model named streamPCQ-OR. The proposed model has been compared with other advanced PCQA methods on the WPC5.0, BASICS and M-PCCD databases, and experimental results show that our model has excellent performance while having very low computational complexity, providing a reliable choice for time-critical applications. To facilitate subsequent research, the database and source code will be publicly released at https://github.com/qdushl/Waterloo-Point-Cloud-Database-5.0.
♻ ☆ Evaluating Semantic Variation in Text-to-Image Synthesis: A Causal Perspective
Accurate interpretation and visualization of human instructions are crucial for text-to-image (T2I) synthesis. However, current models struggle to capture semantic variations from word order changes, and existing evaluations, relying on indirect metrics like text-image similarity, fail to reliably assess these challenges. This often obscures poor performance on complex or uncommon linguistic patterns by the focus on frequent word combinations. To address these deficiencies, we propose a novel metric called SemVarEffect and a benchmark named SemVarBench, designed to evaluate the causality between semantic variations in inputs and outputs in T2I synthesis. Semantic variations are achieved through two types of linguistic permutations, while avoiding easily predictable literal variations. Experiments reveal that the CogView-3-Plus and Ideogram 2 performed the best, achieving a score of 0.2/1. Semantic variations in object relations are less understood than attributes, scoring 0.07/1 compared to 0.17-0.19/1. We found that cross-modal alignment in UNet or Transformers plays a crucial role in handling semantic variations, a factor previously overlooked by a focus on textual encoders. Our work establishes an effective evaluation framework that advances the T2I synthesis community's exploration of human instruction understanding. Our benchmark and code are available at https://github.com/zhuxiangru/SemVarBench .
comment: The only change in the current version update is the replacement of the template with a more precise one
♻ ☆ MixEval-X: Any-to-Any Evaluations from Real-World Data Mixtures
Perceiving and generating diverse modalities are crucial for AI models to effectively learn from and engage with real-world signals, necessitating reliable evaluations for their development. We identify two major issues in current evaluations: (1) inconsistent standards, shaped by different communities with varying protocols and maturity levels; and (2) significant query, grading, and generalization biases. To address these, we introduce MixEval-X, the first any-to-any, real-world benchmark designed to optimize and standardize evaluations across diverse input and output modalities. We propose multi-modal benchmark mixture and adaptation-rectification pipelines to reconstruct real-world task distributions, ensuring evaluations generalize effectively to real-world use cases. Extensive meta-evaluations show our approach effectively aligns benchmark samples with real-world task distributions. Meanwhile, MixEval-X's model rankings correlate strongly with that of crowd-sourced real-world evaluations (up to 0.98) while being much more efficient. We provide comprehensive leaderboards to rerank existing models and organizations and offer insights to enhance understanding of multi-modal evaluations and inform future research.
♻ ☆ Synthesizing Sentiment-Controlled Feedback For Multimodal Text and Image Data
The ability to generate sentiment-controlled feedback in response to multimodal inputs comprising text and images addresses a critical gap in human-computer interaction. This capability allows systems to provide empathetic, accurate, and engaging responses, with useful applications in education, healthcare, marketing, and customer service. To this end, we have constructed a large-scale Controllable Multimodal Feedback Synthesis (CMFeed) dataset and propose a controllable feedback synthesis system. The system features an encoder, decoder, and controllability block for textual and visual inputs. It extracts features using a transformer and Faster R-CNN networks, combining them to generate feedback. The CMFeed dataset includes images, texts, reactions to the posts, human comments with relevance scores, and reactions to these comments. These reactions train the model to produce feedback with specified sentiments, achieving a sentiment classification accuracy of 77.23\%, which is 18.82\% higher than the accuracy without controllability. The system also incorporates a similarity module for assessing feedback relevance through rank-based metrics and an interpretability technique to analyze the contributions of textual and visual features during feedback generation. Access to the CMFeed dataset and the system's code is available at https://github.com/MIntelligence-Group/CMFeed.
♻ ☆ An automatic mixing speech enhancement system for multi-track audio
We propose a speech enhancement system for multitrack audio. The system will minimize auditory masking while allowing one to hear multiple simultaneous speakers. The system can be used in multiple communication scenarios e.g., teleconferencing, invoice gaming, and live streaming. The ITU-R BS.1387 Perceptual Evaluation of Audio Quality (PEAQ) model is used to evaluate the amount of masking in the audio signals. Different audio effects e.g., level balance, equalization, dynamic range compression, and spatialization are applied via an iterative Harmony searching algorithm that aims to minimize the masking. In the subjective listening test, the designed system can compete with mixes by professional sound engineers and outperforms mixes by existing auto-mixing systems.
comment: 5 pages
Computation and Language 226
☆ How Numerical Precision Affects Mathematical Reasoning Capabilities of LLMs
Despite the remarkable success of Transformer-based Large Language Models (LLMs) across various domains, understanding and enhancing their mathematical capabilities remains a significant challenge. In this paper, we conduct a rigorous theoretical analysis of LLMs' mathematical abilities, with a specific focus on their arithmetic performances. We identify numerical precision as a key factor that influences their effectiveness in mathematical tasks. Our results show that Transformers operating with low numerical precision fail to address arithmetic tasks, such as iterated addition and integer multiplication, unless the model size grows super-polynomially with respect to the input length. In contrast, Transformers with standard numerical precision can efficiently handle these tasks with significantly smaller model sizes. We further support our theoretical findings through empirical experiments that explore the impact of varying numerical precision on arithmetic tasks, providing valuable insights for improving the mathematical reasoning capabilities of LLMs.
☆ Can MLLMs Understand the Deep Implication Behind Chinese Images?
As the capabilities of Multimodal Large Language Models (MLLMs) continue to improve, the need for higher-order capability evaluation of MLLMs is increasing. However, there is a lack of work evaluating MLLM for higher-order perception and understanding of Chinese visual content. To fill the gap, we introduce the **C**hinese **I**mage **I**mplication understanding **Bench**mark, **CII-Bench**, which aims to assess the higher-order perception and understanding capabilities of MLLMs for Chinese images. CII-Bench stands out in several ways compared to existing benchmarks. Firstly, to ensure the authenticity of the Chinese context, images in CII-Bench are sourced from the Chinese Internet and manually reviewed, with corresponding answers also manually crafted. Additionally, CII-Bench incorporates images that represent Chinese traditional culture, such as famous Chinese traditional paintings, which can deeply reflect the model's understanding of Chinese traditional culture. Through extensive experiments on CII-Bench across multiple MLLMs, we have made significant findings. Initially, a substantial gap is observed between the performance of MLLMs and humans on CII-Bench. The highest accuracy of MLLMs attains 64.4%, where as human accuracy averages 78.2%, peaking at an impressive 81.0%. Subsequently, MLLMs perform worse on Chinese traditional culture images, suggesting limitations in their ability to understand high-level semantics and lack a deep knowledge base of Chinese traditional culture. Finally, it is observed that most models exhibit enhanced accuracy when image emotion hints are incorporated into the prompts. We believe that CII-Bench will enable MLLMs to gain a better understanding of Chinese semantics and Chinese-specific images, advancing the journey towards expert artificial general intelligence (AGI). Our project is publicly available at https://cii-bench.github.io/.
comment: 32 pages,18 figures. Project Page: https://cii-bench.github.io/ Code: https://github.com/MING_X/CII-Bench Dataset: https://huggingface.co/datasets/m-a-p/CII-Bench
☆ Retrospective Learning from Interactions
Multi-turn interactions between large language models (LLMs) and users naturally include implicit feedback signals. If an LLM responds in an unexpected way to an instruction, the user is likely to signal it by rephrasing the request, expressing frustration, or pivoting to an alternative task. Such signals are task-independent and occupy a relatively constrained subspace of language, allowing the LLM to identify them even if it fails on the actual task. This creates an avenue for continually learning from interactions without additional annotations. We introduce ReSpect, a method to learn from such signals in past interactions via retrospection. We deploy ReSpect in a new multimodal interaction scenario, where humans instruct an LLM to solve an abstract reasoning task with a combinatorial solution space. Through thousands of interactions with humans, we show how ReSpect gradually improves task completion rate from 31% to 82%, all without any external annotation.
☆ Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generation
In this paper, we introduce Janus, an autoregressive framework that unifies multimodal understanding and generation. Prior research often relies on a single visual encoder for both tasks, such as Chameleon. However, due to the differing levels of information granularity required by multimodal understanding and generation, this approach can lead to suboptimal performance, particularly in multimodal understanding. To address this issue, we decouple visual encoding into separate pathways, while still leveraging a single, unified transformer architecture for processing. The decoupling not only alleviates the conflict between the visual encoder's roles in understanding and generation, but also enhances the framework's flexibility. For instance, both the multimodal understanding and generation components can independently select their most suitable encoding methods. Experiments show that Janus surpasses previous unified model and matches or exceeds the performance of task-specific models. The simplicity, high flexibility, and effectiveness of Janus make it a strong candidate for next-generation unified multimodal models.
comment: Technical Report
☆ SimLayerKV: A Simple Framework for Layer-Level KV Cache Reduction
Recent advancements in large language models (LLMs) have extended their capabilities to handle long contexts. However, increasing the number of model layers and the length of input sequences significantly escalates the memory required to store key-value (KV) cache, posing challenges for efficient inference. To mitigate this issue, we present SimLayerKV, a simple yet effective method that reduces inter-layer KV cache redundancies by selectively dropping cache in identified lazy layers. Our approach is based on the observation that certain layers in long-context LLMs exhibit "lazy" behavior, contributing less to modeling long-range dependencies compared to non-lazy layers. By analyzing attention weight patterns, we find that the behavior of these lazy layers is consistent across tokens during generation for a given input. This insight motivates our SimLayerKV, which identifies lazy layers and reduces their KV cache accordingly. SimLayerKV is training-free, generalizable, and can be implemented with only seven lines of code. We conduct extensive experiments on three representative LLMs, e.g., LLaMA2-7B, LLaMA3-8B, and Mistral-7B across 16 tasks from the LongBench benchmark. The results demonstrate that SimLayerKV achieves a KV cache compression ratio of 5$\times$ with only a 1.2% performance drop when combined with 4-bit quantization. Our code is available at https://github.com/sail-sg/SimLayerKV.
☆ A Unified View of Delta Parameter Editing in Post-Trained Large-Scale Models
Post-training has emerged as a crucial paradigm for adapting large-scale pre-trained models to various tasks, whose effects are fully reflected by delta parameters (i.e., the disparity between post-trained and pre-trained parameters). While numerous studies have explored delta parameter properties via operations like pruning, quantization, low-rank approximation, and extrapolation, a unified framework for systematically examining these characteristics has been lacking. In this paper, we propose a novel perspective based on Riemann sum approximation of the loss function to elucidate delta parameter editing operations. Our analysis categorizes existing methods into three classes based on their post-editing performance: competitive, decreased, and improved, explaining how they are expressed by the Riemann sum approximation term and how they alter the model performance. Extensive experiments on both visual and language models, including ViT, LLaMA 3, Qwen 2, and Mistral, corroborate our theoretical findings. Furthermore, we introduce extensions to existing techniques like DARE and BitDelta, highlighting their limitations in leveraging the properties of delta parameters and reorganizing them into general expressions to enhance the applicability and effectiveness of delta parameter editing in post-trained models.
☆ A Common Pitfall of Margin-based Language Model Alignment: Gradient Entanglement
Reinforcement Learning from Human Feedback (RLHF) has become the predominant approach for language model (LM) alignment. At its core, RLHF uses a margin-based loss for preference optimization, specifying ideal LM behavior only by the difference between preferred and dispreferred responses. In this paper, we identify a common pitfall of margin-based methods -- the under-specification of ideal LM behavior on preferred and dispreferred responses individually, which leads to two unintended consequences as the margin increases: (1) The probability of dispreferred (e.g., unsafe) responses may increase, resulting in potential safety alignment failures. (2) The probability of preferred responses may decrease, even when those responses are ideal. We demystify the reasons behind these problematic behaviors: margin-based losses couple the change in the preferred probability to the gradient of the dispreferred one, and vice versa, often preventing the preferred probability from increasing while the dispreferred one decreases, and thus causing a synchronized increase or decrease in both probabilities. We term this effect, inherent in margin-based objectives, gradient entanglement. Formally, we derive conditions for general margin-based alignment objectives under which gradient entanglement becomes concerning: the inner product of the gradients of preferred and dispreferred log-probabilities is large relative to the individual gradient norms. We theoretically investigate why such inner products can be large when aligning language models and empirically validate our findings. Empirical implications of our framework extend to explaining important differences in the training dynamics of various preference optimization algorithms, and suggesting potential algorithm designs to mitigate the under-specification issue of margin-based methods and thereby improving language model alignment.
☆ AgentOccam: A Simple Yet Strong Baseline for LLM-Based Web Agents
Autonomy via agents using large language models (LLMs) for personalized, standardized tasks boosts human efficiency. Automating web tasks (like booking hotels within a budget) is increasingly sought after. Fulfilling practical needs, the web agent also serves as an important proof-of-concept example for various agent grounding scenarios, with its success promising advancements in many future applications. Prior research often handcrafts web agent strategies (e.g., prompting templates, multi-agent systems, search methods, etc.) and the corresponding in-context examples, which may not generalize well across all real-world scenarios. On the other hand, there has been limited study on the misalignment between a web agent's observation/action representation and the pre-training data of the LLM it's based on. This discrepancy is especially notable when LLMs are primarily trained for language completion rather than tasks involving embodied navigation actions and symbolic web elements. Our study enhances an LLM-based web agent by simply refining its observation and action space to better align with the LLM's capabilities. This approach enables our base agent to significantly outperform previous methods on a wide variety of web tasks. Specifically, on WebArena, a benchmark featuring general-purpose web interaction tasks, our agent AgentOccam surpasses the previous state-of-the-art and concurrent work by 9.8 (+29.4%) and 5.9 (+15.8%) absolute points respectively, and boosts the success rate by 26.6 points (+161%) over similar plain web agents with its observation and action space alignment. We achieve this without using in-context examples, new agent roles, online feedback or search strategies. AgentOccam's simple design highlights LLMs' impressive zero-shot performance on web tasks, and underlines the critical role of carefully tuning observation and action spaces for LLM-based agents.
☆ Harnessing Webpage UIs for Text-Rich Visual Understanding
Text-rich visual understanding-the ability to process environments where dense textual content is integrated with visuals-is crucial for multimodal large language models (MLLMs) to interact effectively with structured environments. To enhance this capability, we propose synthesizing general multimodal instructions from webpage UIs using text-based large language models (LLMs). Despite lacking direct visual input, text-based LLMs are able to process structured text representations from webpage accessibility trees. These instructions are then paired with UI screenshots to train multimodal models. We introduce MultiUI, a dataset containing 7.3 million samples from 1 million websites, covering diverse multimodal tasks and UI layouts. Models trained on MultiUI not only excel in web UI tasks-achieving up to a 48\% improvement on VisualWebBench and a 19.1\% boost in action accuracy on a web agent dataset Mind2Web-but also generalize surprisingly well to non-web UI tasks and even to non-UI domains, such as document understanding, OCR, and chart interpretation. These results highlight the broad applicability of web UI data for advancing text-rich visual understanding across various scenarios.
☆ De-mark: Watermark Removal in Large Language Models
Watermarking techniques offer a promising way to identify machine-generated content via embedding covert information into the contents generated from language models (LMs). However, the robustness of the watermarking schemes has not been well explored. In this paper, we present De-mark, an advanced framework designed to remove n-gram-based watermarks effectively. Our method utilizes a novel querying strategy, termed random selection probing, which aids in assessing the strength of the watermark and identifying the red-green list within the n-gram watermark. Experiments on popular LMs, such as Llama3 and ChatGPT, demonstrate the efficiency and effectiveness of De-mark in watermark removal and exploitation tasks.
☆ A Watermark for Order-Agnostic Language Models
Statistical watermarking techniques are well-established for sequentially decoded language models (LMs). However, these techniques cannot be directly applied to order-agnostic LMs, as the tokens in order-agnostic LMs are not generated sequentially. In this work, we introduce Pattern-mark, a pattern-based watermarking framework specifically designed for order-agnostic LMs. We develop a Markov-chain-based watermark generator that produces watermark key sequences with high-frequency key patterns. Correspondingly, we propose a statistical pattern-based detection algorithm that recovers the key sequence during detection and conducts statistical tests based on the count of high-frequency patterns. Our extensive evaluations on order-agnostic LMs, such as ProteinMPNN and CMLM, demonstrate Pattern-mark's enhanced detection efficiency, generation quality, and robustness, positioning it as a superior watermarking technique for order-agnostic LMs.
☆ BenTo: Benchmark Task Reduction with In-Context Transferability
Evaluating large language models (LLMs) is costly: it requires the generation and examination of LLM outputs on a large-scale benchmark of various tasks. This paper investigates how to efficiently reduce the tasks used to benchmark LLMs without affecting the evaluation quality. Our study reveals that task transferability and relevance provide critical information to identify the most representative subset of tasks via optimizing a facility location function. We propose a practically efficient metric for estimating the transferability between two tasks via in-context learning (ICL). By analyzing the pairwise transferability, we can reduce tasks in a modern LLM benchmark (e.g., MMLU or FLAN) to 5% while inducing only a <4% difference to the evaluation on the original benchmark. Compared to prior works, our method is training-free, gradient-free, and highly efficient requiring ICL only.
☆ Modeling Future Conversation Turns to Teach LLMs to Ask Clarifying Questions
Large language models (LLMs) must often respond to highly ambiguous user requests. In such cases, the LLM's best response may be to ask a clarifying question to elicit more information. We observe existing LLMs often respond by presupposing a single interpretation of such ambiguous requests, frustrating users who intended a different interpretation. We speculate this is caused by current preference data labeling practice, where LLM responses are evaluated only on their prior contexts. To address this, we propose to assign preference labels by simulating their expected outcomes in the future turns. This allows LLMs to learn to ask clarifying questions when it can generate responses that are tailored to each user interpretation in future turns. In experiments on open-domain QA, we compare systems that trained using our proposed preference labeling methods against standard methods, which assign preferences based on only prior context. We evaluate systems based on their ability to ask clarifying questions that can recover each user's interpretation and expected answer, and find that our training with our proposed method trains LLMs to ask clarifying questions with a 5% improvement in F1 measured against the answer set from different interpretations of each query
☆ Looking Inward: Language Models Can Learn About Themselves by Introspection
Humans acquire knowledge by observing the external world, but also by introspection. Introspection gives a person privileged access to their current state of mind (e.g., thoughts and feelings) that is not accessible to external observers. Can LLMs introspect? We define introspection as acquiring knowledge that is not contained in or derived from training data but instead originates from internal states. Such a capability could enhance model interpretability. Instead of painstakingly analyzing a model's internal workings, we could simply ask the model about its beliefs, world models, and goals. More speculatively, an introspective model might self-report on whether it possesses certain internal states such as subjective feelings or desires and this could inform us about the moral status of these states. Such self-reports would not be entirely dictated by the model's training data. We study introspection by finetuning LLMs to predict properties of their own behavior in hypothetical scenarios. For example, "Given the input P, would your output favor the short- or long-term option?" If a model M1 can introspect, it should outperform a different model M2 in predicting M1's behavior even if M2 is trained on M1's ground-truth behavior. The idea is that M1 has privileged access to its own behavioral tendencies, and this enables it to predict itself better than M2 (even if M2 is generally stronger). In experiments with GPT-4, GPT-4o, and Llama-3 models (each finetuned to predict itself), we find that the model M1 outperforms M2 in predicting itself, providing evidence for introspection. Notably, M1 continues to predict its behavior accurately even after we intentionally modify its ground-truth behavior. However, while we successfully elicit introspection on simple tasks, we are unsuccessful on more complex tasks or those requiring out-of-distribution generalization.
comment: 15 pages, 9 figures
☆ PopAlign: Diversifying Contrasting Patterns for a More Comprehensive Alignment
Alignment of large language models (LLMs) involves training models on preference-contrastive output pairs to adjust their responses according to human preferences. To obtain such contrastive pairs, traditional methods like RLHF and RLAIF rely on limited contrasting patterns, such as varying model variants or decoding temperatures. This singularity leads to two issues: (1) alignment is not comprehensive; and thereby (2) models are susceptible to jailbreaking attacks. To address these issues, we investigate how to construct more comprehensive and diversified contrasting patterns to enhance preference data (RQ1) and verify the impact of the diversification of contrasting patterns on model alignment (RQ2). For RQ1, we propose PopAlign, a framework that integrates diversified contrasting patterns across the prompt, model, and pipeline levels, introducing six contrasting strategies that do not require additional feedback labeling procedures. Regarding RQ2, we conduct thorough experiments demonstrating that PopAlign significantly outperforms existing methods, leading to more comprehensive alignment.
comment: 28 pages
☆ Quantity vs. Quality of Monolingual Source Data in Automatic Text Translation: Can It Be Too Little If It Is Too Good?
Monolingual data, being readily available in large quantities, has been used to upscale the scarcely available parallel data to train better models for automatic translation. Self-learning, where a model is made to learn from its output, is one approach to exploit such data. However, it has been shown that too much of this data can be detrimental to the performance of the model if the available parallel data is comparatively extremely low. In this study, we investigate whether the monolingual data can also be too little and if this reduction, based on quality, has any effect on the performance of the translation model. Experiments have shown that on English-German low-resource NMT, it is often better to select only the most useful additional data, based on quality or closeness to the domain of the test data, than utilizing all of the available data.
☆ Optimal Quantization for Matrix Multiplication
Recent work in machine learning community proposed multiple methods for performing lossy compression (quantization) of large matrices. This quantization is important for accelerating matrix multiplication (main component of large language models), which is often bottlenecked by the speed of loading these matrices from memory. Unlike classical vector quantization and rate-distortion theory, the goal of these new compression algorithms is to be able to approximate not the matrices themselves, but their matrix product. Specifically, given a pair of real matrices $A,B$ an encoder (compressor) is applied to each of them independently producing descriptions with $R$ bits per entry. These representations subsequently are used by the decoder to estimate matrix product $A^\top B$. In this work, we provide a non-asymptotic lower bound on the mean squared error of this approximation (as a function of rate $R$) for the case of matrices $A,B$ with iid Gaussian entries. Algorithmically, we construct a universal quantizer based on nested lattices with an explicit guarantee of approximation error for any (non-random) pair of matrices $A$, $B$ in terms of only Frobenius norms $\|A\|_F, \|B\|_F$ and $\|A^\top B\|_F$. For iid Gaussian matrices our quantizer achieves the lower bound and is, thus, asymptotically optimal. A practical low-complexity version of our quantizer achieves performance quite close to optimal. In information-theoretic terms we derive rate-distortion function for matrix multiplication of iid Gaussian matrices.
☆ The Mystery of the Pathological Path-star Task for Language Models EMNLP 2024
The recently introduced path-star task is a minimal task designed to exemplify limitations to the abilities of language models (Bachmann and Nagarajan, 2024). It involves a path-star graph where multiple arms radiate from a single starting node and each node is unique. Given the start node and a specified target node that ends an arm, the task is to generate the arm containing that target node. This is straightforward for a human but surprisingly difficult for language models, which did not outperform the random baseline. The authors hypothesized this is due to a deficiency in teacher-forcing and the next-token prediction paradigm. We demonstrate the task is learnable using teacher-forcing in alternative settings and that the issue is partially due to representation. We introduce a regularization method using structured samples of the same graph but with differing target nodes, improving results across a variety of model types. We provide RASP proofs showing the task is theoretically solvable. Finally, we find settings where an encoder-only model can consistently solve the task.
comment: EMNLP 2024 Main
☆ Aggregation Artifacts in Subjective Tasks Collapse Large Language Models' Posteriors
In-context Learning (ICL) has become the primary method for performing natural language tasks with Large Language Models (LLMs). The knowledge acquired during pre-training is crucial for this few-shot capability, providing the model with task priors. However, recent studies have shown that ICL predominantly relies on retrieving task priors rather than "learning" to perform tasks. This limitation is particularly evident in complex subjective domains such as emotion and morality, where priors significantly influence posterior predictions. In this work, we examine whether this is the result of the aggregation used in corresponding datasets, where trying to combine low-agreement, disparate annotations might lead to annotation artifacts that create detrimental noise in the prompt. Moreover, we evaluate the posterior bias towards certain annotators by grounding our study in appropriate, quantitative measures of LLM priors. Our results indicate that aggregation is a confounding factor in the modeling of subjective tasks, and advocate focusing on modeling individuals instead. However, aggregation does not explain the entire gap between ICL and the state of the art, meaning other factors in such tasks also account for the observed phenomena. Finally, by rigorously studying annotator-level labels, we find that it is possible for minority annotators to both better align with LLMs and have their perspectives further amplified.
comment: 12 pages, 7 figures, 2 tables
☆ Knowledge-Aware Query Expansion with Large Language Models for Textual and Relational Retrieval
Large language models (LLMs) have been used to generate query expansions augmenting original queries for improving information search. Recent studies also explore providing LLMs with initial retrieval results to generate query expansions more grounded to document corpus. However, these methods mostly focus on enhancing textual similarities between search queries and target documents, overlooking document relations. For queries like "Find me a highly rated camera for wildlife photography compatible with my Nikon F-Mount lenses", existing methods may generate expansions that are semantically similar but structurally unrelated to user intents. To handle such semi-structured queries with both textual and relational requirements, in this paper we propose a knowledge-aware query expansion framework, augmenting LLMs with structured document relations from knowledge graph (KG). To further address the limitation of entity-based scoring in existing KG-based methods, we leverage document texts as rich KG node representations and use document-based relation filtering for our Knowledge-Aware Retrieval (KAR). Extensive experiments on three datasets of diverse domains show the advantages of our method compared against state-of-the-art baselines on textual and relational semi-structured retrieval.
☆ MobA: A Two-Level Agent System for Efficient Mobile Task Automation
Current mobile assistants are limited by dependence on system APIs or struggle with complex user instructions and diverse interfaces due to restricted comprehension and decision-making abilities. To address these challenges, we propose MobA, a novel Mobile phone Agent powered by multimodal large language models that enhances comprehension and planning capabilities through a sophisticated two-level agent architecture. The high-level Global Agent (GA) is responsible for understanding user commands, tracking history memories, and planning tasks. The low-level Local Agent (LA) predicts detailed actions in the form of function calls, guided by sub-tasks and memory from the GA. Integrating a Reflection Module allows for efficient task completion and enables the system to handle previously unseen complex tasks. MobA demonstrates significant improvements in task execution efficiency and completion rate in real-life evaluations, underscoring the potential of MLLM-empowered mobile assistants.
comment: 27 pages, 6 figures, and 5 tables. We will release our source code in a few days
☆ LLM-Human Pipeline for Cultural Context Grounding of Conversations
Conversations often adhere to well-understood social norms that vary across cultures. For example, while "addressing parents by name" is commonplace in the West, it is rare in most Asian cultures. Adherence or violation of such norms often dictates the tenor of conversations. Humans are able to navigate social situations requiring cultural awareness quite adeptly. However, it is a hard task for NLP models. In this paper, we tackle this problem by introducing a "Cultural Context Schema" for conversations. It comprises (1) conversational information such as emotions, dialogue acts, etc., and (2) cultural information such as social norms, violations, etc. We generate ~110k social norm and violation descriptions for ~23k conversations from Chinese culture using LLMs. We refine them using automated verification strategies which are evaluated against culturally aware human judgements. We organize these descriptions into meaningful structures we call "Norm Concepts", using an interactive human-in-loop framework. We ground the norm concepts and the descriptions in conversations using symbolic annotation. Finally, we use the obtained dataset for downstream tasks such as emotion, sentiment, and dialogue act detection. We show that it significantly improves the empirical performance.
comment: 19 pages, 9 figures, 7 tables
☆ MIRAGE-Bench: Automatic Multilingual Benchmark Arena for Retrieval-Augmented Generation Systems
Traditional Retrieval-Augmented Generation (RAG) benchmarks rely on different heuristic-based metrics for evaluation, but these require human preferences as ground truth for reference. In contrast, arena-based benchmarks, where two models compete each other, require an expensive Large Language Model (LLM) as a judge for a reliable evaluation. We present an easy and efficient technique to get the best of both worlds. The idea is to train a learning to rank model as a "surrogate" judge using RAG-based evaluation heuristics as input, to produce a synthetic arena-based leaderboard. Using this idea, We develop MIRAGE-Bench, a standardized arena-based multilingual RAG benchmark for 18 diverse languages on Wikipedia. The benchmark is constructed using MIRACL, a retrieval dataset, and extended for multilingual generation evaluation. MIRAGE-Bench evaluates RAG extensively coupling both heuristic features and LLM as a judge evaluator. In our work, we benchmark 19 diverse multilingual-focused LLMs, and achieve a high correlation (Kendall Tau ($\tau$) = 0.909) using our surrogate judge learned using heuristic features with pairwise evaluations and between GPT-4o as a teacher on the MIRAGE-Bench leaderboard using the Bradley-Terry framework. We observe proprietary and large open-source LLMs currently dominate in multilingual RAG. MIRAGE-Bench is available at: https://github.com/vectara/mirage-bench.
☆ On the Role of Attention Heads in Large Language Model Safety
Large language models (LLMs) achieve state-of-the-art performance on multiple language tasks, yet their safety guardrails can be circumvented, leading to harmful generations. In light of this, recent research on safety mechanisms has emerged, revealing that when safety representations or component are suppressed, the safety capability of LLMs are compromised. However, existing research tends to overlook the safety impact of multi-head attention mechanisms, despite their crucial role in various model functionalities. Hence, in this paper, we aim to explore the connection between standard attention mechanisms and safety capability to fill this gap in the safety-related mechanistic interpretability. We propose a novel metric which tailored for multi-head attention, the Safety Head ImPortant Score (Ships), to assess the individual heads' contributions to model safety. Based on this, we generalize Ships to the dataset level and further introduce the Safety Attention Head AttRibution Algorithm (Sahara) to attribute the critical safety attention heads inside the model. Our findings show that the special attention head has a significant impact on safety. Ablating a single safety head allows aligned model (e.g., Llama-2-7b-chat) to respond to 16 times more harmful queries, while only modifying 0.006% of the parameters, in contrast to the ~ 5% modification required in previous studies. More importantly, we demonstrate that attention heads primarily function as feature extractors for safety and models fine-tuned from the same base model exhibit overlapping safety heads through comprehensive experiments. Together, our attribution approach and findings provide a novel perspective for unpacking the black box of safety mechanisms within large models.
comment: 28 pages, 18 figures, 7 tables
☆ Unconstrained Model Merging for Enhanced LLM Reasoning
Recent advancements in building domain-specific large language models (LLMs) have shown remarkable success, especially in tasks requiring reasoning abilities like logical inference over complex relationships and multi-step problem solving. However, creating a powerful all-in-one LLM remains challenging due to the need for proprietary data and vast computational resources. As a resource-friendly alternative, we explore the potential of merging multiple expert models into a single LLM. Existing studies on model merging mainly focus on generalist LLMs instead of domain experts, or the LLMs under the same architecture and size. In this work, we propose an unconstrained model merging framework that accommodates both homogeneous and heterogeneous model architectures with a focus on reasoning tasks. A fine-grained layer-wise weight merging strategy is designed for homogeneous models merging, while heterogeneous model merging is built upon the probabilistic distribution knowledge derived from instruction-response fine-tuning data. Across 7 benchmarks and 9 reasoning-optimized LLMs, we reveal key findings that combinatorial reasoning emerges from merging which surpasses simple additive effects. We propose that unconstrained model merging could serve as a foundation for decentralized LLMs, marking a notable progression from the existing centralized LLM framework. This evolution could enhance wider participation and stimulate additional advancement in the field of artificial intelligence, effectively addressing the constraints posed by centralized models.
comment: Under review
☆ Exploring the Design Space of Visual Context Representation in Video MLLMs
Video Multimodal Large Language Models (MLLMs) have shown remarkable capability of understanding the video semantics on various downstream tasks. Despite the advancements, there is still a lack of systematic research on visual context representation, which refers to the scheme to select frames from a video and further select the tokens from a frame. In this paper, we explore the design space for visual context representation, and aim to improve the performance of video MLLMs by finding more effective representation schemes. Firstly, we formulate the task of visual context representation as a constrained optimization problem, and model the language modeling loss as a function of the number of frames and the number of embeddings (or tokens) per frame, given the maximum visual context window size. Then, we explore the scaling effects in frame selection and token selection respectively, and fit the corresponding function curve by conducting extensive empirical experiments. We examine the effectiveness of typical selection strategies and present empirical findings to determine the two factors. Furthermore, we study the joint effect of frame selection and token selection, and derive the optimal formula for determining the two factors. We demonstrate that the derived optimal settings show alignment with the best-performed results of empirical experiments. Our code and model are available at: https://github.com/RUCAIBox/Opt-Visor.
comment: Long Video MLLM; work in progress
☆ Pose-Based Sign Language Appearance Transfer
We introduce a method for transferring the signer's appearance in sign language skeletal poses while preserving the sign content. Using estimated poses, we transfer the appearance of one signer to another, maintaining natural movements and transitions. This approach improves pose-based rendering and sign stitching while obfuscating identity. Our experiments show that while the method reduces signer identification accuracy, it slightly harms sign recognition performance, highlighting a tradeoff between privacy and utility. Our code is available at \url{https://github.com/sign-language-processing/pose-anonymization}.
☆ HEALTH-PARIKSHA: Assessing RAG Models for Health Chatbots in Real-World Multilingual Settings
Assessing the capabilities and limitations of large language models (LLMs) has garnered significant interest, yet the evaluation of multiple models in real-world scenarios remains rare. Multilingual evaluation often relies on translated benchmarks, which typically do not capture linguistic and cultural nuances present in the source language. This study provides an extensive assessment of 24 LLMs on real world data collected from Indian patients interacting with a medical chatbot in Indian English and 4 other Indic languages. We employ a uniform Retrieval Augmented Generation framework to generate responses, which are evaluated using both automated techniques and human evaluators on four specific metrics relevant to our application. We find that models vary significantly in their performance and that instruction tuned Indic models do not always perform well on Indic language queries. Further, we empirically show that factual correctness is generally lower for responses to Indic queries compared to English queries. Finally, our qualitative work shows that code-mixed and culturally relevant queries in our dataset pose challenges to evaluated models.
comment: Under Review
☆ signwriting-evaluation: Effective Sign Language Evaluation via SignWriting
The lack of automatic evaluation metrics tailored for SignWriting presents a significant obstacle in developing effective transcription and translation models for signed languages. This paper introduces a comprehensive suite of evaluation metrics specifically designed for SignWriting, including adaptations of standard metrics such as \texttt{BLEU} and \texttt{chrF}, the application of \texttt{CLIPScore} to SignWriting images, and a novel symbol distance metric unique to our approach. We address the distinct challenges of evaluating single signs versus continuous signing and provide qualitative demonstrations of metric efficacy through score distribution analyses and nearest-neighbor searches within the SignBank corpus. Our findings reveal the strengths and limitations of each metric, offering valuable insights for future advancements using SignWriting. This work contributes essential tools for evaluating SignWriting models, facilitating progress in the field of sign language processing. Our code is available at \url{https://github.com/sign-language-processing/signwriting-evaluation}.
☆ ORCHID: A Chinese Debate Corpus for Target-Independent Stance Detection and Argumentative Dialogue Summarization EMNLP 2023
Dialogue agents have been receiving increasing attention for years, and this trend has been further boosted by the recent progress of large language models (LLMs). Stance detection and dialogue summarization are two core tasks of dialogue agents in application scenarios that involve argumentative dialogues. However, research on these tasks is limited by the insufficiency of public datasets, especially for non-English languages. To address this language resource gap in Chinese, we present ORCHID (Oral Chinese Debate), the first Chinese dataset for benchmarking target-independent stance detection and debate summarization. Our dataset consists of 1,218 real-world debates that were conducted in Chinese on 476 unique topics, containing 2,436 stance-specific summaries and 14,133 fully annotated utterances. Besides providing a versatile testbed for future research, we also conduct an empirical study on the dataset and propose an integrated task. The results show the challenging nature of the dataset and suggest a potential of incorporating stance detection in summarization for argumentative dialogue.
comment: In EMNLP 2023
☆ VL-GLUE: A Suite of Fundamental yet Challenging Visuo-Linguistic Reasoning Tasks
Deriving inference from heterogeneous inputs (such as images, text, and audio) is an important skill for humans to perform day-to-day tasks. A similar ability is desirable for the development of advanced Artificial Intelligence (AI) systems. While state-of-the-art models are rapidly closing the gap with human-level performance on diverse computer vision and NLP tasks separately, they struggle to solve tasks that require joint reasoning over visual and textual modalities. Inspired by GLUE (Wang et. al., 2018)- a multitask benchmark for natural language understanding, we propose VL-GLUE in this paper. VL-GLUE consists of over 100k samples spanned across seven different tasks, which at their core require visuo-linguistic reasoning. Moreover, our benchmark comprises of diverse image types (from synthetically rendered figures, and day-to-day scenes to charts and complex diagrams) and includes a broad variety of domain-specific text (from cooking, politics, and sports to high-school curricula), demonstrating the need for multi-modal understanding in the real-world. We show that this benchmark is quite challenging for existing large-scale vision-language models and encourage development of systems that possess robust visuo-linguistic reasoning capabilities.
comment: 18 pages, 7 figures
☆ Red and blue language: Word choices in the Trump & Harris 2024 presidential debate
Political debates are a peculiar type of political discourse, in which candidates directly confront one another, addressing not only the the moderator's questions, but also their opponent's statements, as well as the concerns of voters from both parties and undecided voters. Therefore, language is adjusted to meet specific expectations and achieve persuasion. We analyse how the language of Trump and Harris during the debate (September 10th 2024) differs in relation to the following semantic and pragmatic features, for which we formulated targeted hypotheses: framing values and ideology, appealing to emotion, using words with different degrees of concreteness and specificity, addressing others through singular or plural pronouns. Our findings include: differences in the use of figurative frames (Harris often framing issues around recovery and empowerment, Trump often focused on crisis and decline); similar use of emotional language, with Trump showing a slight higher tendency toward negativity and toward less subjective language compared to Harris; no significant difference in the specificity of candidates' responses; similar use of abstract language, with Trump showing more variability than Harris, depending on the subject discussed; differences in addressing the opponent, with Trump not mentioning Harris by name, while Harris referring to Trump frequently; different uses of pronouns, with Harris using both singular and plural pronouns equally, while Trump using more singular pronouns. The results are discussed in relation to previous literature on Red and Blue language, which refers to distinct linguistic patterns associated with conservative (Red) and liberal (Blue) political ideologies.
comment: Submitted to PLOS ONE, under review
☆ A new approach for fine-tuning sentence transformers for intent classification and out-of-scope detection tasks
In virtual assistant (VA) systems it is important to reject or redirect user queries that fall outside the scope of the system. One of the most accurate approaches for out-of-scope (OOS) rejection is to combine it with the task of intent classification on in-scope queries, and to use methods based on the similarity of embeddings produced by transformer-based sentence encoders. Typically, such encoders are fine-tuned for the intent-classification task, using cross-entropy loss. Recent work has shown that while this produces suitable embeddings for the intent-classification task, it also tends to disperse in-scope embeddings over the full sentence embedding space. This causes the in-scope embeddings to potentially overlap with OOS embeddings, thereby making OOS rejection difficult. This is compounded when OOS data is unknown. To mitigate this issue our work proposes to regularize the cross-entropy loss with an in-scope embedding reconstruction loss learned using an auto-encoder. Our method achieves a 1-4% improvement in the area under the precision-recall curve for rejecting out-of-sample (OOS) instances, without compromising intent classification performance.
comment: Appearing at Empirical Methods in Natural Language Processing 2025 - Industry Track
☆ SimpleToM: Exposing the Gap between Explicit ToM Inference and Implicit ToM Application in LLMs
While prior work has explored whether large language models (LLMs) possess a "theory of mind" (ToM) - the ability to attribute mental states to oneself and others - there has been little work testing whether LLMs can implicitly apply such knowledge to predict behavior, or to judge whether an observed behavior is rational. Such skills are critical for appropriate interaction in social environments. We create a new dataset, SimpleTom, containing concise, diverse stories (e.g., "The can of Pringles has moldy chips in it. Mary picks up the can in the supermarket and walks to the cashier."), each with three questions that test different degrees of ToM reasoning, asking models to predict (a) mental state ("Is Mary aware of the mold?"), (b) behavior ("Will Mary pay for the chips or report the mold?"), and (c) judgment ("Mary paid for the chips. Was that reasonable?"). To our knowledge, SimpleToM is the first dataset to systematically explore downstream reasoning requiring knowledge of mental states in realistic scenarios. Our experimental results are intriguing: While most models can reliably predict mental state on our dataset (a), they often fail to correctly predict the behavior (b), and fare even worse at judging whether given behaviors are reasonable (c), despite being correctly aware of the protagonist's mental state should make such secondary predictions obvious. We further show that we can help models do better at (b) and (c) via interventions such as reminding the model of its earlier mental state answer and mental-state-specific chain-of-thought prompting, raising the action prediction accuracies (e.g., from 49.5% to 93.5% for GPT-4o) and judgment accuracies (e.g., from 15.3% to 94.7% in GPT-4o). While this shows that models can be coaxed to perform well, it requires task-specific interventions, and the natural model performances remain low, a cautionary tale for LLM deployment.
☆ An Active Learning Framework for Inclusive Generation by Large Language Models
Ensuring that Large Language Models (LLMs) generate text representative of diverse sub-populations is essential, particularly when key concepts related to under-represented groups are scarce in the training data. We address this challenge with a novel clustering-based active learning framework, enhanced with knowledge distillation. The proposed framework transforms the intermediate outputs of the learner model, enabling effective active learning for generative tasks for the first time. Integration of clustering and knowledge distillation yields more representative models without prior knowledge of underlying data distribution and overbearing human efforts. We validate our approach in practice through case studies in counter-narration and style transfer. We construct two new datasets in tandem with model training, showing a performance improvement of 2%-10% over baseline models. Our results also show more consistent performance across various data subgroups and increased lexical diversity, underscoring our model's resilience to skewness in available data. Further, our results show that the data acquired via our approach improves the performance of secondary models not involved in the learning loop, showcasing practical utility of the framework.
☆ Latent Space Chain-of-Embedding Enables Output-free LLM Self-Evaluation
LLM self-evaluation relies on the LLM's own ability to estimate response correctness, which can greatly improve its deployment reliability. In this research track, we propose the Chain-of-Embedding (CoE) in the latent space to enable LLMs to perform output-free self-evaluation. CoE consists of all progressive hidden states produced during the inference time, which can be treated as the latent thinking path of LLMs. We find that when LLMs respond correctly and incorrectly, their CoE features differ, these discrepancies assist us in estimating LLM response correctness. Experiments in four diverse domains and seven LLMs fully demonstrate the effectiveness of our method. Meanwhile, its label-free design intent without any training and millisecond-level computational cost ensure real-time feedback in large-scale scenarios. More importantly, we provide interesting insights into LLM response correctness from the perspective of hidden state changes inside LLMs.
comment: 33 pages, 18 figures, 12 tables
☆ A Comparative Study on Reasoning Patterns of OpenAI's o1 Model
Enabling Large Language Models (LLMs) to handle a wider range of complex tasks (e.g., coding, math) has drawn great attention from many researchers. As LLMs continue to evolve, merely increasing the number of model parameters yields diminishing performance improvements and heavy computational costs. Recently, OpenAI's o1 model has shown that inference strategies (i.e., Test-time Compute methods) can also significantly enhance the reasoning capabilities of LLMs. However, the mechanisms behind these methods are still unexplored. In our work, to investigate the reasoning patterns of o1, we compare o1 with existing Test-time Compute methods (BoN, Step-wise BoN, Agent Workflow, and Self-Refine) by using OpenAI's GPT-4o as a backbone on general reasoning benchmarks in three domains (i.e., math, coding, commonsense reasoning). Specifically, first, our experiments show that the o1 model has achieved the best performance on most datasets. Second, as for the methods of searching diverse responses (e.g., BoN), we find the reward models' capability and the search space both limit the upper boundary of these methods. Third, as for the methods that break the problem into many sub-problems, the Agent Workflow has achieved better performance than Step-wise BoN due to the domain-specific system prompt for planning better reasoning processes. Fourth, it is worth mentioning that we have summarized six reasoning patterns of o1, and provided a detailed analysis on several reasoning benchmarks.
☆ H2OVL-Mississippi Vision Language Models Technical Report
Smaller vision-language models (VLMs) are becoming increasingly important for privacy-focused, on-device applications due to their ability to run efficiently on consumer hardware for processing enterprise commercial documents and images. These models require strong language understanding and visual capabilities to enhance human-machine interaction. To address this need, we present H2OVL-Mississippi, a pair of small VLMs trained on 37 million image-text pairs using 240 hours of compute on 8 x H100 GPUs. H2OVL-Mississippi-0.8B is a tiny model with 0.8 billion parameters that specializes in text recognition, achieving state of the art performance on the Text Recognition portion of OCRBench and surpassing much larger models in this area. Additionally, we are releasing H2OVL-Mississippi-2B, a 2 billion parameter model for general use cases, exhibiting highly competitive metrics across various academic benchmarks. Both models build upon our prior work with H2O-Danube language models, extending their capabilities into the visual domain. We release them under the Apache 2.0 license, making VLMs accessible to everyone, democratizing document AI and visual LLMs.
☆ MeNTi: Bridging Medical Calculator and LLM Agent with Nested Tool Calling
Integrating tools into Large Language Models (LLMs) has facilitated the widespread application. Despite this, in specialized downstream task contexts, reliance solely on tools is insufficient to fully address the complexities of the real world. This particularly restricts the effective deployment of LLMs in fields such as medicine. In this paper, we focus on the downstream tasks of medical calculators, which use standardized tests to assess an individual's health status. We introduce MeNTi, a universal agent architecture for LLMs. MeNTi integrates a specialized medical toolkit and employs meta-tool and nested calling mechanisms to enhance LLM tool utilization. Specifically, it achieves flexible tool selection and nested tool calling to address practical issues faced in intricate medical scenarios, including calculator selection, slot filling, and unit conversion. To assess the capabilities of LLMs for quantitative assessment throughout the clinical process of calculator scenarios, we introduce CalcQA. This benchmark requires LLMs to use medical calculators to perform calculations and assess patient health status. CalcQA is constructed by professional physicians and includes 100 case-calculator pairs, complemented by a toolkit of 281 medical tools. The experimental results demonstrate significant performance improvements with our framework. This research paves new directions for applying LLMs in demanding scenarios of medicine.
☆ Large Language Models as Narrative-Driven Recommenders
Narrative-driven recommenders aim to provide personalized suggestions for user requests expressed in free-form text such as "I want to watch a thriller with a mind-bending story, like Shutter Island." Although large language models (LLMs) have been shown to excel in processing general natural language queries, their effectiveness for handling such recommendation requests remains relatively unexplored. To close this gap, we compare the performance of 38 open- and closed-source LLMs of various sizes, such as LLama 3.2 and GPT-4o, in a movie recommendation setting. For this, we utilize a gold-standard, crowdworker-annotated dataset of posts from reddit's movie suggestion community and employ various prompting strategies, including zero-shot, identity, and few-shot prompting. Our findings demonstrate the ability of LLMs to generate contextually relevant movie recommendations, significantly outperforming other state-of-the-art approaches, such as doc2vec. While we find that closed-source and large-parameterized models generally perform best, medium-sized open-source models remain competitive, being only slightly outperformed by their more computationally expensive counterparts. Furthermore, we observe no significant differences across prompting strategies for most models, underscoring the effectiveness of simple approaches such as zero-shot prompting for narrative-driven recommendations. Overall, this work offers valuable insights for recommender system researchers as well as practitioners aiming to integrate LLMs into real-world recommendation tools.
comment: Under review; 19 pages
☆ Enhancing Fact Retrieval in PLMs through Truthfulness
Pre-trained Language Models (PLMs) encode various facts about the world at their pre-training phase as they are trained to predict the next or missing word in a sentence. There has a been an interest in quantifying and improving the amount of facts that can be extracted from PLMs, as they have been envisioned to act as soft knowledge bases, which can be queried in natural language. Different approaches exist to enhance fact retrieval from PLM. Recent work shows that the hidden states of PLMs can be leveraged to determine the truthfulness of the PLMs' inputs. Leveraging this finding to improve factual knowledge retrieval remains unexplored. In this work, we investigate the use of a helper model to improve fact retrieval. The helper model assesses the truthfulness of an input based on the corresponding hidden states representations from the PLMs. We evaluate this approach on several masked PLMs and show that it enhances fact retrieval by up to 33\%. Our findings highlight the potential of hidden states representations from PLMs in improving their factual knowledge retrieval.
☆ Integrating Temporal Representations for Dynamic Memory Retrieval and Management in Large Language Models
Conventional dialogue agents often struggle with effective memory recall, leading to redundant retrieval and inadequate management of unique user associations. To address this, we propose SynapticRAG, a novel approach integrating synaptic dynamics into Retrieval-Augmented Generation (RAG). SynapticRAG integrates temporal representations into memory vectors, mimicking biological synapses by differentiating events based on occurrence times and dynamically updating memory significance. This model employs temporal scoring for memory connections and a synaptic-inspired propagation control mechanism. Experiments across English, Japanese, and Chinese datasets demonstrate SynapticRAG's superiority over existing methods, including traditional RAG, with up to 14.66\% improvement in memory retrieval accuracy. Our approach advances context-aware dialogue AI systems by enhancing long-term context maintenance and specific information extraction from conversations.
☆ Bias in the Mirror : Are LLMs opinions robust to their own adversarial attacks ?
Large language models (LLMs) inherit biases from their training data and alignment processes, influencing their responses in subtle ways. While many studies have examined these biases, little work has explored their robustness during interactions. In this paper, we introduce a novel approach where two instances of an LLM engage in self-debate, arguing opposing viewpoints to persuade a neutral version of the model. Through this, we evaluate how firmly biases hold and whether models are susceptible to reinforcing misinformation or shifting to harmful viewpoints. Our experiments span multiple LLMs of varying sizes, origins, and languages, providing deeper insights into bias persistence and flexibility across linguistic and cultural contexts.
☆ GeoCoder: Solving Geometry Problems by Generating Modular Code through Vision-Language Models
Geometry problem-solving demands advanced reasoning abilities to process multimodal inputs and employ mathematical knowledge effectively. Vision-language models (VLMs) have made significant progress in various multimodal tasks. Yet, they still struggle with geometry problems and are significantly limited by their inability to perform mathematical operations not seen during pre-training, such as calculating the cosine of an arbitrary angle, and by difficulties in correctly applying relevant geometry formulas. To overcome these challenges, we present GeoCoder, which leverages modular code-finetuning to generate and execute code using a predefined geometry function library. By executing the code, we achieve accurate and deterministic calculations, contrasting the stochastic nature of autoregressive token prediction, while the function library minimizes errors in formula usage. We also propose a multimodal retrieval-augmented variant of GeoCoder, named RAG-GeoCoder, which incorporates a non-parametric memory module for retrieving functions from the geometry library, thereby reducing reliance on parametric memory. Our modular code-finetuning approach enhances the geometric reasoning capabilities of VLMs, yielding an average improvement of over 16% across various question complexities on the GeomVerse dataset compared to other finetuning methods.
☆ RAG-DDR: Optimizing Retrieval-Augmented Generation Using Differentiable Data Rewards
Retrieval-Augmented Generation (RAG) has proven its effectiveness in mitigating hallucinations in Large Language Models (LLMs) by retrieving knowledge from external resources. To adapt LLMs for RAG pipelines, current approaches use instruction tuning to optimize LLMs, improving their ability to utilize retrieved knowledge. This supervised fine-tuning (SFT) approach focuses on equipping LLMs to handle diverse RAG tasks using different instructions. However, it trains RAG modules to overfit training signals and overlooks the varying data preferences among agents within the RAG system. In this paper, we propose a Differentiable Data Rewards (DDR) method, which end-to-end trains RAG systems by aligning data preferences between different RAG modules. DDR works by collecting the rewards to optimize each agent with a rollout method. This method prompts agents to sample some potential responses as perturbations, evaluates the impact of these perturbations on the whole RAG system, and subsequently optimizes the agent to produce outputs that improve the performance of the RAG system. Our experiments on various knowledge-intensive tasks demonstrate that DDR significantly outperforms the SFT method, particularly for LLMs with smaller-scale parameters that depend more on the retrieved knowledge. Additionally, DDR exhibits a stronger capability to align the data preference between RAG modules. The DDR method makes generation module more effective in extracting key information from documents and mitigating conflicts between parametric memory and external knowledge. All codes are available at https://github.com/OpenMatch/RAG-DDR.
☆ MathGAP: Out-of-Distribution Evaluation on Problems with Arbitrarily Complex Proofs
Large language models (LLMs) can solve arithmetic word problems with high accuracy, but little is known about how well they generalize to problems that are more complex than the ones on which they have been trained. Empirical investigations of such questions are impeded by two major flaws of current evaluations: (i) much of the evaluation data is contaminated, in the sense that it has already been seen during training, and (ii) benchmark datasets do not capture how problem proofs may be arbitrarily complex in various ways. As a step towards addressing these issues, we present a framework for evaluating LLMs on problems that have arbitrarily complex arithmetic proofs, called MathGAP. MathGAP generates problems that follow fixed proof specifications -- along with chain-of-thought reasoning annotations -- enabling systematic studies on generalization with respect to arithmetic proof complexity. We apply MathGAP to analyze how in-context learning interacts with generalization to problems that have more complex proofs. We find that among the models tested, most show a significant decrease in performance as proofs get deeper and wider. This effect is more pronounced in complex, nonlinear proof structures, which are challenging even for GPT-4o. Surprisingly, providing in-context examples from the same distribution as the test set is not always beneficial for performance. In particular, zero-shot prompting as well as demonstrating a diverse range of examples that are less complex than the test data sometimes yield similar or higher accuracies.
comment: Preprint
☆ Enhancing Text Generation in Joint NLG/NLU Learning Through Curriculum Learning, Semi-Supervised Training, and Advanced Optimization Techniques
Text generation is the automated process of producing written or spoken language using computational methods. It involves generating coherent and contextually relevant text based on predefined rules or learned patterns. However, challenges in text generation arise from maintaining coherence, ensuring diversity and creativity, and avoiding biases or inappropriate content. This research paper developed a novel approach to improve text generation in the context of joint Natural Language Generation (NLG) and Natural Language Understanding (NLU) learning. The data is prepared by gathering and preprocessing annotated datasets, including cleaning, tokenization, stemming, and stop-word removal. Feature extraction techniques such as POS tagging, Bag of words, and Term Frequency-Inverse Document Frequency (TF-IDF) are applied. Transformer-based encoders and decoders, capturing long range dependencies and improving source-target sequence modelling. Pre-trained language models like Optimized BERT are incorporated, along with a Hybrid Redfox Artificial Hummingbird Algorithm (HRAHA). Reinforcement learning with policy gradient techniques, semi-supervised training, improved attention mechanisms, and differentiable approximations like straight-through Gumbel SoftMax estimator are employed to fine-tune the models and handle complex linguistic tasks effectively. The proposed model is implemented using Python.
☆ Repetition Neurons: How Do Language Models Produce Repetitions?
This paper introduces repetition neurons, regarded as skill neurons responsible for the repetition problem in text generation tasks. These neurons are progressively activated more strongly as repetition continues, indicating that they perceive repetition as a task to copy the previous context repeatedly, similar to in-context learning. We identify these repetition neurons by comparing activation values before and after the onset of repetition in texts generated by recent pre-trained language models. We analyze the repetition neurons in three English and one Japanese pre-trained language models and observe similar patterns across them.
☆ Seeing Through VisualBERT: A Causal Adventure on Memetic Landscapes EMNLP
Detecting offensive memes is crucial, yet standard deep neural network systems often remain opaque. Various input attribution-based methods attempt to interpret their behavior, but they face challenges with implicitly offensive memes and non-causal attributions. To address these issues, we propose a framework based on a Structural Causal Model (SCM). In this framework, VisualBERT is trained to predict the class of an input meme based on both meme input and causal concepts, allowing for transparent interpretation. Our qualitative evaluation demonstrates the framework's effectiveness in understanding model behavior, particularly in determining whether the model was right due to the right reason, and in identifying reasons behind misclassification. Additionally, quantitative analysis assesses the significance of proposed modelling choices, such as de-confounding, adversarial learning, and dynamic routing, and compares them with input attribution methods. Surprisingly, we find that input attribution methods do not guarantee causality within our framework, raising questions about their reliability in safety-critical applications. The project page is at: https://newcodevelop.github.io/causality_adventure/
comment: Accepted at EMNLP Findings 2024
☆ IterSelectTune: An Iterative Training Framework for Efficient Instruction-Tuning Data Selection
As large language models (LLMs) continue to advance, instruction tuning has become critical for improving their ability to generate accurate and contextually appropriate responses. Although numerous instruction-tuning datasets have been developed to enhance LLM performance, selecting high-quality instruction data from large source datasets typically demands significant human effort. In this work, we introduce $\textbf{IterSelectTune}$, an efficient, cost-effective iterative training policy for selecting high-quality instruction data with no human involvement and limited reliance on GPT-4. By fine-tuning on approximately 20\% of the source data, our method consistently outperforms models fine-tuned on the full dataset across multiple benchmarks and public test datasets. These results highlight the effectiveness of our approach in enhancing LLM performance while reducing the computational resources required for instruction tuning.
☆ Progressive Mixed-Precision Decoding for Efficient LLM Inference
In spite of the great potential of large language models (LLMs) across various tasks, their deployment on resource-constrained devices remains challenging due to their excessive computational and memory demands. Quantization has emerged as an effective solution by storing weights in reduced precision. However, utilizing low precisions (i.e.~2/3-bit) to substantially alleviate the memory-boundedness of LLM decoding, still suffers from prohibitive performance drop. In this work, we argue that existing approaches fail to explore the diversity in computational patterns, redundancy, and sensitivity to approximations of the different phases of LLM inference, resorting to a uniform quantization policy throughout. Instead, we propose a novel phase-aware method that selectively allocates precision during different phases of LLM inference, achieving both strong context extraction during prefill and efficient memory bandwidth utilization during decoding. To further address the memory-boundedness of the decoding phase, we introduce Progressive Mixed-Precision Decoding (PMPD), a technique that enables the gradual lowering of precision deeper in the generated sequence, together with a spectrum of precision-switching schedulers that dynamically drive the precision-lowering decisions in either task-adaptive or prompt-adaptive manner. Extensive evaluation across diverse language tasks shows that when targeting Nvidia GPUs, PMPD achieves 1.4$-$12.2$\times$ speedup in matrix-vector multiplications over fp16 models, while when targeting an LLM-optimized NPU, our approach delivers a throughput gain of 3.8$-$8.0$\times$ over fp16 models and up to 1.54$\times$ over uniform quantization approaches while preserving the output quality.
☆ Breaking the Manual Annotation Bottleneck: Creating a Comprehensive Legal Case Criticality Dataset through Semi-Automated Labeling
Predicting case criticality helps legal professionals in the court system manage large volumes of case law. This paper introduces the Criticality Prediction dataset, a new resource for evaluating the potential influence of Swiss Federal Supreme Court decisions on future jurisprudence. Unlike existing approaches that rely on resource-intensive manual annotations, we semi-automatically derive labels leading to a much larger dataset than otherwise possible. Our dataset features a two-tier labeling system: (1) the LD-Label, which identifies cases published as Leading Decisions (LD), and (2) the Citation-Label, which ranks cases by their citation frequency and recency. This allows for a more nuanced evaluation of case importance. We evaluate several multilingual models, including fine-tuned variants and large language models, and find that fine-tuned models consistently outperform zero-shot baselines, demonstrating the need for task-specific adaptation. Our contributions include the introduction of this task and the release of a multilingual dataset to the research community.
☆ MedINST: Meta Dataset of Biomedical Instructions
The integration of large language model (LLM) techniques in the field of medical analysis has brought about significant advancements, yet the scarcity of large, diverse, and well-annotated datasets remains a major challenge. Medical data and tasks, which vary in format, size, and other parameters, require extensive preprocessing and standardization for effective use in training LLMs. To address these challenges, we introduce MedINST, the Meta Dataset of Biomedical Instructions, a novel multi-domain, multi-task instructional meta-dataset. MedINST comprises 133 biomedical NLP tasks and over 7 million training samples, making it the most comprehensive biomedical instruction dataset to date. Using MedINST as the meta dataset, we curate MedINST32, a challenging benchmark with different task difficulties aiming to evaluate LLMs' generalization ability. We fine-tune several LLMs on MedINST and evaluate on MedINST32, showcasing enhanced cross-task generalization.
☆ Unlocking Legal Knowledge: A Multilingual Dataset for Judicial Summarization in Switzerland
Legal research is a time-consuming task that most lawyers face on a daily basis. A large part of legal research entails looking up relevant caselaw and bringing it in relation to the case at hand. Lawyers heavily rely on summaries (also called headnotes) to find the right cases quickly. However, not all decisions are annotated with headnotes and writing them is time-consuming. Automated headnote creation has the potential to make hundreds of thousands of decisions more accessible for legal research in Switzerland alone. To kickstart this, we introduce the Swiss Leading Decision Summarization ( SLDS) dataset, a novel cross-lingual resource featuring 18K court rulings from the Swiss Federal Supreme Court (SFSC), in German, French, and Italian, along with German headnotes. We fine-tune and evaluate three mT5 variants, along with proprietary models. Our analysis highlights that while proprietary models perform well in zero-shot and one-shot settings, fine-tuned smaller models still provide a strong competitive edge. We publicly release the dataset to facilitate further research in multilingual legal summarization and the development of assistive technologies for legal professionals
☆ Parameter-efficient Adaptation of Multilingual Multimodal Models for Low-resource ASR
Automatic speech recognition (ASR) for low-resource languages remains a challenge due to the scarcity of labeled training data. Parameter-efficient fine-tuning and text-only adaptation are two popular methods that have been used to address such low-resource settings. In this work, we investigate how these techniques can be effectively combined using a multilingual multimodal model like SeamlessM4T. Multimodal models are able to leverage unlabeled text via text-only adaptation with further parameter-efficient ASR fine-tuning, thus boosting ASR performance. We also show cross-lingual transfer from a high-resource language, achieving up to a relative 17% WER reduction over a baseline in a zero-shot setting without any labeled speech.
☆ NLIP_Lab-IITH Multilingual MT System for WAT24 MT Shared Task
This paper describes NLIP Lab's multilingual machine translation system for the WAT24 shared task on multilingual Indic MT task for 22 scheduled languages belonging to 4 language families. We explore pre-training for Indic languages using alignment agreement objectives. We utilize bi-lingual dictionaries to substitute words from source sentences. Furthermore, we fine-tuned language direction-specific multilingual translation models using small and high-quality seed data. Our primary submission is a 243M parameters multilingual translation model covering 22 Indic languages. In the IN22-Gen benchmark, we achieved an average chrF++ score of 46.80 and 18.19 BLEU score for the En-Indic direction. In the Indic-En direction, we achieved an average chrF++ score of 56.34 and 30.82 BLEU score. In the In22-Conv benchmark, we achieved an average chrF++ score of 43.43 and BLEU score of 16.58 in the En-Indic direction, and in the Indic-En direction, we achieved an average of 52.44 and 29.77 for chrF++ and BLEU respectively. Our model\footnote{Our code and models are available at \url{https://github.com/maharajbrahma/WAT2024-MultiIndicMT}} is competitive with IndicTransv1 (474M parameter model).
comment: WMT 24 WAT Shared Task IndicMultiMT (Best System)
☆ Similarity-Dissimilarity Loss with Supervised Contrastive Learning for Multi-label Classification
Supervised contrastive learning has been explored in making use of label information for multi-label classification, but determining positive samples in multi-label scenario remains challenging. Previous studies have examined strategies for identifying positive samples, considering label overlap proportion between anchors and samples. However, they ignore various relations between given anchors and samples, as well as how to dynamically adjust the weights in contrastive loss functions based on different relations, leading to great ambiguity. In this paper, we introduce five distinct relations between multi-label samples and propose a Similarity-Dissimilarity Loss with contrastive learning for multi-label classification. Our loss function re-weights the loss by computing the similarity and dissimilarity between positive samples and a given anchor based on the introduced relations. We mainly conduct experiments for multi-label text classification on MIMIC datasets, then further extend the evaluation on MS-COCO. The Experimental results show that our proposed loss effectively improves the performance on all encoders under supervised contrastive learning paradigm, demonstrating its effectiveness and robustness.
☆ Think Thrice Before You Act: Progressive Thought Refinement in Large Language Models
Recent advancements in large language models (LLMs) have demonstrated that progressive refinement, rather than providing a single answer, results in more accurate and thoughtful outputs. However, existing methods often rely heavily on supervision signals to evaluate previous responses, making it difficult to assess output quality in more open-ended scenarios effectively. Additionally, these methods are typically designed for specific tasks, which limits their generalization to new domains. To address these limitations, we propose Progressive Thought Refinement (PTR), a framework that enables LLMs to refine their responses progressively. PTR operates in two phases: (1) Thought data construction stage: We propose a weak and strong model collaborative selection strategy to build a high-quality progressive refinement dataset to ensure logical consistency from thought to answers, and the answers are gradually refined in each round. (2) Thought-Mask Fine-Tuning Phase: We design a training structure to mask the "thought" and adjust loss weights to encourage LLMs to refine prior thought, teaching them to implicitly understand "how to improve" rather than "what is correct." Experimental results show that PTR significantly enhances LLM performance across ten diverse tasks (avg. from 49.6% to 53.5%) without task-specific fine-tuning. Notably, in more open-ended tasks, LLMs also demonstrate substantial improvements in the quality of responses beyond mere accuracy, suggesting that PTR truly teaches LLMs to self-improve over time.
comment: 10 pages, 4 figures
☆ Attr-Int: A Simple and Effective Entity Alignment Framework for Heterogeneous Knowledge Graphs
Entity alignment (EA) refers to the task of linking entities in different knowledge graphs (KGs). Existing EA methods rely heavily on structural isomorphism. However, in real-world KGs, aligned entities usually have non-isomorphic neighborhood structures, which paralyses the application of these structure-dependent methods. In this paper, we investigate and tackle the problem of entity alignment between heterogeneous KGs. First, we propose two new benchmarks to closely simulate real-world EA scenarios of heterogeneity. Then we conduct extensive experiments to evaluate the performance of representative EA methods on the new benchmarks. Finally, we propose a simple and effective entity alignment framework called Attr-Int, in which innovative attribute information interaction methods can be seamlessly integrated with any embedding encoder for entity alignment, improving the performance of existing entity alignment techniques. Experiments demonstrate that our framework outperforms the state-of-the-art approaches on two new benchmarks.
☆ MoR: Mixture of Ranks for Low-Rank Adaptation Tuning
Low-Rank Adaptation (LoRA) drives research to align its performance with full fine-tuning. However, significant challenges remain: (1) Simply increasing the rank size of LoRA does not effectively capture high-rank information, which leads to a performance bottleneck.(2) MoE-style LoRA methods substantially increase parameters and inference latency, contradicting the goals of efficient fine-tuning and ease of application. To address these challenges, we introduce Mixture of Ranks (MoR), which learns rank-specific information for different tasks based on input and efficiently integrates multi-rank information. We firstly propose a new framework that equates the integration of multiple LoRAs to expanding the rank of LoRA. Moreover, we hypothesize that low-rank LoRA already captures sufficient intrinsic information, and MoR can derive high-rank information through mathematical transformations of the low-rank components. Thus, MoR can reduces the learning difficulty of LoRA and enhances its multi-task capabilities. MoR achieves impressive results, with MoR delivering a 1.31\% performance improvement while using only 93.93\% of the parameters compared to baseline methods.
comment: 11 pages, 7 figures
☆ Towards Hybrid Intelligence in Journalism: Findings and Lessons Learnt from a Collaborative Analysis of Greek Political Rhetoric by ChatGPT and Humans
This chapter introduces a research project titled "Analyzing the Political Discourse: A Collaboration Between Humans and Artificial Intelligence", which was initiated in preparation for Greece's 2023 general elections. The project focused on the analysis of political leaders' campaign speeches, employing Artificial Intelligence (AI), in conjunction with an interdisciplinary team comprising journalists, a political scientist, and data scientists. The chapter delves into various aspects of political discourse analysis, including sentiment analysis, polarization, populism, topic detection, and Named Entities Recognition (NER). This experimental study investigates the capabilities of large language model (LLMs), and in particular OpenAI's ChatGPT, for analyzing political speech, evaluates its strengths and weaknesses, and highlights the essential role of human oversight in using AI in journalism projects and potentially other societal sectors. The project stands as an innovative example of human-AI collaboration (known also as "hybrid intelligence") within the realm of digital humanities, offering valuable insights for future initiatives.
☆ Linguistically Grounded Analysis of Language Models using Shapley Head Values
Understanding how linguistic knowledge is encoded in language models is crucial for improving their generalisation capabilities. In this paper, we investigate the processing of morphosyntactic phenomena, by leveraging a recently proposed method for probing language models via Shapley Head Values (SHVs). Using the English language BLiMP dataset, we test our approach on two widely used models, BERT and RoBERTa, and compare how linguistic constructions such as anaphor agreement and filler-gap dependencies are handled. Through quantitative pruning and qualitative clustering analysis, we demonstrate that attention heads responsible for processing related linguistic phenomena cluster together. Our results show that SHV-based attributions reveal distinct patterns across both models, providing insights into how language models organize and process linguistic information. These findings support the hypothesis that language models learn subnetworks corresponding to linguistic theory, with potential implications for cross-linguistic model analysis and interpretability in Natural Language Processing (NLP).
☆ Cross-Lingual Auto Evaluation for Assessing Multilingual LLMs
Evaluating machine-generated text remains a significant challenge in NLP, especially for non-English languages. Current methodologies, including automated metrics, human assessments, and LLM-based evaluations, predominantly focus on English, revealing a significant gap in multilingual evaluation frameworks. We introduce the Cross Lingual Auto Evaluation (CIA) Suite, an extensible framework that includes evaluator LLMs (Hercule) and a novel test set (Recon) specifically designed for multilingual evaluation. Our test set features 500 human-annotated instructions spanning various task capabilities along with human judgment scores across six languages. This would enable benchmarking of general-purpose multilingual LLMs and facilitate meta-evaluation of Evaluator LLMs. The proposed model, Hercule, is a cross-lingual evaluation model that addresses the scarcity of reference answers in the target language by learning to assign scores to responses based on easily available reference answers in English. Our experiments demonstrate that Hercule aligns more closely with human judgments compared to proprietary models, demonstrating the effectiveness of such cross-lingual evaluation in low resource scenarios. Further, it is also effective in zero-shot evaluation on unseen languages. This study is the first comprehensive examination of cross-lingual evaluation using LLMs, presenting a scalable and effective approach for multilingual assessment. All code, datasets, and models will be publicly available to enable further research in this important area.
☆ Metacognitive Monitoring: A Human Ability Beyond Generative Artificial Intelligence
Large language models (LLMs) have shown impressive alignment with human cognitive processes, raising questions about the extent of their similarity to human cognition. This study investigates whether LLMs, specifically ChatGPT, possess metacognitive monitoring abilities akin to humans-particularly in predicting memory performance on an item-by-item basis. We employed a cross-agent prediction model to compare the metacognitive performance of humans and ChatGPT in a language-based memory task involving garden-path sentences preceded by either fitting or unfitting context sentences. Both humans and ChatGPT rated the memorability of these sentences; humans then completed a surprise recognition memory test. Our findings reveal a significant positive relationship between humans' memorability ratings and their actual recognition performance, indicating reliable metacognitive monitoring. In contrast, ChatGPT did not exhibit a similar predictive capability. Bootstrapping analyses demonstrated that none of the GPT models tested (GPT-3.5-turbo, GPT-4-turbo, GPT-4o) could accurately predict human memory performance on a per-item basis. This suggests that, despite their advanced language processing abilities and alignment with human cognition at the object level, current LLMs lack the metacognitive mechanisms that enable humans to anticipate their memory performance. These results highlight a fundamental difference between human and AI cognition at the metacognitive level. Addressing this gap is crucial for developing AI systems capable of effective self-monitoring and adaptation to human needs, thereby enhancing human-AI interactions across domains such as education and personalized learning.
comment: 28 pages, 2 figures. arXiv admin note: substantial text overlap with arXiv:2403.05152
☆ On the Use of Audio to Improve Dialogue Policies
With the significant progress of speech technologies, spoken goal-oriented dialogue systems are becoming increasingly popular. One of the main modules of a dialogue system is typically the dialogue policy, which is responsible for determining system actions. This component usually relies only on audio transcriptions, being strongly dependent on their quality and ignoring very important extralinguistic information embedded in the user's speech. In this paper, we propose new architectures to add audio information by combining speech and text embeddings using a Double Multi-Head Attention component. Our experiments show that audio embedding-aware dialogue policies outperform text-based ones, particularly in noisy transcription scenarios, and that how text and audio embeddings are combined is crucial to improve performance. We obtained a 9.8% relative improvement in the User Request Score compared to an only-text-based dialogue system on the DSTC2 dataset.
comment: IberSpeech 2024
☆ Remember, Retrieve and Generate: Understanding Infinite Visual Concepts as Your Personalized Assistant
The development of large language models (LLMs) has significantly enhanced the capabilities of multimodal LLMs (MLLMs) as general assistants. However, lack of user-specific knowledge still restricts their application in human's daily life. In this paper, we introduce the Retrieval Augmented Personalization (RAP) framework for MLLMs' personalization. Starting from a general MLLM, we turn it into a personalized assistant in three steps. (a) Remember: We design a key-value database to store user-related information, e.g., user's name, avatar and other attributes. (b) Retrieve: When the user initiates a conversation, RAP will retrieve relevant information from the database using a multimodal retriever. (c) Generate: The input query and retrieved concepts' information are fed into MLLMs to generate personalized, knowledge-augmented responses. Unlike previous methods, RAP allows real-time concept editing via updating the external database. To further improve generation quality and alignment with user-specific information, we design a pipeline for data collection and create a specialized dataset for personalized training of MLLMs. Based on the dataset, we train a series of MLLMs as personalized multimodal assistants. By pretraining on large-scale dataset, RAP-MLLMs can generalize to infinite visual concepts without additional finetuning. Our models demonstrate outstanding flexibility and generation quality across a variety of tasks, such as personalized image captioning, question answering and visual recognition. The code, data and models are available at https://github.com/Hoar012/RAP-MLLM.
☆ LAR-ECHR: A New Legal Argument Reasoning Task and Dataset for Cases of the European Court of Human Rights
We present Legal Argument Reasoning (LAR), a novel task designed to evaluate the legal reasoning capabilities of Large Language Models (LLMs). The task requires selecting the correct next statement (from multiple choice options) in a chain of legal arguments from court proceedings, given the facts of the case. We constructed a dataset (LAR-ECHR) for this task using cases from the European Court of Human Rights (ECHR). We evaluated seven general-purpose LLMs on LAR-ECHR and found that (a) the ranking of the models is aligned with that of LegalBench, an established US-based legal reasoning benchmark, even though LAR-ECHR is based on EU law, (b) LAR-ECHR distinguishes top models more clearly, compared to LegalBench, (c) even the best model (GPT-4o) obtains 75.8% accuracy on LAR-ECHR, indicating significant potential for further model improvement. The process followed to construct LAR-ECHR can be replicated with cases from other legal systems.
comment: Published in Natural Legal Language Processing (NLLP) 2024 workshop
☆ Representation Learning of Structured Data for Medical Foundation Models NeurIPS 2024
Large Language Models (LLMs) have demonstrated remarkable performance across various domains, including healthcare. However, their ability to effectively represent structured non-textual data, such as the alphanumeric medical codes used in records like ICD-10 or SNOMED-CT, is limited and has been particularly exposed in recent research. This paper examines the challenges LLMs face in processing medical codes due to the shortcomings of current tokenization methods. As a result, we introduce the UniStruct architecture to design a multimodal medical foundation model of unstructured text and structured data, which addresses these challenges by adapting subword tokenization techniques specifically for the structured medical codes. Our approach is validated through model pre-training on both an extensive internal medical database and a public repository of structured medical records. Trained on over 1 billion tokens on the internal medical database, the proposed model achieves up to a 23% improvement in evaluation metrics, with around 2% gain attributed to our proposed tokenization. Additionally, when evaluated on the EHRSHOT public benchmark with a 1/1000 fraction of the pre-training data, the UniStruct model improves performance on over 42% of the downstream tasks. Our approach not only enhances the representation and generalization capabilities of patient-centric models but also bridges a critical gap in representation learning models' ability to handle complex structured medical data, alongside unstructured text.
comment: NeurIPS 2024 Workshop on Unifying Representations in Neural Models (UniReps 2024)
☆ Cerberus: Efficient Inference with Adaptive Parallel Decoding and Sequential Knowledge Enhancement
Large language models (LLMs) often face a bottleneck in inference speed due to their reliance on auto-regressive decoding. Recently, parallel decoding has shown significant promise in enhancing inference efficiency. However, we have identified two key issues with existing parallel decoding frameworks: (1) decoding heads fail to balance prediction accuracy and the parallelism of execution, and (2) parallel decoding is not a universal solution, as it can bring unnecessary overheads at some challenging decoding steps. To address these issues, we propose Cerberus, an adaptive parallel decoding framework introduces the gating mechanism to enable the LLMs to adaptively choose appropriate decoding approaches at each decoding step, along with introducing a new paradigm of decoding heads that introduce the sequential knowledge while maintaining execution parallelism. The experiment results demonstrate that the Cerberus can achieve up to 2.12x speed up compared to auto-regressive decoding, and outperforms one of the leading parallel decoding frameworks, Medusa, with a 10% - 30% increase in acceleration and superior generation quality.
☆ Do LLMs Overcome Shortcut Learning? An Evaluation of Shortcut Challenges in Large Language Models
Large Language Models (LLMs) have shown remarkable capabilities in various natural language processing tasks. However, LLMs may rely on dataset biases as shortcuts for prediction, which can significantly impair their robustness and generalization capabilities. This paper presents Shortcut Suite, a comprehensive test suite designed to evaluate the impact of shortcuts on LLMs' performance, incorporating six shortcut types, five evaluation metrics, and four prompting strategies. Our extensive experiments yield several key findings: 1) LLMs demonstrate varying reliance on shortcuts for downstream tasks, significantly impairing their performance. 2) Larger LLMs are more likely to utilize shortcuts under zero-shot and few-shot in-context learning prompts. 3) Chain-of-thought prompting notably reduces shortcut reliance and outperforms other prompting strategies, while few-shot prompts generally underperform compared to zero-shot prompts. 4) LLMs often exhibit overconfidence in their predictions, especially when dealing with datasets that contain shortcuts. 5) LLMs generally have a lower explanation quality in shortcut-laden datasets, with errors falling into three types: distraction, disguised comprehension, and logical fallacy. Our findings offer new insights for evaluating robustness and generalization in LLMs and suggest potential directions for mitigating the reliance on shortcuts. The code is available at \url {https://github.com/yyhappier/ShortcutSuite.git}.
☆ Probing-RAG: Self-Probing to Guide Language Models in Selective Document Retrieval
Retrieval-Augmented Generation (RAG) enhances language models by retrieving and incorporating relevant external knowledge. However, traditional retrieve-and-generate processes may not be optimized for real-world scenarios, where queries might require multiple retrieval steps or none at all. In this paper, we propose a Probing-RAG, which utilizes the hidden state representations from the intermediate layers of language models to adaptively determine the necessity of additional retrievals for a given query. By employing a pre-trained prober, Probing-RAG effectively captures the model's internal cognition, enabling reliable decision-making about retrieving external documents. Experimental results across five open-domain QA datasets demonstrate that Probing-RAG outperforms previous methods while reducing the number of redundant retrieval steps.
comment: 6 figures, 13 tables
☆ Do LLMs Have Political Correctness? Analyzing Ethical Biases and Jailbreak Vulnerabilities in AI Systems
Although large language models (LLMs) demonstrate impressive proficiency in various tasks, they present potential safety risks, such as `jailbreaks', where malicious inputs can coerce LLMs into generating harmful content. To address these issues, many LLM developers have implemented various safety measures to align these models. This alignment involves several techniques, including data filtering during pre-training, supervised fine-tuning, reinforcement learning from human feedback, and red-teaming exercises. These methods often introduce deliberate and intentional biases similar to Political Correctness (PC) to ensure the ethical behavior of LLMs. In this paper, we delve into the intentional biases injected into LLMs for safety purposes and examine methods to circumvent these safety alignment techniques. Notably, these intentional biases result in a jailbreaking success rate in GPT-4o models that differs by 20% between non-binary and cisgender keywords and by 16% between white and black keywords, even when the other parts of the prompts are identical. We introduce the concept of PCJailbreak, highlighting the inherent risks posed by these safety-induced biases. Additionally, we propose an efficient defense method PCDefense, which prevents jailbreak attempts by injecting defense prompts prior to generation. PCDefense stands as an appealing alternative to Guard Models, such as Llama-Guard, that require additional inference cost after text generation. Our findings emphasize the urgent need for LLM developers to adopt a more responsible approach when designing and implementing safety measures.
☆ Fine-Tuning Language Models on Multiple Datasets for Citation Intention Classification EMNLP 2024
Citation intention Classification (CIC) tools classify citations by their intention (e.g., background, motivation) and assist readers in evaluating the contribution of scientific literature. Prior research has shown that pretrained language models (PLMs) such as SciBERT can achieve state-of-the-art performance on CIC benchmarks. PLMs are trained via self-supervision tasks on a large corpus of general text and can quickly adapt to CIC tasks via moderate fine-tuning on the corresponding dataset. Despite their advantages, PLMs can easily overfit small datasets during fine-tuning. In this paper, we propose a multi-task learning (MTL) framework that jointly fine-tunes PLMs on a dataset of primary interest together with multiple auxiliary CIC datasets to take advantage of additional supervision signals. We develop a data-driven task relation learning (TRL) method that controls the contribution of auxiliary datasets to avoid negative transfer and expensive hyper-parameter tuning. We conduct experiments on three CIC datasets and show that fine-tuning with additional datasets can improve the PLMs' generalization performance on the primary dataset. PLMs fine-tuned with our proposed framework outperform the current state-of-the-art models by 7% to 11% on small datasets while aligning with the best-performing model on a large dataset.
comment: To be appear as a Findings paper at EMNLP 2024
☆ Mitigating Hallucinations in Large Vision-Language Models via Summary-Guided Decoding
Large Vision-Language Models (LVLMs) demonstrate impressive capabilities in generating detailed and coherent responses from visual inputs. However, they are prone to generate hallucinations due to an over-reliance on language priors. To address this issue, we investigate the language priors in LVLMs and make two key observations: (1) Even when predicting the tokens associated with image-related part-of-speech (POS), models increasingly rely on linguistic priors as the token sequences grow, thereby amplifying hallucinations. (2) Methods that directly calibrate LVLM's output distribution to mitigate language priors can lead to a degradation in text quality or even exacerbate hallucinations. Based on these findings, we propose a novel method, Summary-Guided Decoding (SGD). This method naturally encourages the model to focus more on image information by reducing the text context through summaries, while controlling only the image-related POS tokens to maintain text quality. Through experiments, we demonstrate that SGD achieves state-of-the-art performance on object hallucination benchmarks. Furthermore, in terms of the trade-off between precision and recall, SGD achieves Pareto optimality among the existing methods. Lastly, we observe that although existing methods struggle to balance the reduction of object hallucinations with maintaining text quality, SGD demonstrates robustness in handling this challenge.
☆ Computational Approaches to Arabic-English Code-Switching
Natural Language Processing (NLP) is a vital computational method for addressing language processing, analysis, and generation. NLP tasks form the core of many daily applications, from automatic text correction to speech recognition. While significant research has focused on NLP tasks for the English language, less attention has been given to Modern Standard Arabic and Dialectal Arabic. Globalization has also contributed to the rise of Code-Switching (CS), where speakers mix languages within conversations and even within individual words (intra-word CS). This is especially common in Arab countries, where people often switch between dialects or between dialects and a foreign language they master. CS between Arabic and English is frequent in Egypt, especially on social media. Consequently, a significant amount of code-switched content can be found online. Such code-switched data needs to be investigated and analyzed for several NLP tasks to tackle the challenges of this multilingual phenomenon and Arabic language challenges. No work has been done before for several integral NLP tasks on Arabic-English CS data. In this work, we focus on the Named Entity Recognition (NER) task and other tasks that help propose a solution for the NER task on CS data, e.g., Language Identification. This work addresses this gap by proposing and applying state-of-the-art techniques for Modern Standard Arabic and Arabic-English NER. We have created the first annotated CS Arabic-English corpus for the NER task. Also, we apply two enhancement techniques to improve the NER tagger on CS data using CS contextual embeddings and data augmentation techniques. All methods showed improvements in the performance of the NER taggers on CS data. Finally, we propose several intra-word language identification approaches to determine the language type of a mixed text and identify whether it is a named entity or not.
comment: PhD thesis
☆ Mitigating Biases to Embrace Diversity: A Comprehensive Annotation Benchmark for Toxic Language EMNLP
This study introduces a prescriptive annotation benchmark grounded in humanities research to ensure consistent, unbiased labeling of offensive language, particularly for casual and non-mainstream language uses. We contribute two newly annotated datasets that achieve higher inter-annotator agreement between human and language model (LLM) annotations compared to original datasets based on descriptive instructions. Our experiments show that LLMs can serve as effective alternatives when professional annotators are unavailable. Moreover, smaller models fine-tuned on multi-source LLM-annotated data outperform models trained on larger, single-source human-annotated datasets. These findings highlight the value of structured guidelines in reducing subjective variability, maintaining performance with limited data, and embracing language diversity. Content Warning: This article only analyzes offensive language for academic purposes. Discretion is advised.
comment: 12 pages, 9 figures, EMNLP-NLP4DH 2024
☆ Reference-Based Post-OCR Processing with LLM for Diacritic Languages
Extracting fine-grained OCR text from aged documents in diacritic languages remains challenging due to unexpected artifacts, time-induced degradation, and lack of datasets. While standalone spell correction approaches have been proposed, they show limited performance for historical documents due to numerous possible OCR error combinations and differences between modern and classical corpus distributions. We propose a method utilizing available content-focused ebooks as a reference base to correct imperfect OCR-generated text, supported by large language models. This technique generates high-precision pseudo-page-to-page labels for diacritic languages, where small strokes pose significant challenges in historical conditions. The pipeline eliminates various types of noise from aged documents and addresses issues such as missing characters, words, and disordered sequences. Our post-processing method, which generated a large OCR dataset of classical Vietnamese books, achieved a mean grading score of 8.72 on a 10-point scale. This outperformed the state-of-the-art transformer-based Vietnamese spell correction model, which scored 7.03 when evaluated on a sampled subset of the dataset. We also trained a baseline OCR model to assess and compare it with well-known engines. Experimental results demonstrate the strength of our baseline model compared to widely used open-source solutions. The resulting dataset will be released publicly to support future studies.
☆ Advancing Large Language Model Attribution through Self-Improving EMNLP 2024
Teaching large language models (LLMs) to generate text with citations to evidence sources can mitigate hallucinations and enhance verifiability in information-seeking systems. However, improving this capability requires high-quality attribution data, which is costly and labor-intensive. Inspired by recent advances in self-improvement that enhance LLMs without manual annotation, we present START, a Self-Taught AttRibuTion framework for iteratively improving the attribution capability of LLMs. First, to prevent models from stagnating due to initially insufficient supervision signals, START leverages the model to self-construct synthetic training data for warming up. To further self-improve the model's attribution ability, START iteratively utilizes fine-grained preference supervision signals constructed from its sampled responses to encourage robust, comprehensive, and attributable generation. Experiments on three open-domain question-answering datasets, covering long-form QA and multi-step reasoning, demonstrate significant performance gains of 25.13% on average without relying on human annotations and more advanced models. Further analysis reveals that START excels in aggregating information across multiple sources.
comment: Accepted by EMNLP 2024 Main Conference
☆ Learning to Route with Confidence Tokens
Large language models (LLMs) have demonstrated impressive performance on several tasks and are increasingly deployed in real-world applications. However, especially in high-stakes settings, it becomes vital to know when the output of an LLM may be unreliable. Depending on whether an answer is trustworthy, a system can then choose to route the question to another expert, or otherwise fall back on a safe default behavior. In this work, we study the extent to which LLMs can reliably indicate confidence in their answers, and how this notion of confidence can translate into downstream accuracy gains. We propose Self-REF, a lightweight training strategy to teach LLMs to express confidence in whether their answers are correct in a reliable manner. Self-REF introduces confidence tokens into the LLM, from which a confidence score can be extracted. Compared to conventional approaches such as verbalizing confidence and examining token probabilities, we demonstrate empirically that confidence tokens show significant improvements in downstream routing and rejection learning tasks.
☆ BANTH: A Multi-label Hate Speech Detection Dataset for Transliterated Bangla
The proliferation of transliterated texts in digital spaces has emphasized the need for detecting and classifying hate speech in languages beyond English, particularly in low-resource languages. As online discourse can perpetuate discrimination based on target groups, e.g. gender, religion, and origin, multi-label classification of hateful content can help in comprehending hate motivation and enhance content moderation. While previous efforts have focused on monolingual or binary hate classification tasks, no work has yet addressed the challenge of multi-label hate speech classification in transliterated Bangla. We introduce BanTH, the first multi-label transliterated Bangla hate speech dataset comprising 37.3k samples. The samples are sourced from YouTube comments, where each instance is labeled with one or more target groups, reflecting the regional demographic. We establish novel transformer encoder-based baselines by further pre-training on transliterated Bangla corpus. We also propose a novel translation-based LLM prompting strategy for transliterated text. Experiments reveal that our further pre-trained encoders are achieving state-of-the-art performance on the BanTH dataset, while our translation-based prompting outperforms other strategies in the zero-shot setting. The introduction of BanTH not only fills a critical gap in hate speech research for Bangla but also sets the stage for future exploration into code-mixed and multi-label classification challenges in underrepresented languages.
☆ SeerAttention: Learning Intrinsic Sparse Attention in Your LLMs
Attention is the cornerstone of modern Large Language Models (LLMs). Yet its quadratic complexity limits the efficiency and scalability of LLMs, especially for those with a long-context window. A promising approach addressing this limitation is to leverage the sparsity in attention. However, existing sparsity-based solutions predominantly rely on predefined patterns or heuristics to approximate sparsity. This practice falls short to fully capture the dynamic nature of attention sparsity in language-based tasks. This paper argues that attention sparsity should be learned rather than predefined. To this end, we design SeerAttention, a new Attention mechanism that augments the conventional attention with a learnable gate that adaptively selects significant blocks in an attention map and deems the rest blocks sparse. Such block-level sparsity effectively balances accuracy and speedup. To enable efficient learning of the gating network, we develop a customized FlashAttention implementation that extracts the block-level ground truth of attention map with minimum overhead. SeerAttention not only applies to post-training, but also excels in long-context fine-tuning. Our results show that at post-training stages, SeerAttention significantly outperforms state-of-the-art static or heuristic-based sparse attention methods, while also being more versatile and flexible to adapt to varying context lengths and sparsity ratios. When applied to long-context fine-tuning with YaRN, SeerAttention can achieve a remarkable 90% sparsity ratio at a 32k context length with minimal perplexity loss, offering a 5.67x speedup over FlashAttention-2.
☆ Breaking Chains: Unraveling the Links in Multi-Hop Knowledge Unlearning
Large language models (LLMs) serve as giant information stores, often including personal or copyrighted data, and retraining them from scratch is not a viable option. This has led to the development of various fast, approximate unlearning techniques to selectively remove knowledge from LLMs. Prior research has largely focused on minimizing the probabilities of specific token sequences by reversing the language modeling objective. However, these methods still leave LLMs vulnerable to adversarial attacks that exploit indirect references. In this work, we examine the limitations of current unlearning techniques in effectively erasing a particular type of indirect prompt: multi-hop queries. Our findings reveal that existing methods fail to completely remove multi-hop knowledge when one of the intermediate hops is unlearned. To address this issue, we propose MUNCH, a simple uncertainty-based approach that breaks down multi-hop queries into subquestions and leverages the uncertainty of the unlearned model in final decision-making. Empirical results demonstrate the effectiveness of our framework, and MUNCH can be easily integrated with existing unlearning techniques, making it a flexible and useful solution for enhancing unlearning processes.
comment: 16 pages, 5 figures
☆ Roadmap towards Superhuman Speech Understanding using Large Language Models
The success of large language models (LLMs) has prompted efforts to integrate speech and audio data, aiming to create general foundation models capable of processing both textual and non-textual inputs. Recent advances, such as GPT-4o, highlight the potential for end-to-end speech LLMs, which preserves non-semantic information and world knowledge for deeper speech understanding. To guide the development of speech LLMs, we propose a five-level roadmap, ranging from basic automatic speech recognition (ASR) to advanced superhuman models capable of integrating non-semantic information with abstract acoustic knowledge for complex tasks. Moreover, we design a benchmark, SAGI Bechmark, that standardizes critical aspects across various tasks in these five levels, uncovering challenges in using abstract acoustic knowledge and completeness of capability. Our findings reveal gaps in handling paralinguistic cues and abstract acoustic knowledge, and we offer future directions. This paper outlines a roadmap for advancing speech LLMs, introduces a benchmark for evaluation, and provides key insights into their current limitations and potential.
☆ CLaMP 2: Multimodal Music Information Retrieval Across 101 Languages Using Large Language Models
Challenges in managing linguistic diversity and integrating various musical modalities are faced by current music information retrieval systems. These limitations reduce their effectiveness in a global, multimodal music environment. To address these issues, we introduce CLaMP 2, a system compatible with 101 languages that supports both ABC notation (a text-based musical notation format) and MIDI (Musical Instrument Digital Interface) for music information retrieval. CLaMP 2, pre-trained on 1.5 million ABC-MIDI-text triplets, includes a multilingual text encoder and a multimodal music encoder aligned via contrastive learning. By leveraging large language models, we obtain refined and consistent multilingual descriptions at scale, significantly reducing textual noise and balancing language distribution. Our experiments show that CLaMP 2 achieves state-of-the-art results in both multilingual semantic search and music classification across modalities, thus establishing a new standard for inclusive and global music information retrieval.
comment: 17 pages, 10 figures, 4 tables
☆ From Babbling to Fluency: Evaluating the Evolution of Language Models in Terms of Human Language Acquisition
We examine the language capabilities of language models (LMs) from the critical perspective of human language acquisition. Building on classical language development theories, we propose a three-stage framework to assess the abilities of LMs, ranging from preliminary word understanding to complex grammar and complex logical reasoning. Using this framework, we evaluate the generative capacities of LMs using methods from linguistic research. Results indicate that although recent LMs outperform earlier models in overall performance, their developmental trajectory does not strictly follow the path of human language acquisition. Notably, in generation tasks, LMs are more similar to human performance in areas where information is easier to extract from the corpus, such as average word length, clauses, and auxiliary verbs. Newer LMs did not exhibit significant progress in terms of specific dimensions, such as clauses and auxiliary verbs, where the variation across corpora is relatively limited. Register theory offers a plausible explanation for these observations, suggesting that the linguistic features of the training data have a substantial impact on the models' abilities.
☆ A Systematic Investigation of Knowledge Retrieval and Selection for Retrieval Augmented Generation
Retrieval-augmented generation (RAG) has emerged as a powerful method for enhancing natural language generation by integrating external knowledge into a model's output. While prior work has demonstrated the importance of improving knowledge retrieval for boosting generation quality, the role of knowledge selection remains less clear. In this paper, we perform a comprehensive analysis of how knowledge retrieval and selection influence downstream generation performance in RAG systems. By simulating different retrieval and selection conditions through a controlled mixture of gold and distractor knowledge, we assess the impact of these factors on generation outcomes. Our findings indicate that the downstream generator model's capability, as well as the complexity of the task and dataset, significantly influence the impact of knowledge retrieval and selection on the overall RAG system performance. In typical scenarios, improving the knowledge recall score is key to enhancing generation outcomes, with the knowledge selector providing a limited additional benefit when a strong generator model is used on clear, well-defined tasks. For weaker generator models or more ambiguous tasks and datasets, the knowledge F1 score becomes a critical factor, and the knowledge selector plays a more prominent role in improving overall performance.
☆ Automatic Translation Alignment Pipeline for Multilingual Digital Editions of Literary Works
This paper investigates the application of translation alignment algorithms in the creation of a Multilingual Digital Edition (MDE) of Alessandro Manzoni's Italian novel "I promessi sposi" ("The Betrothed"), with translations in eight languages (English, Spanish, French, German, Dutch, Polish, Russian and Chinese) from the 19th and 20th centuries. We identify key requirements for the MDE to improve both the reader experience and support for translation studies. Our research highlights the limitations of current state-of-the-art algorithms when applied to the translation of literary texts and outlines an automated pipeline for MDE creation. This pipeline transforms raw texts into web-based, side-by-side representations of original and translated texts with different rendering options. In addition, we propose new metrics for evaluating the alignment of literary translations and suggest visualization techniques for future analysis.
comment: 18 pages, Computational Humanities Research Conference, December 4-6, 2024, Aarhus, Denmark
☆ Disentangling Likes and Dislikes in Personalized Generative Explainable Recommendation
Recent research on explainable recommendation generally frames the task as a standard text generation problem, and evaluates models simply based on the textual similarity between the predicted and ground-truth explanations. However, this approach fails to consider one crucial aspect of the systems: whether their outputs accurately reflect the users' (post-purchase) sentiments, i.e., whether and why they would like and/or dislike the recommended items. To shed light on this issue, we introduce new datasets and evaluation methods that focus on the users' sentiments. Specifically, we construct the datasets by explicitly extracting users' positive and negative opinions from their post-purchase reviews using an LLM, and propose to evaluate systems based on whether the generated explanations 1) align well with the users' sentiments, and 2) accurately identify both positive and negative opinions of users on the target items. We benchmark several recent models on our datasets and demonstrate that achieving strong performance on existing metrics does not ensure that the generated explanations align well with the users' sentiments. Lastly, we find that existing models can provide more sentiment-aware explanations when the users' (predicted) ratings for the target items are directly fed into the models as input. We will release our code and datasets upon acceptance.
☆ Atomic Calibration of LLMs in Long-Form Generations
Large language models (LLMs) often suffer from hallucinations, posing significant challenges for real-world applications. Confidence calibration, which estimates the underlying uncertainty of model predictions, is essential to enhance the LLMs' trustworthiness. Existing research on LLM calibration has primarily focused on short-form tasks, providing a single confidence score at the response level (macro calibration). However, this approach is insufficient for long-form generations, where responses often contain more complex statements and may include both accurate and inaccurate information. Therefore, we introduce atomic calibration, a novel approach that evaluates factuality calibration at a fine-grained level by breaking down long responses into atomic claims. We classify confidence elicitation methods into discriminative and generative types and demonstrate that their combination can enhance calibration. Our extensive experiments on various LLMs and datasets show that atomic calibration is well-suited for long-form generation and can also improve macro calibration results. Additionally, atomic calibration reveals insightful patterns in LLM confidence throughout the generation process.
☆ Large Language Models are Easily Confused: A Quantitative Metric, Security Implications and Typological Analysis
Language Confusion is a phenomenon where Large Language Models (LLMs) generate text that is neither in the desired language, nor in a contextually appropriate language. This phenomenon presents a critical challenge in text generation by LLMs, often appearing as erratic and unpredictable behavior. We hypothesize that there are linguistic regularities to this inherent vulnerability in LLMs and shed light on patterns of language confusion across LLMs. We introduce a novel metric, Language Confusion Entropy, designed to directly measure and quantify this confusion, based on language distributions informed by linguistic typology and lexical variation. Comprehensive comparisons with the Language Confusion Benchmark (Marchisio et al., 2024) confirm the effectiveness of our metric, revealing patterns of language confusion across LLMs. We further link language confusion to LLM security, and find patterns in the case of multilingual embedding inversion attacks. Our analysis demonstrates that linguistic typology offers theoretically grounded interpretation, and valuable insights into leveraging language similarities as a prior for LLM alignment and security.
comment: 17 pages, 6 figures, 14 tables
☆ SPIN: Self-Supervised Prompt INjection
Large Language Models (LLMs) are increasingly used in a variety of important applications, yet their safety and reliability remain as major concerns. Various adversarial and jailbreak attacks have been proposed to bypass the safety alignment and cause the model to produce harmful responses. We introduce Self-supervised Prompt INjection (SPIN) which can detect and reverse these various attacks on LLMs. As our self-supervised prompt defense is done at inference-time, it is also compatible with existing alignment and adds an additional layer of safety for defense. Our benchmarks demonstrate that our system can reduce the attack success rate by up to 87.9%, while maintaining the performance on benign user requests. In addition, we discuss the situation of an adaptive attacker and show that our method is still resilient against attackers who are aware of our defense.
☆ Web Agents with World Models: Learning and Leveraging Environment Dynamics in Web Navigation
Large language models (LLMs) have recently gained much attention in building autonomous agents. However, the performance of current LLM-based web agents in long-horizon tasks is far from optimal, often yielding errors such as repeatedly buying a non-refundable flight ticket. By contrast, humans can avoid such an irreversible mistake, as we have an awareness of the potential outcomes (e.g., losing money) of our actions, also known as the "world model". Motivated by this, our study first starts with preliminary analyses, confirming the absence of world models in current LLMs (e.g., GPT-4o, Claude-3.5-Sonnet, etc.). Then, we present a World-model-augmented (WMA) web agent, which simulates the outcomes of its actions for better decision-making. To overcome the challenges in training LLMs as world models predicting next observations, such as repeated elements across observations and long HTML inputs, we propose a transition-focused observation abstraction, where the prediction objectives are free-form natural language descriptions exclusively highlighting important state differences between time steps. Experiments on WebArena and Mind2Web show that our world models improve agents' policy selection without training and demonstrate our agents' cost- and time-efficiency compared to recent tree-search-based agents.
comment: Work in progress
☆ Proof Flow: Preliminary Study on Generative Flow Network Language Model Tuning for Formal Reasoning
Reasoning is a fundamental substrate for solving novel and complex problems. Deliberate efforts in learning and developing frameworks around System 2 reasoning have made great strides, yet problems of sufficient complexity remain largely out of reach for open models. To address this gap, we examine the potential of Generative Flow Networks as a fine-tuning method for LLMs to unlock advanced reasoning capabilities. In this paper, we present a proof of concept in the domain of formal reasoning, specifically in the Neural Theorem Proving (NTP) setting, where proofs specified in a formal language such as Lean can be deterministically and objectively verified. Unlike classical reward-maximization reinforcement learning, which frequently over-exploits high-reward actions and fails to effectively explore the state space, GFlowNets have emerged as a promising approach for sampling compositional objects, improving generalization, and enabling models to maintain diverse hypotheses. Our early results demonstrate GFlowNet fine-tuning's potential for enhancing model performance in a search setting, which is especially relevant given the paradigm shift towards inference time compute scaling and "thinking slowly."
☆ CBT-Bench: Evaluating Large Language Models on Assisting Cognitive Behavior Therapy
There is a significant gap between patient needs and available mental health support today. In this paper, we aim to thoroughly examine the potential of using Large Language Models (LLMs) to assist professional psychotherapy. To this end, we propose a new benchmark, CBT-BENCH, for the systematic evaluation of cognitive behavioral therapy (CBT) assistance. We include three levels of tasks in CBT-BENCH: I: Basic CBT knowledge acquisition, with the task of multiple-choice questions; II: Cognitive model understanding, with the tasks of cognitive distortion classification, primary core belief classification, and fine-grained core belief classification; III: Therapeutic response generation, with the task of generating responses to patient speech in CBT therapy sessions. These tasks encompass key aspects of CBT that could potentially be enhanced through AI assistance, while also outlining a hierarchy of capability requirements, ranging from basic knowledge recitation to engaging in real therapeutic conversations. We evaluated representative LLMs on our benchmark. Experimental results indicate that while LLMs perform well in reciting CBT knowledge, they fall short in complex real-world scenarios requiring deep analysis of patients' cognitive structures and generating effective responses, suggesting potential future work.
☆ Anchored Alignment for Self-Explanations Enhancement
In this work, we introduce a methodology for alignment designed to enhance the ability of large language models (LLMs) to articulate their reasoning (self-explanation) even in the absence of annotated rationale explanations. Our alignment methodology comprises three key components: explanation quality assessment, self-instruction dataset generation, and model alignment. Additionally, we present a novel technique called Alignment with Anchor Preference Pairs, which improves the selection of preference pairs by categorizing model outputs into three groups: consistently correct, consistently incorrect, and variable. By applying tailored strategies to each category, we enhance the effectiveness of Direct Preference Optimization (DPO). Our experimental results demonstrate that this approach significantly improves explanation quality while maintaining accuracy compared to other fine-tuning strategies.
☆ FaithBench: A Diverse Hallucination Benchmark for Summarization by Modern LLMs
Summarization is one of the most common tasks performed by large language models (LLMs), especially in applications like Retrieval-Augmented Generation (RAG). However, existing evaluations of hallucinations in LLM-generated summaries, and evaluations of hallucination detection models both suffer from a lack of diversity and recency in the LLM and LLM families considered. This paper introduces FaithBench, a summarization hallucination benchmark comprising challenging hallucinations made by 10 modern LLMs from 8 different families, with ground truth annotations by human experts. ``Challenging'' here means summaries on which popular, state-of-the-art hallucination detection models, including GPT-4o-as-a-judge, disagreed on. Our results show GPT-4o and GPT-3.5-Turbo produce the least hallucinations. However, even the best hallucination detection models have near 50\% accuracies on FaithBench, indicating lots of room for future improvement. The repo is https://github.com/vectara/FaithBench
☆ BQA: Body Language Question Answering Dataset for Video Large Language Models
A large part of human communication relies on nonverbal cues such as facial expressions, eye contact, and body language. Unlike language or sign language, such nonverbal communication lacks formal rules, requiring complex reasoning based on commonsense understanding. Enabling current Video Large Language Models (VideoLLMs) to accurately interpret body language is a crucial challenge, as human unconscious actions can easily cause the model to misinterpret their intent. To address this, we propose a dataset, BQA, a body language question answering dataset, to validate whether the model can correctly interpret emotions from short clips of body language comprising 26 emotion labels of videos of body language. We evaluated various VideoLLMs on BQA and revealed that understanding body language is challenging, and our analyses of the wrong answers by VideoLLMs show that certain VideoLLMs made significantly biased answers depending on the age group and ethnicity of the individuals in the video. The dataset is available.
☆ Measuring Free-Form Decision-Making Inconsistency of Language Models in Military Crisis Simulations
There is an increasing interest in using language models (LMs) for automated decision-making, with multiple countries actively testing LMs to aid in military crisis decision-making. To scrutinize relying on LM decision-making in high-stakes settings, we examine the inconsistency of responses in a crisis simulation ("wargame"), similar to reported tests conducted by the US military. Prior work illustrated escalatory tendencies and varying levels of aggression among LMs but were constrained to simulations with pre-defined actions. This was due to the challenges associated with quantitatively measuring semantic differences and evaluating natural language decision-making without relying on pre-defined actions. In this work, we query LMs for free form responses and use a metric based on BERTScore to measure response inconsistency quantitatively. Leveraging the benefits of BERTScore, we show that the inconsistency metric is robust to linguistic variations that preserve semantic meaning in a question-answering setting across text lengths. We show that all five tested LMs exhibit levels of inconsistency that indicate semantic differences, even when adjusting the wargame setting, anonymizing involved conflict countries, or adjusting the sampling temperature parameter $T$. Further qualitative evaluation shows that models recommend courses of action that share few to no similarities. We also study the impact of different prompt sensitivity variations on inconsistency at temperature $T = 0$. We find that inconsistency due to semantically equivalent prompt variations can exceed response inconsistency from temperature sampling for most studied models across different levels of ablations. Given the high-stakes nature of military deployment, we recommend further consideration be taken before using LMs to inform military decisions or other cases of high-stakes decision-making.
☆ Meta-DiffuB: A Contextualized Sequence-to-Sequence Text Diffusion Model with Meta-Exploration
The diffusion model, a new generative modeling paradigm, has achieved significant success in generating images, audio, video, and text. It has been adapted for sequence-to-sequence text generation (Seq2Seq) through DiffuSeq, termed S2S Diffusion. Existing S2S-Diffusion models predominantly rely on fixed or hand-crafted rules to schedule noise during the diffusion and denoising processes. However, these models are limited by non-contextualized noise, which fails to fully consider the characteristics of Seq2Seq tasks. In this paper, we propose the Meta-DiffuB framework - a novel scheduler-exploiter S2S-Diffusion paradigm designed to overcome the limitations of existing S2S-Diffusion models. We employ Meta-Exploration to train an additional scheduler model dedicated to scheduling contextualized noise for each sentence. Our exploiter model, an S2S-Diffusion model, leverages the noise scheduled by our scheduler model for updating and generation. Meta-DiffuB achieves state-of-the-art performance compared to previous S2S-Diffusion models and fine-tuned pre-trained language models (PLMs) across four Seq2Seq benchmark datasets. We further investigate and visualize the impact of Meta-DiffuB's noise scheduling on the generation of sentences with varying difficulties. Additionally, our scheduler model can function as a "plug-and-play" model to enhance DiffuSeq without the need for fine-tuning during the inference stage.
☆ Failing Forward: Improving Generative Error Correction for ASR with Synthetic Data and Retrieval Augmentation
Generative Error Correction (GEC) has emerged as a powerful post-processing method to enhance the performance of Automatic Speech Recognition (ASR) systems. However, we show that GEC models struggle to generalize beyond the specific types of errors encountered during training, limiting their ability to correct new, unseen errors at test time, particularly in out-of-domain (OOD) scenarios. This phenomenon amplifies with named entities (NEs), where, in addition to insufficient contextual information or knowledge about the NEs, novel NEs keep emerging. To address these issues, we propose DARAG (Data- and Retrieval-Augmented Generative Error Correction), a novel approach designed to improve GEC for ASR in in-domain (ID) and OOD scenarios. We augment the GEC training dataset with synthetic data generated by prompting LLMs and text-to-speech models, thereby simulating additional errors from which the model can learn. For OOD scenarios, we simulate test-time errors from new domains similarly and in an unsupervised fashion. Additionally, to better handle named entities, we introduce retrieval-augmented correction by augmenting the input with entities retrieved from a database. Our approach is simple, scalable, and both domain- and language-agnostic. We experiment on multiple datasets and settings, showing that DARAG outperforms all our baselines, achieving 8\% -- 30\% relative WER improvements in ID and 10\% -- 33\% improvements in OOD settings.
comment: Preprint. Under Review
☆ The Geometry of Numerical Reasoning: Language Models Compare Numeric Properties in Linear Subspaces
This paper investigates whether large language models (LLMs) utilize numerical attributes encoded in a low-dimensional subspace of the embedding space when answering logical comparison questions (e.g., Was Cristiano born before Messi?). We first identified these subspaces using partial least squares regression, which effectively encodes the numerical attributes associated with the entities in comparison prompts. Further, we demonstrate causality by intervening in these subspaces to manipulate hidden states, thereby altering the LLM's comparison outcomes. Experimental results show that our findings hold for different numerical attributes, indicating that LLMs utilize the linearly encoded information for numerical reasoning.
☆ Evaluating Self-Generated Documents for Enhancing Retrieval-Augmented Generation with Large Language Models
In retrieval-augmented generation systems, the integration of self-generated documents (SGDs) alongside retrieved content has emerged as a promising strategy for enhancing the performance of large language model. However, previous research primarily focuses on optimizing the use of SGDs, with the inherent properties of SGDs remaining underexplored. Therefore, this paper conducts a comprehensive analysis of different types of SGDs and experiments on various knowledge-intensive tasks. We develop a taxonomy of SGDs grounded in Systemic Functional Linguistics (SFL) to compare the influence of different SGD categories. Our findings offer key insights into what kinds of SGDs most effectively contribute to improving LLM's performance. The results and further fusion methods based on SGD categories also provide practical guidelines for taking better advantage of SGDs to achieve significant advancements in knowledge-driven QA tasks with RAG.
comment: Under Review
☆ MCQG-SRefine: Multiple Choice Question Generation and Evaluation with Iterative Self-Critique, Correction, and Comparison Feedback
Automatic question generation (QG) is essential for AI and NLP, particularly in intelligent tutoring, dialogue systems, and fact verification. Generating multiple-choice questions (MCQG) for professional exams, like the United States Medical Licensing Examination (USMLE), is particularly challenging, requiring domain expertise and complex multi-hop reasoning for high-quality questions. However, current large language models (LLMs) like GPT-4 struggle with professional MCQG due to outdated knowledge, hallucination issues, and prompt sensitivity, resulting in unsatisfactory quality and difficulty. To address these challenges, we propose MCQG-SRefine, an LLM self-refine-based (Critique and Correction) framework for converting medical cases into high-quality USMLE-style questions. By integrating expert-driven prompt engineering with iterative self-critique and self-correction feedback, MCQG-SRefine significantly enhances human expert satisfaction regarding both the quality and difficulty of the questions. Furthermore, we introduce an LLM-as-Judge-based automatic metric to replace the complex and costly expert evaluation process, ensuring reliable and expert-aligned assessments.
comment: Equal contribution for the first two authors
☆ aiXcoder-7B: A Lightweight and Effective Large Language Model for Code Completion
Large Language Models (LLMs) have been widely used in code completion, and researchers are focusing on scaling up LLMs to improve their accuracy. However, larger LLMs will increase the response time of code completion and decrease the developers' productivity. In this paper, we propose a lightweight and effective LLM for code completion named aiXcoder-7B. Compared to existing LLMs, aiXcoder-7B achieves higher code completion accuracy while having smaller scales (i.e., 7 billion parameters). We attribute the superiority of aiXcoder-7B to three key factors: (1) Multi-objective training. We employ three training objectives, one of which is our proposed Structured Fill-In-the-Middle (SFIM). SFIM considers the syntax structures in code and effectively improves the performance of LLMs for code. (2) Diverse data sampling strategies. They consider inter-file relationships and enhance the capability of LLMs in understanding cross-file contexts. (3) Extensive high-quality data. We establish a rigorous data collection pipeline and consume a total of 1.2 trillion unique tokens for training aiXcoder-7B. This vast volume of data enables aiXcoder-7B to learn a broad distribution of code. We evaluate aiXcoder-7B in five popular code completion benchmarks and a new benchmark collected by this paper. The results show that aiXcoder-7B outperforms the latest six LLMs with similar sizes and even surpasses four larger LLMs (e.g., StarCoder2-15B and CodeLlama-34B), positioning aiXcoder-7B as a lightweight and effective LLM for academia and industry. Finally, we summarize three valuable insights for helping practitioners train the next generations of LLMs for code. aiXcoder-7B has been open-souced and gained significant attention. As of the submission date, aiXcoder-7B has received 2,193 GitHub Stars.
comment: aiXcoder-7B is available at https://github.com/aixcoder-plugin/aiXcoder-7B/tree/main
☆ Chain of Ideas: Revolutionizing Research in Novel Idea Development with LLM Agents
Effective research ideation is a critical step for scientific research. However, the exponential increase in scientific literature makes it challenging for researchers to stay current with recent advances and identify meaningful research directions. Recent developments in large language models~(LLMs) suggest a promising avenue for automating the generation of novel research ideas. However, existing methods for idea generation either trivially prompt LLMs or directly expose LLMs to extensive literature without indicating useful information. Inspired by the research process of human researchers, we propose a Chain-of-Ideas~(CoI) agent, an LLM-based agent that organizes relevant literature in a chain structure to effectively mirror the progressive development in a research domain. This organization facilitates LLMs to capture the current advancements in research, thereby enhancing their ideation capabilities. Furthermore, we propose Idea Arena, an evaluation protocol that can comprehensively evaluate idea generation methods from different perspectives, aligning closely with the preferences of human researchers. Experimental results indicate that the CoI agent consistently outperforms other methods and shows comparable quality as humans in research idea generation. Moreover, our CoI agent is budget-friendly, with a minimum cost of \$0.50 to generate a candidate idea and its corresponding experimental design.
comment: 10 pages,5 figures, conference
☆ Router-Tuning: A Simple and Effective Approach for Enabling Dynamic-Depth in Transformers
Traditional transformer models often allocate a fixed amount of computational resources to every input token, leading to inefficient and unnecessary computation. To address this, the Mixture of Depths (MoD) was introduced to dynamically adjust the computational depth by skipping less important layers. Despite its promise, current MoD approaches remain under-explored and face two main challenges: (1) \textit{high training costs due to the need to train the entire model along with the routers that determine which layers to skip}, and (2) \textit{the risk of performance degradation when important layers are bypassed}. In response to the first issue, we propose Router-Tuning, a method that fine-tunes only the router on a small dataset, drastically reducing the computational overhead associated with full model training. For the second challenge, we propose MindSkip, which deploys \textit{Attention with Dynamic Depths}. This method preserves the model's performance while significantly enhancing computational and memory efficiency. Extensive experiments demonstrate that our approach delivers competitive results while dramatically improving the computation efficiency, e.g., 21\% speedup and only a 0.2\% performance drop. The code is released at \url{https://github.com/CASE-Lab-UMD/Router-Tuning}.
☆ AdaSwitch: Adaptive Switching between Small and Large Agents for Effective Cloud-Local Collaborative Learning EMNLP 2024
Recent advancements in large language models (LLMs) have been remarkable. Users face a choice between using cloud-based LLMs for generation quality and deploying local-based LLMs for lower computational cost. The former option is typically costly and inefficient, while the latter usually fails to deliver satisfactory performance for reasoning steps requiring deliberate thought processes. In this work, we propose a novel LLM utilization paradigm that facilitates the collaborative operation of large cloud-based LLMs and smaller local-deployed LLMs. Our framework comprises two primary modules: the local agent instantiated with a relatively smaller LLM, handling less complex reasoning steps, and the cloud agent equipped with a larger LLM, managing more intricate reasoning steps. This collaborative processing is enabled through an adaptive mechanism where the local agent introspectively identifies errors and proactively seeks assistance from the cloud agent, thereby effectively integrating the strengths of both locally-deployed and cloud-based LLMs, resulting in significant enhancements in task completion performance and efficiency. We evaluate AdaSwitch across 7 benchmarks, ranging from mathematical reasoning and complex question answering, using various types of LLMs to instantiate the local and cloud agents. The empirical results show that AdaSwitch effectively improves the performance of the local agent, and sometimes achieves competitive results compared to the cloud agent while utilizing much less computational overhead.
comment: EMNLP 2024 Main Conference
☆ EH-MAM: Easy-to-Hard Masked Acoustic Modeling for Self-Supervised Speech Representation Learning
In this paper, we present EH-MAM (Easy-to-Hard adaptive Masked Acoustic Modeling), a novel self-supervised learning approach for speech representation learning. In contrast to the prior methods that use random masking schemes for Masked Acoustic Modeling (MAM), we introduce a novel selective and adaptive masking strategy. Specifically, during SSL training, we progressively introduce harder regions to the model for reconstruction. Our approach automatically selects hard regions and is built on the observation that the reconstruction loss of individual frames in MAM can provide natural signals to judge the difficulty of solving the MAM pre-text task for that frame. To identify these hard regions, we employ a teacher model that first predicts the frame-wise losses and then decides which frames to mask. By learning to create challenging problems, such as identifying harder frames and solving them simultaneously, the model is able to learn more effective representations and thereby acquire a more comprehensive understanding of the speech. Quantitatively, EH-MAM outperforms several state-of-the-art baselines across various low-resource speech recognition and SUPERB benchmarks by 5%-10%. Additionally, we conduct a thorough analysis to show that the regions masked by EH-MAM effectively capture useful context across speech frames.
☆ An Evolved Universal Transformer Memory
Prior methods propose to offset the escalating costs of modern foundation models by dropping specific parts of their contexts with hand-designed rules, while attempting to preserve their original performance. We overcome this trade-off with Neural Attention Memory Models (NAMMs), introducing a learned network for memory management that improves both the performance and efficiency of transformers. We evolve NAMMs atop pre-trained transformers to provide different latent contexts focusing on the most relevant information for individual layers and attention heads.NAMMs are universally applicable to any model using self-attention as they condition exclusively on the values in the produced attention matrices. Learning NAMMs on a small set of problems, we achieve substantial performance improvements across multiple long-context benchmarks while cutting the model's input contexts up to a fraction of the original sizes. We show the generality of our conditioning enables zero-shot transfer of NAMMs trained only on language to entirely new transformer architectures even across input modalities, with their benefits carrying over to vision and reinforcement learning.
comment: 29 pages, 14 figures. Preprint, under submission. Source code is available at https://github.com/SakanaAI/evo-memory
☆ SLM-Mod: Small Language Models Surpass LLMs at Content Moderation
Large language models (LLMs) have shown promise in many natural language understanding tasks, including content moderation. However, these models can be expensive to query in real-time and do not allow for a community-specific approach to content moderation. To address these challenges, we explore the use of open-source small language models (SLMs) for community-specific content moderation tasks. We fine-tune and evaluate SLMs (less than 15B parameters) by comparing their performance against much larger open- and closed-sourced models. Using 150K comments from 15 popular Reddit communities, we find that SLMs outperform LLMs at content moderation -- 11.5% higher accuracy and 25.7% higher recall on average across all communities. We further show the promise of cross-community content moderation, which has implications for new communities and the development of cross-platform moderation techniques. Finally, we outline directions for future work on language model based content moderation. Code and links to HuggingFace models can be found at https://github.com/AGoyal0512/SLM-Mod.
comment: Preprint: 15 pages, 8 figures, 8 pages
☆ Better to Ask in English: Evaluation of Large Language Models on English, Low-resource and Cross-Lingual Settings
Large Language Models (LLMs) are trained on massive amounts of data, enabling their application across diverse domains and tasks. Despite their remarkable performance, most LLMs are developed and evaluated primarily in English. Recently, a few multi-lingual LLMs have emerged, but their performance in low-resource languages, especially the most spoken languages in South Asia, is less explored. To address this gap, in this study, we evaluate LLMs such as GPT-4, Llama 2, and Gemini to analyze their effectiveness in English compared to other low-resource languages from South Asia (e.g., Bangla, Hindi, and Urdu). Specifically, we utilized zero-shot prompting and five different prompt settings to extensively investigate the effectiveness of the LLMs in cross-lingual translated prompts. The findings of the study suggest that GPT-4 outperformed Llama 2 and Gemini in all five prompt settings and across all languages. Moreover, all three LLMs performed better for English language prompts than other low-resource language prompts. This study extensively investigates LLMs in low-resource language contexts to highlight the improvements required in LLMs and language-specific resources to develop more generally purposed NLP applications.
☆ Mapping Bias in Vision Language Models: Signposts, Pitfalls, and the Road Ahead NAACL 2025
As Vision Language Models (VLMs) gain widespread use, their fairness remains under-explored. In this paper, we analyze demographic biases across five models and six datasets. We find that portrait datasets like UTKFace and CelebA are the best tools for bias detection, finding gaps in performance and fairness between LLaVa and CLIP models. However, scene based datasets like PATA, VLStereoSet fail to be useful benchmarks for bias due to their construction. As for pronoun based datasets like VisoGender, we receive mixed signals as only some subsets of the data are useful in providing insights. To alleviate this problem, we introduce a more difficult version of VisoGender to serve as a more rigorous evaluation. Based on these results, we call for more effective and carefully designed datasets to ensure VLMs are both fair and reliable.
comment: Under Review at NAACL 2025
☆ Data Defenses Against Large Language Models
Large language models excel at performing inference over text to extract information, summarize information, or generate additional text. These inference capabilities are implicated in a variety of ethical harms spanning surveillance, labor displacement, and IP/copyright theft. While many policy, legal, and technical mitigations have been proposed to counteract these harms, these mitigations typically require cooperation from institutions that move slower than technical advances (i.e., governments) or that have few incentives to act to counteract these harms (i.e., the corporations that create and profit from these LLMs). In this paper, we define and build "data defenses" -- a novel strategy that directly empowers data owners to block LLMs from performing inference on their data. We create data defenses by developing a method to automatically generate adversarial prompt injections that, when added to input text, significantly reduce the ability of LLMs to accurately infer personally identifying information about the subject of the input text or to use copyrighted text in inference. We examine the ethics of enabling such direct resistance to LLM inference, and argue that making data defenses that resist and subvert LLMs enables the realization of important values such as data ownership, data sovereignty, and democratic control over AI systems. We verify that our data defenses are cheap and fast to generate, work on the latest commercial and open-source LLMs, resistance to countermeasures, and are robust to several different attack settings. Finally, we consider the security implications of LLM data defenses and outline several future research directions in this area. Our code is available at https://github.com/wagnew3/LLMDataDefenses and a tool for using our defenses to protect text against LLM inference is at https://wagnew3.github.io/LLM-Data-Defenses/.
☆ Retrieval-Enhanced Named Entity Recognition
When combined with In-Context Learning, a technique that enables models to adapt to new tasks by incorporating task-specific examples or demonstrations directly within the input prompt, autoregressive language models have achieved good performance in a wide range of tasks and applications. However, this combination has not been properly explored in the context of named entity recognition, where the structure of this task poses unique challenges. We propose RENER (Retrieval-Enhanced Named Entity Recognition), a technique for named entity recognition using autoregressive language models based on In-Context Learning and information retrieval techniques. When presented with an input text, RENER fetches similar examples from a dataset of training examples that are used to enhance a language model to recognize named entities from this input text. RENER is modular and independent of the underlying language model and information retrieval algorithms. Experimental results show that in the CrossNER collection we achieve state-of-the-art performance with the proposed technique and that information retrieval can increase the F-score by up to 11 percentage points.
comment: 13 pages, 6 figures, 3 tables
☆ Learning to Summarize from LLM-generated Feedback
Developing effective text summarizers remains a challenge due to issues like hallucinations, key information omissions, and verbosity in LLM-generated summaries. This work explores using LLM-generated feedback to improve summary quality by aligning the summaries with human preferences for faithfulness, completeness, and conciseness. We introduce FeedSum, a large-scale dataset containing multi-dimensional LLM feedback on summaries of varying quality across diverse domains. Our experiments show how feedback quality, dimensionality, and granularity influence preference learning, revealing that high-quality, multi-dimensional, fine-grained feedback significantly improves summary generation. We also compare two methods for using this feedback: supervised fine-tuning and direct preference optimization. Finally, we introduce SummLlama3-8b, a model that outperforms the nearly 10x larger Llama3-70b-instruct in generating human-preferred summaries, demonstrating that smaller models can achieve superior performance with appropriate training. The full dataset will be released soon. The SummLlama3-8B model is now available at https://huggingface.co/DISLab/SummLlama3-8B.
☆ Controllable Generation via Locally Constrained Resampling
Autoregressive models have demonstrated an unprecedented ability at modeling the intricacies of natural language. However, they continue to struggle with generating complex outputs that adhere to logical constraints. Sampling from a fully-independent distribution subject to a constraint is hard. Sampling from an autoregressive distribution subject to a constraint is doubly hard: We have to contend not only with the hardness of the constraint but also the distribution's lack of structure. We propose a tractable probabilistic approach that performs Bayesian conditioning to draw samples subject to a constraint. Our approach considers the entire sequence, leading to a more globally optimal constrained generation than current greedy methods. Starting from a model sample, we induce a local, factorized distribution which we can tractably condition on the constraint. To generate samples that satisfy the constraint, we sample from the conditional distribution, correct for biases in the samples and resample. The resulting samples closely approximate the target distribution and are guaranteed to satisfy the constraints. We evaluate our approach on several tasks, including LLM detoxification and solving Sudoku puzzles. We show that by disallowing a list of toxic expressions our approach is able to steer the model's outputs away from toxic generations, outperforming similar approaches to detoxification. We conclude by showing that our approach achieves a perfect accuracy on Sudoku compared to <50% for GPT4-o and Gemini 1.5.
comment: arXiv admin note: text overlap with arXiv:2312.03905
☆ A Little Human Data Goes A Long Way
Faced with an expensive human annotation process, creators of NLP systems increasingly turn to synthetic data generation. While this method shows promise, the extent to which synthetic data can replace human annotation is poorly understood. We investigate the use of synthetic data in Fact Verification (FV) and Question Answering (QA) by studying the effects of incrementally replacing human generated data with synthetic points on eight diverse datasets. Strikingly, replacing up to 90% of the training data only marginally decreases performance, but replacing the final 10% leads to severe declines. We find that models trained on purely synthetic data can be reliably improved by including as few as 125 human generated data points. We show that matching the performance gain of just a little additional human data (only 200 points) requires an order of magnitude more synthetic data and estimate price ratios at which human annotation would be a more cost-effective solution. Our results suggest that even when human annotation at scale is infeasible, there is great value to having a small proportion of the dataset being human generated.
☆ Be My Donor. Transfer the NLP Datasets Between the Languages Using LLM
In this work, we investigated how one can use the LLM to transfer the dataset and its annotation from one language to another. This is crucial since sharing the knowledge between different languages could boost certain underresourced directions in the target language, saving lots of efforts in data annotation or quick prototyping. We experiment with English and Russian pairs translating the DEFT corpus. This corpus contains three layers of annotation dedicated to term-definition pair mining, which is a rare annotation type for Russian. We provide a pipeline for the annotation transferring using ChatGPT3.5-turbo and Llama-3.1-8b as core LLMs. In the end, we train the BERT-based models on the translated dataset to establish a baseline.
☆ Efficient Vision-Language Models by Summarizing Visual Tokens into Compact Registers
Recent advancements in vision-language models (VLMs) have expanded their potential for real-world applications, enabling these models to perform complex reasoning on images. In the widely used fully autoregressive transformer-based models like LLaVA, projected visual tokens are prepended to textual tokens. Oftentimes, visual tokens are significantly more than prompt tokens, resulting in increased computational overhead during both training and inference. In this paper, we propose Visual Compact Token Registers (Victor), a method that reduces the number of visual tokens by summarizing them into a smaller set of register tokens. Victor adds a few learnable register tokens after the visual tokens and summarizes the visual information into these registers using the first few layers in the language tower of VLMs. After these few layers, all visual tokens are discarded, significantly improving computational efficiency for both training and inference. Notably, our method is easy to implement and requires a small number of new trainable parameters with minimal impact on model performance. In our experiment, with merely 8 visual registers--about 1% of the original tokens--Victor shows less than a 4% accuracy drop while reducing the total training time by 43% and boosting the inference throughput by 3.3X.
☆ UCFE: A User-Centric Financial Expertise Benchmark for Large Language Models
This paper introduces the UCFE: User-Centric Financial Expertise benchmark, an innovative framework designed to evaluate the ability of large language models (LLMs) to handle complex real-world financial tasks. UCFE benchmark adopts a hybrid approach that combines human expert evaluations with dynamic, task-specific interactions to simulate the complexities of evolving financial scenarios. Firstly, we conducted a user study involving 804 participants, collecting their feedback on financial tasks. Secondly, based on this feedback, we created our dataset that encompasses a wide range of user intents and interactions. This dataset serves as the foundation for benchmarking 12 LLM services using the LLM-as-Judge methodology. Our results show a significant alignment between benchmark scores and human preferences, with a Pearson correlation coefficient of 0.78, confirming the effectiveness of the UCFE dataset and our evaluation approach. UCFE benchmark not only reveals the potential of LLMs in the financial sector but also provides a robust framework for assessing their performance and user satisfaction.The benchmark dataset and evaluation code are available.
☆ Towards Cross-Cultural Machine Translation with Retrieval-Augmented Generation from Multilingual Knowledge Graphs EMNLP 2024
Translating text that contains entity names is a challenging task, as cultural-related references can vary significantly across languages. These variations may also be caused by transcreation, an adaptation process that entails more than transliteration and word-for-word translation. In this paper, we address the problem of cross-cultural translation on two fronts: (i) we introduce XC-Translate, the first large-scale, manually-created benchmark for machine translation that focuses on text that contains potentially culturally-nuanced entity names, and (ii) we propose KG-MT, a novel end-to-end method to integrate information from a multilingual knowledge graph into a neural machine translation model by leveraging a dense retrieval mechanism. Our experiments and analyses show that current machine translation systems and large language models still struggle to translate texts containing entity names, whereas KG-MT outperforms state-of-the-art approaches by a large margin, obtaining a 129% and 62% relative improvement compared to NLLB-200 and GPT-4, respectively.
comment: Accepted at EMNLP 2024
☆ From Isolated Conversations to Hierarchical Schemas: Dynamic Tree Memory Representation for LLMs
Recent advancements in large language models have significantly improved their context windows, yet challenges in effective long-term memory management remain. We introduce MemTree, an algorithm that leverages a dynamic, tree-structured memory representation to optimize the organization, retrieval, and integration of information, akin to human cognitive schemas. MemTree organizes memory hierarchically, with each node encapsulating aggregated textual content, corresponding semantic embeddings, and varying abstraction levels across the tree's depths. Our algorithm dynamically adapts this memory structure by computing and comparing semantic embeddings of new and existing information to enrich the model's context-awareness. This approach allows MemTree to handle complex reasoning and extended interactions more effectively than traditional memory augmentation methods, which often rely on flat lookup tables. Evaluations on benchmarks for multi-turn dialogue understanding and document question answering show that MemTree significantly enhances performance in scenarios that demand structured memory management.
☆ Learning Multimodal Cues of Children's Uncertainty SIGDIAL 2023
Understanding uncertainty plays a critical role in achieving common ground (Clark et al.,1983). This is especially important for multimodal AI systems that collaborate with users to solve a problem or guide the user through a challenging concept. In this work, for the first time, we present a dataset annotated in collaboration with developmental and cognitive psychologists for the purpose of studying nonverbal cues of uncertainty. We then present an analysis of the data, studying different roles of uncertainty and its relationship with task difficulty and performance. Lastly, we present a multimodal machine learning model that can predict uncertainty given a real-time video clip of a participant, which we find improves upon a baseline multimodal transformer model. This work informs research on cognitive coordination between human-human and human-AI and has broad implications for gesture understanding and generation. The anonymized version of our data and code will be publicly available upon the completion of the required consent forms and data sheets.
comment: SIGDIAL 2023
☆ Learning Metadata-Agnostic Representations for Text-to-SQL In-Context Example Selection NeurIPS 2024
In-context learning (ICL) is a powerful paradigm where large language models (LLMs) benefit from task demonstrations added to the prompt. Yet, selecting optimal demonstrations is not trivial, especially for complex or multi-modal tasks where input and output distributions differ. We hypothesize that forming task-specific representations of the input is key. In this paper, we propose a method to align representations of natural language questions and those of SQL queries in a shared embedding space. Our technique, dubbed MARLO - Metadata-Agnostic Representation Learning for Text-tO-SQL - uses query structure to model querying intent without over-indexing on underlying database metadata (i.e. tables, columns, or domain-specific entities of a database referenced in the question or query). This allows MARLO to select examples that are structurally and semantically relevant for the task rather than examples that are spuriously related to a certain domain or question phrasing. When used to retrieve examples based on question similarity, MARLO shows superior performance compared to generic embedding models (on average +2.9\%pt. in execution accuracy) on the Spider benchmark. It also outperforms the next best method that masks metadata information by +0.8\%pt. in execution accuracy on average, while imposing a significantly lower inference latency.
comment: Accepted to NeurIPS 2024 Table Representation Learning workshop
☆ Efficient Retrieval of Temporal Event Sequences from Textual Descriptions
Retrieving temporal event sequences from textual descriptions is essential for applications such as analyzing e-commerce behavior, monitoring social media activities, and tracking criminal incidents. In this paper, we introduce TPP-LLM-Embedding, a unified model for efficiently embedding and retrieving event sequences based on natural language descriptions. Built on the TPP-LLM framework, which integrates large language models with temporal point processes, our model encodes both event types and times, generating a sequence-level representation through pooling. Textual descriptions are embedded using the same architecture, ensuring a shared embedding space for both sequences and descriptions. We optimize a contrastive loss based on similarity between these embeddings, bringing matching pairs closer and separating non-matching ones. TPP-LLM-Embedding enables efficient retrieval and demonstrates superior performance compared to baseline models across diverse datasets.
☆ Style-Compress: An LLM-Based Prompt Compression Framework Considering Task-Specific Styles EMNLP 2024
Prompt compression condenses contexts while maintaining their informativeness for different usage scenarios. It not only shortens the inference time and reduces computational costs during the usage of large language models, but also lowers expenses when using closed-source models. In a preliminary study, we discover that when instructing language models to compress prompts, different compression styles (e.g., extractive or abstractive) impact performance of compressed prompts on downstream tasks. Building on this insight, we propose Style-Compress, a lightweight framework that adapts a smaller language model to compress prompts for a larger model on a new task without additional training. Our approach iteratively generates and selects effective compressed prompts as task-specific demonstrations through style variation and in-context learning, enabling smaller models to act as efficient compressors with task-specific examples. Style-Compress outperforms two baseline compression models in four tasks: original prompt reconstruction, text summarization, multi-hop QA, and CoT reasoning. In addition, with only 10 samples and 100 queries for adaptation, prompts compressed by Style-Compress achieve performance on par with or better than original prompts at a compression ratio of 0.25 or 0.5.
comment: EMNLP 2024 Findings
☆ From Barriers to Tactics: A Behavioral Science-Informed Agentic Workflow for Personalized Nutrition Coaching
Effective management of cardiometabolic conditions requires sustained positive nutrition habits, often hindered by complex and individualized barriers. Direct human management is simply not scalable, while previous attempts aimed at automating nutrition coaching lack the personalization needed to address these diverse challenges. This paper introduces a novel LLM-powered agentic workflow designed to provide personalized nutrition coaching by directly targeting and mitigating patient-specific barriers. Grounded in behavioral science principles, the workflow leverages a comprehensive mapping of nutrition-related barriers to corresponding evidence-based strategies. A specialized LLM agent intentionally probes for and identifies the root cause of a patient's dietary struggles. Subsequently, a separate LLM agent delivers tailored tactics designed to overcome those specific barriers with patient context. We designed and validated our approach through a user study with individuals with cardiometabolic conditions, demonstrating the system's ability to accurately identify barriers and provide personalized guidance. Furthermore, we conducted a large-scale simulation study, grounding on real patient vignettes and expert-validated metrics, to evaluate the system's performance across a wide range of scenarios. Our findings demonstrate the potential of this LLM-powered agentic workflow to improve nutrition coaching by providing personalized, scalable, and behaviorally-informed interventions.
comment: 22 pages
☆ Graph Neural Flows for Unveiling Systemic Interactions Among Irregularly Sampled Time Series NeurIPS 2024
Interacting systems are prevalent in nature. It is challenging to accurately predict the dynamics of the system if its constituent components are analyzed independently. We develop a graph-based model that unveils the systemic interactions of time series observed at irregular time points, by using a directed acyclic graph to model the conditional dependencies (a form of causal notation) of the system components and learning this graph in tandem with a continuous-time model that parameterizes the solution curves of ordinary differential equations (ODEs). Our technique, a graph neural flow, leads to substantial enhancements over non-graph-based methods, as well as graph-based methods without the modeling of conditional dependencies. We validate our approach on several tasks, including time series classification and forecasting, to demonstrate its efficacy.
comment: NeurIPS 2024. Code is available at https://github.com/gmerca/GNeuralFlow
☆ Measuring and Modifying the Readability of English Texts with GPT-4
The success of Large Language Models (LLMs) in other domains has raised the question of whether LLMs can reliably assess and manipulate the readability of text. We approach this question empirically. First, using a published corpus of 4,724 English text excerpts, we find that readability estimates produced ``zero-shot'' from GPT-4 Turbo and GPT-4o mini exhibit relatively high correlation with human judgments (r = 0.76 and r = 0.74, respectively), out-performing estimates derived from traditional readability formulas and various psycholinguistic indices. Then, in a pre-registered human experiment (N = 59), we ask whether Turbo can reliably make text easier or harder to read. We find evidence to support this hypothesis, though considerable variance in human judgments remains unexplained. We conclude by discussing the limitations of this approach, including limited scope, as well as the validity of the ``readability'' construct and its dependence on context, audience, and goal.
comment: 9 pages, 6 figures, workshop TSAR 2024
☆ Generating Signed Language Instructions in Large-Scale Dialogue Systems NAACL 2024
We introduce a goal-oriented conversational AI system enhanced with American Sign Language (ASL) instructions, presenting the first implementation of such a system on a worldwide multimodal conversational AI platform. Accessible through a touch-based interface, our system receives input from users and seamlessly generates ASL instructions by leveraging retrieval methods and cognitively based gloss translations. Central to our design is a sign translation module powered by Large Language Models, alongside a token-based video retrieval system for delivering instructional content from recipes and wikiHow guides. Our development process is deeply rooted in a commitment to community engagement, incorporating insights from the Deaf and Hard-of-Hearing community, as well as experts in cognitive and ASL learning sciences. The effectiveness of our signing instructions is validated by user feedback, achieving ratings on par with those of the system in its non-signing variant. Additionally, our system demonstrates exceptional performance in retrieval accuracy and text-generation quality, measured by metrics such as BERTScore. We have made our codebase and datasets publicly accessible at https://github.com/Merterm/signed-dialogue, and a demo of our signed instruction video retrieval system is available at https://huggingface.co/spaces/merterm/signed-instructions.
comment: 2024 Annual Conference of the North American Chapter of the Association for Computational Linguistics (NAACL 2024) Industry Track
☆ LLMs are Biased Teachers: Evaluating LLM Bias in Personalized Education
With the increasing adoption of large language models (LLMs) in education, concerns about inherent biases in these models have gained prominence. We evaluate LLMs for bias in the personalized educational setting, specifically focusing on the models' roles as "teachers". We reveal significant biases in how models generate and select educational content tailored to different demographic groups, including race, ethnicity, sex, gender, disability status, income, and national origin. We introduce and apply two bias score metrics--Mean Absolute Bias (MAB) and Maximum Difference Bias (MDB)--to analyze 9 open and closed state-of-the-art LLMs. Our experiments, which utilize over 17,000 educational explanations across multiple difficulty levels and topics, uncover that models perpetuate both typical and inverted harmful stereotypes.
comment: 46 Pages, 55 Figures, dataset release pending publication
☆ Personalized Adaptation via In-Context Preference Learning
Reinforcement Learning from Human Feedback (RLHF) is widely used to align Language Models (LMs) with human preferences. However, existing approaches often neglect individual user preferences, leading to suboptimal personalization. We present the Preference Pretrained Transformer (PPT), a novel approach for adaptive personalization using online user feedback. PPT leverages the in-context learning capabilities of transformers to dynamically adapt to individual preferences. Our approach consists of two phases: (1) an offline phase where we train a single policy model using a history-dependent loss function, and (2) an online phase where the model adapts to user preferences through in-context learning. We demonstrate PPT's effectiveness in a contextual bandit setting, showing that it achieves personalized adaptation superior to existing methods while significantly reducing the computational costs. Our results suggest the potential of in-context learning for scalable and efficient personalization in large language models.
☆ RiTeK: A Dataset for Large Language Models Complex Reasoning over Textual Knowledge Graphs
Answering complex real-world questions often requires accurate retrieval from textual knowledge graphs (TKGs). The scarcity of annotated data, along with intricate topological structures, makes this task particularly challenging. As the nature of relational path information could enhance the inference ability of Large Language Models (LLMs), efficiently retrieving more complex relational path information from TKGs presents another key challenge. To tackle these challenges, we first develop a Dataset for LLMs Complex Reasoning over Textual Knowledge Graphs (RiTeK) with a broad topological structure coverage.We synthesize realistic user queries that integrate diverse topological structures, relational information, and complex textual descriptions. We conduct rigorous expert evaluation to validate the quality of our synthesized queries. And then, we introduce an enhanced Monte Carlo Tree Search (MCTS) method, Relational MCTS, to automatically extract relational path information from textual graphs for specific queries. Our dataset mainly covers the medical domain as the relation types and entity are complex and publicly available. Experimental results indicate that RiTeK poses significant challenges for current retrieval and LLM systems, while the proposed Relational MCTS method enhances LLM inference ability and achieves state-of-the-art performance on RiTeK.
☆ Are LLMs Models of Distributional Semantics? A Case Study on Quantifiers
Distributional semantics is the linguistic theory that a word's meaning can be derived from its distribution in natural language (i.e., its use). Language models are commonly viewed as an implementation of distributional semantics, as they are optimized to capture the statistical features of natural language. It is often argued that distributional semantics models should excel at capturing graded/vague meaning based on linguistic conventions, but struggle with truth-conditional reasoning and symbolic processing. We evaluate this claim with a case study on vague (e.g. "many") and exact (e.g. "more than half") quantifiers. Contrary to expectations, we find that, across a broad range of models of various types, LLMs align more closely with human judgements on exact quantifiers versus vague ones. These findings call for a re-evaluation of the assumptions underpinning what distributional semantics models are, as well as what they can capture.
comment: 9 Pages, 3 Figures
☆ Debiasing Large Vision-Language Models by Ablating Protected Attribute Representations NeurIPS
Large Vision Language Models (LVLMs) such as LLaVA have demonstrated impressive capabilities as general-purpose chatbots that can engage in conversations about a provided input image. However, their responses are influenced by societal biases present in their training datasets, leading to undesirable differences in how the model responds when presented with images depicting people of different demographics. In this work, we propose a novel debiasing framework for LVLMs by directly ablating biased attributes during text generation to avoid generating text related to protected attributes, or even representing them internally. Our method requires no training and a relatively small amount of representative biased outputs (~1000 samples). Our experiments show that not only can we can minimize the propensity of LVLMs to generate text related to protected attributes, but we can even use synthetic data to inform the ablation while retaining captioning performance on real data such as COCO. Furthermore, we find the resulting generations from a debiased LVLM exhibit similar accuracy as a baseline biased model, showing that debiasing effects can be achieved without sacrificing model performance.
comment: NeurIPS workshop on SafeGenAI, 10 pages, 2 figures
☆ Detecting AI-Generated Texts in Cross-Domains
Existing tools to detect text generated by a large language model (LLM) have met with certain success, but their performance can drop when dealing with texts in new domains. To tackle this issue, we train a ranking classifier called RoBERTa-Ranker, a modified version of RoBERTa, as a baseline model using a dataset we constructed that includes a wider variety of texts written by humans and generated by various LLMs. We then present a method to fine-tune RoBERTa-Ranker that requires only a small amount of labeled data in a new domain. Experiments show that this fine-tuned domain-aware model outperforms the popular DetectGPT and GPTZero on both in-domain and cross-domain texts, where AI-generated texts may either be in a different domain or generated by a different LLM not used to generate the training datasets. This approach makes it feasible and economical to build a single system to detect AI-generated texts across various domains.
☆ From Single to Multi: How LLMs Hallucinate in Multi-Document Summarization
Although many studies have investigated and reduced hallucinations in large language models (LLMs) for single-document tasks, research on hallucination in multi-document summarization (MDS) tasks remains largely unexplored. Specifically, it is unclear how the challenges arising from handling multiple documents (e.g., repetition and diversity of information) affect models outputs. In this work, we investigate how hallucinations manifest in LLMs when summarizing topic-specific information from multiple documents. Since no benchmarks exist for investigating hallucinations in MDS, we use existing news and conversation datasets, annotated with topic-specific insights, to create two novel multi-document benchmarks. When evaluating 5 LLMs on our benchmarks, we observe that on average, up to 75% of the content in LLM-generated summary is hallucinated, with hallucinations more likely to occur towards the end of the summaries. Moreover, when summarizing non-existent topic-related information, gpt-3.5-turbo and GPT-4o still generate summaries about 79.35% and 44% of the time, raising concerns about their tendency to fabricate content. To understand the characteristics of these hallucinations, we manually evaluate 700+ insights and find that most errors stem from either failing to follow instructions or producing overly generic insights. Motivated by these observations, we investigate the efficacy of simple post-hoc baselines in mitigating hallucinations but find them only moderately effective. Our results underscore the need for more effective approaches to systematically mitigate hallucinations in MDS. We release our dataset and code at github.com/megagonlabs/Hallucination_MDS.
☆ FinQAPT: Empowering Financial Decisions with End-to-End LLM-driven Question Answering Pipeline
Financial decision-making hinges on the analysis of relevant information embedded in the enormous volume of documents in the financial domain. To address this challenge, we developed FinQAPT, an end-to-end pipeline that streamlines the identification of relevant financial reports based on a query, extracts pertinent context, and leverages Large Language Models (LLMs) to perform downstream tasks. To evaluate the pipeline, we experimented with various techniques to optimize the performance of each module using the FinQA dataset. We introduced a novel clustering-based negative sampling technique to enhance context extraction and a novel prompting method called Dynamic N-shot Prompting to boost the numerical question-answering capabilities of LLMs. At the module level, we achieved state-of-the-art accuracy on FinQA, attaining an accuracy of 80.6\%. However, at the pipeline level, we observed decreased performance due to challenges in extracting relevant context from financial reports. We conducted a detailed error analysis of each module and the end-to-end pipeline, pinpointing specific challenges that must be addressed to develop a robust solution for handling complex financial tasks.
comment: Accepted in ICAIF 2024, 8 pages, 5 figures, 4 tables
☆ Identifying High Consideration E-Commerce Search Queries EMNLP 2024
In e-commerce, high consideration search missions typically require careful and elaborate decision making, and involve a substantial research investment from customers. We consider the task of identifying High Consideration (HC) queries. Identifying such queries enables e-commerce sites to better serve user needs using targeted experiences such as curated QA widgets that help users reach purchase decisions. We explore the task by proposing an Engagement-based Query Ranking (EQR) approach, focusing on query ranking to indicate potential engagement levels with query-related shopping knowledge content during product search. Unlike previous studies on predicting trends, EQR prioritizes query-level features related to customer behavior, finance, and catalog information rather than popularity signals. We introduce an accurate and scalable method for EQR and present experimental results demonstrating its effectiveness. Offline experiments show strong ranking performance. Human evaluation shows a precision of 96% for HC queries identified by our model. The model was commercially deployed, and shown to outperform human-selected queries in terms of downstream customer impact, as measured through engagement.
comment: Accepted by EMNLP 2024 (Industry Track)
☆ Boosting LLM Translation Skills without General Ability Loss via Rationale Distillation
Large Language Models (LLMs) have achieved impressive results across numerous NLP tasks but still encounter difficulties in machine translation. Traditional methods to improve translation have typically involved fine-tuning LLMs using parallel corpora. However, vanilla fine-tuning often leads to catastrophic forgetting of the instruction-following capabilities and alignment with human preferences, compromising their broad general abilities and introducing potential security risks. These abilities, which are developed using proprietary and unavailable training data, make existing continual instruction tuning methods ineffective. To overcome this issue, we propose a novel approach called RaDis (Rationale Distillation). RaDis harnesses the strong generative capabilities of LLMs to create rationales for training data, which are then "replayed" to prevent forgetting. These rationales encapsulate general knowledge and safety principles, acting as self-distillation targets to regulate the training process. By jointly training on both reference translations and self-generated rationales, the model can learn new translation skills while preserving its overall general abilities. Extensive experiments demonstrate that our method enhances machine translation performance while maintaining the broader capabilities of LLMs across other tasks. This work presents a pathway for creating more versatile LLMs that excel in specialized tasks without compromising generality and safety.
♻ ☆ Towards Multilingual LLM Evaluation for European Languages
The rise of Large Language Models (LLMs) has revolutionized natural language processing across numerous languages and tasks. However, evaluating LLM performance in a consistent and meaningful way across multiple European languages remains challenging, especially due to the scarcity of language-parallel multilingual benchmarks. We introduce a multilingual evaluation approach tailored for European languages. We employ translated versions of five widely-used benchmarks to assess the capabilities of 40 LLMs across 21 European languages. Our contributions include examining the effectiveness of translated benchmarks, assessing the impact of different translation services, and offering a multilingual evaluation framework for LLMs that includes newly created datasets: EU20-MMLU, EU20-HellaSwag, EU20-ARC, EU20-TruthfulQA, and EU20-GSM8K. The benchmarks and results are made publicly available to encourage further research in multilingual LLM evaluation.
♻ ☆ Retrieval Augmented Generation or Long-Context LLMs? A Comprehensive Study and Hybrid Approach EMNLP 2024
Retrieval Augmented Generation (RAG) has been a powerful tool for Large Language Models (LLMs) to efficiently process overly lengthy contexts. However, recent LLMs like Gemini-1.5 and GPT-4 show exceptional capabilities to understand long contexts directly. We conduct a comprehensive comparison between RAG and long-context (LC) LLMs, aiming to leverage the strengths of both. We benchmark RAG and LC across various public datasets using three latest LLMs. Results reveal that when resourced sufficiently, LC consistently outperforms RAG in terms of average performance. However, RAG's significantly lower cost remains a distinct advantage. Based on this observation, we propose Self-Route, a simple yet effective method that routes queries to RAG or LC based on model self-reflection. Self-Route significantly reduces the computation cost while maintaining a comparable performance to LC. Our findings provide a guideline for long-context applications of LLMs using RAG and LC.
comment: Accepted to EMNLP 2024 industry track
♻ ☆ Many-Shot In-Context Learning NeurIPS
Large language models (LLMs) excel at few-shot in-context learning (ICL) -- learning from a few examples provided in context at inference, without any weight updates. Newly expanded context windows allow us to investigate ICL with hundreds or thousands of examples -- the many-shot regime. Going from few-shot to many-shot, we observe significant performance gains across a wide variety of generative and discriminative tasks. While promising, many-shot ICL can be bottlenecked by the available amount of human-generated examples. To mitigate this limitation, we explore two new settings: Reinforced and Unsupervised ICL. Reinforced ICL uses model-generated chain-of-thought rationales in place of human examples. Unsupervised ICL removes rationales from the prompt altogether, and prompts the model only with domain-specific questions. We find that both Reinforced and Unsupervised ICL can be quite effective in the many-shot regime, particularly on complex reasoning tasks. Finally, we demonstrate that, unlike few-shot learning, many-shot learning is effective at overriding pretraining biases, can learn high-dimensional functions with numerical inputs, and performs comparably to fine-tuning. We also find that inference cost increases linearly in the many-shot regime, and frontier LLMs benefit from many-shot ICL to varying degrees. Our analysis also reveals the limitations of next-token prediction loss as an indicator of downstream ICL performance.
comment: NeurIPS (Spotlight)
♻ ☆ Dynamic Topic Language Model on Heterogeneous Children's Mental Health Clinical Notes
Mental health diseases affect children's lives and well-beings which have received increased attention since the COVID-19 pandemic. Analyzing psychiatric clinical notes with topic models is critical to evaluating children's mental status over time. However, few topic models are built for longitudinal settings, and most existing approaches fail to capture temporal trajectories for each document. To address these challenges, we develop a dynamic topic model with consistent topics and individualized temporal dependencies on the evolving document metadata. Our model preserves the semantic meaning of discovered topics over time and incorporates heterogeneity among documents. In particular, when documents can be categorized, we propose a classifier-free approach to maximize topic heterogeneity across different document groups. We also present an efficient variational optimization procedure adapted for the multistage longitudinal setting. In this case study, we apply our method to the psychiatric clinical notes from a large tertiary pediatric hospital in Southern California and achieve a 38% increase in the overall coherence of extracted topics. Our real data analysis reveals that children tend to express more negative emotions during state shutdowns and more positive when schools reopen. Furthermore, it suggests that sexual and gender minority (SGM) children display more pronounced reactions to major COVID-19 events and a greater sensitivity to vaccine-related news than non-SGM children. This study examines children's mental health progression during the pandemic and offers clinicians valuable insights to recognize disparities in children's mental health related to their sexual and gender identities.
♻ ☆ The Impact of Visual Information in Chinese Characters: Evaluating Large Models' Ability to Recognize and Utilize Radicals
The glyphic writing system of Chinese incorporates information-rich visual features in each character, such as radicals that provide hints about meaning or pronunciation. However, there has been no investigation into whether contemporary Large Language Models (LLMs) and Vision-Language Models (VLMs) can harness these sub-character features in Chinese through prompting. In this study, we establish a benchmark to evaluate LLMs' and VLMs' understanding of visual elements in Chinese characters, including radicals, composition structures, strokes, and stroke counts. Our results reveal that models surprisingly exhibit some, but still limited, knowledge of the visual information, regardless of whether images of characters are provided. To incite models' ability to use radicals, we further experiment with incorporating radicals into the prompts for Chinese language processing (CLP) tasks. We observe consistent improvement in Part-Of-Speech tagging when providing additional information about radicals, suggesting the potential to enhance CLP by integrating sub-character information.
♻ ☆ Superlatives in Context: Modeling the Implicit Semantics of Superlatives
Superlatives are used to single out elements with a maximal/minimal property. Semantically, superlatives perform a set comparison: something (or some things) has the min/max property out of a set. As such, superlatives provide an ideal phenomenon for studying implicit phenomena and discourse restrictions. While this comparison set is often not explicitly defined, its (implicit) restrictions can be inferred from the discourse context the expression appears in. In this work we provide an extensive computational study on the semantics of superlatives. We propose a unified account of superlative semantics which allows us to derive a broad-coverage annotation schema. Using this unified schema we annotated a multi-domain dataset of superlatives and their semantic interpretations. We specifically focus on interpreting implicit or ambiguous superlative expressions, by analyzing how the discourse context restricts the set of interpretations. In a set of experiments we then analyze how well models perform at variations of predicting superlative semantics, with and without context. We show that the fine-grained semantics of superlatives in context can be challenging for contemporary models, including GPT-4.
comment: 11 pages
♻ ☆ Larger Language Models Don't Care How You Think: Why Chain-of-Thought Prompting Fails in Subjective Tasks
In-Context Learning (ICL) in Large Language Models (LLM) has emerged as the dominant technique for performing natural language tasks, as it does not require updating the model parameters with gradient-based methods. ICL promises to "adapt" the LLM to perform the present task at a competitive or state-of-the-art level at a fraction of the computational cost. ICL can be augmented by incorporating the reasoning process to arrive at the final label explicitly in the prompt, a technique called Chain-of-Thought (CoT) prompting. However, recent work has found that ICL relies mostly on the retrieval of task priors and less so on "learning" to perform tasks, especially for complex subjective domains like emotion and morality, where priors ossify posterior predictions. In this work, we examine whether "enabling" reasoning also creates the same behavior in LLMs, wherein the format of CoT retrieves reasoning priors that remain relatively unchanged despite the evidence in the prompt. We find that, surprisingly, CoT indeed suffers from the same posterior collapse as ICL for larger language models. Code is avalaible at https://github.com/gchochla/cot-priors.
comment: 5 pages, 2 figures, 1 table. arXiv admin note: text overlap with arXiv:2403.17125
♻ ☆ Natural Language Processing Methods for the Study of Protein-Ligand Interactions
Recent advances in Natural Language Processing (NLP) have ignited interest in developing effective methods for predicting protein-ligand interactions (PLIs) given their relevance to drug discovery and protein engineering efforts and the ever-growing volume of biochemical sequence and structural data available. The parallels between human languages and the "languages" used to represent proteins and ligands have enabled the use of NLP machine learning approaches to advance PLI studies. In this review, we explain where and how such approaches have been applied in the recent literature and discuss useful mechanisms such as long short-term memory, transformers, and attention. We conclude with a discussion of the current limitations of NLP methods for the study of PLIs as well as key challenges that need to be addressed in future work.
comment: 52 Pages and 3 Figures
♻ ☆ Can Large Language Models Generate High-quality Patent Claims?
Large language models (LLMs) have shown exceptional performance across various text generation tasks but remain under-explored in the patent domain, which offers highly structured and precise language. This paper constructs a dataset to investigate the performance of current LLMs in patent claim generation. Our results demonstrate that generating claims based on patent descriptions outperforms previous research relying on abstracts. Interestingly, current patent-specific LLMs perform much worse than state-of-the-art general LLMs, highlighting the necessity for future research on in-domain LLMs. We also find that LLMs can produce high-quality first independent claims, but their performances markedly decrease for subsequent dependent claims. Moreover, fine-tuning can enhance the completeness of inventions' features, conceptual clarity, and feature linkage. Among the tested LLMs, GPT-4 demonstrates the best performance in comprehensive human evaluations by patent experts, with better feature coverage, conceptual clarity, and technical coherence. Despite these capabilities, comprehensive revision and modification are still necessary to pass rigorous patent scrutiny and ensure legal robustness.
comment: 16 pages, 2 figures, 12 tables
♻ ☆ Modeling Human Subjectivity in LLMs Using Explicit and Implicit Human Factors in Personas EMNLP 2024
Large language models (LLMs) are increasingly being used in human-centered social scientific tasks, such as data annotation, synthetic data creation, and engaging in dialog. However, these tasks are highly subjective and dependent on human factors, such as one's environment, attitudes, beliefs, and lived experiences. Thus, it may be the case that employing LLMs (which do not have such human factors) in these tasks results in a lack of variation in data, failing to reflect the diversity of human experiences. In this paper, we examine the role of prompting LLMs with human-like personas and asking the models to answer as if they were a specific human. This is done explicitly, with exact demographics, political beliefs, and lived experiences, or implicitly via names prevalent in specific populations. The LLM personas are then evaluated via (1) subjective annotation task (e.g., detecting toxicity) and (2) a belief generation task, where both tasks are known to vary across human factors. We examine the impact of explicit vs. implicit personas and investigate which human factors LLMs recognize and respond to. Results show that explicit LLM personas show mixed results when reproducing known human biases, but generally fail to demonstrate implicit biases. We conclude that LLMs may capture the statistical patterns of how people speak, but are generally unable to model the complex interactions and subtleties of human perceptions, potentially limiting their effectiveness in social science applications.
comment: Accepted at Findings of EMNLP 2024
♻ ☆ uDistil-Whisper: Label-Free Data Filtering for Knowledge Distillation in Low-Data Regimes
Recent work on distilling Whisper's knowledge into small models using pseudo-labels shows promising performance while reducing the size by up to 50\%. This results in small, efficient, and dedicated models. However, a critical step of distillation from pseudo-labels involves filtering high-quality predictions and using only those during training. This step requires ground truth labels to compare and filter low-quality examples making the whole process supervised. In addition to that, the distillation process requires a large amount of data thereby limiting the ability to distill models in low-resource settings. To address this challenge, we propose a distillation framework that does not require any labeled data. Through experimentation, we show that our best distilled models outperform the teacher model by 5-7 points in terms of WER compared to those without filtering and are on par with or perform better than similar supervised data filtering setups. When we scale the data, our models significantly outperform all zero-shot and supervised models. We demonstrate that it is possible to distill large Whisper models into relatively small ones without using any labeled data. Our distilled models are also 25-50\% more compute- and memory-efficient while maintaining performance equal to or better than that of the teacher model.
comment: Work in progress
♻ ☆ K-Level Reasoning: Establishing Higher Order Beliefs in Large Language Models for Strategic Reasoning
Strategic reasoning is a complex yet essential capability for intelligent agents. It requires Large Language Model (LLM) agents to adapt their strategies dynamically in multi-agent environments. Unlike static reasoning tasks, success in these contexts depends on anticipating other agents' beliefs and actions while continuously adjusting strategies to achieve individual goals. LLMs and LLM agents often struggle with strategic reasoning due to the absence of a reasoning framework that enables them to dynamically infer others' perspectives and adapt to changing environments. Inspired by the Level-K framework from game theory and behavioral economics, which extends reasoning from simple reactions to structured strategic depth, we propose a novel framework: "K-Level Reasoning with Large Language Models (K-R)." This framework employs recursive mechanisms to enable LLMs to achieve varying levels of strategic depth, allowing agents to form higher order beliefs - beliefs about others' beliefs. We validate this framework through rigorous testing on four testbeds: two classical game theory problems and two social intelligence tasks. The results demonstrate the advantages of K-R in strategic reasoning. Our work presents the first recursive implementation of strategic depth in large language models (LLMs). It establishes a foundation for future research into theory of mind and strategic reasoning in LLMs.
♻ ☆ Beyond Coarse-Grained Matching in Video-Text Retrieval ACCV 2024
Video-text retrieval has seen significant advancements, yet the ability of models to discern subtle differences in captions still requires verification. In this paper, we introduce a new approach for fine-grained evaluation. Our approach can be applied to existing datasets by automatically generating hard negative test captions with subtle single-word variations across nouns, verbs, adjectives, adverbs, and prepositions. We perform comprehensive experiments using four state-of-the-art models across two standard benchmarks (MSR-VTT and VATEX) and two specially curated datasets enriched with detailed descriptions (VLN-UVO and VLN-OOPS), resulting in a number of novel insights: 1) our analyses show that the current evaluation benchmarks fall short in detecting a model's ability to perceive subtle single-word differences, 2) our fine-grained evaluation highlights the difficulty models face in distinguishing such subtle variations. To enhance fine-grained understanding, we propose a new baseline that can be easily combined with current methods. Experiments on our fine-grained evaluations demonstrate that this approach enhances a model's ability to understand fine-grained differences.
comment: Accepted to ACCV 2024
♻ ☆ Understanding and Mitigating Language Confusion in LLMs EMNLP 2024
We investigate a surprising limitation of LLMs: their inability to consistently generate text in a user's desired language. We create the Language Confusion Benchmark (LCB) to evaluate such failures, covering 15 typologically diverse languages with existing and newly-created English and multilingual prompts. We evaluate a range of LLMs on monolingual and cross-lingual generation reflecting practical use cases, finding that Llama Instruct and Mistral models exhibit high degrees of language confusion and even the strongest models fail to consistently respond in the correct language. We observe that base and English-centric instruct models are more prone to language confusion, which is aggravated by complex prompts and high sampling temperatures. We find that language confusion can be partially mitigated via few-shot prompting, multilingual SFT and preference tuning. We release our language confusion benchmark, which serves as a first layer of efficient, scalable multilingual evaluation at https://github.com/for-ai/language-confusion.
comment: EMNLP 2024 Main Conference Camera-ready
♻ ☆ ShadowLLM: Predictor-based Contextual Sparsity for Large Language Models EMNLP 2024
The high power consumption and latency-sensitive deployments of large language models (LLMs) have motivated efficiency techniques like quantization and sparsity. Contextual sparsity, where the sparsity pattern is input-dependent, is crucial in LLMs because the permanent removal of attention heads or neurons from LLMs can significantly degrade accuracy. Prior work has attempted to model contextual sparsity using neural networks trained to predict activation magnitudes, which can be used to dynamically prune structures with low predicted activation magnitude. In this paper, we look beyond magnitude-based pruning criteria to assess attention head and neuron importance in LLMs. We develop a novel predictor called ShadowLLM, which can shadow the LLM behavior and enforce better sparsity patterns, resulting in over 15% improvement in end-to-end accuracy compared to prior methods. In addition, ShadowLLM achieves up to a 20% speed-up over the state-of-the-art DejaVu framework. These enhancements are validated on Llama-2 and OPT models with up to 30 billion parameters. Our code is available at \href{https://github.com/abdelfattah-lab/shadow_llm/}{ShadowLLM}.
comment: Accepted to EMNLP 2024 (Main, Long Paper)
♻ ☆ Block-Attention for Efficient RAG
We introduce Block-Attention, an attention mechanism designed to address the increased inference latency and cost in Retrieval-Augmented Generation (RAG) scenarios. Traditional approaches often encode the entire context. Instead, Block-Attention divides retrieved documents into discrete blocks, with each block independently calculating key-value (KV) states except for the final block. In RAG scenarios, by defining each passage as a block, Block-Attention enables us to reuse the KV states of passages that have been seen before, thereby significantly reducing the latency and the computation overhead during inference. The implementation of Block-Attention involves block segmentation, position re-encoding, and fine-tuning the LLM to adapt to the Block-Attention mechanism. Experiments on four RAG benchmarks demonstrate that after block fine-tuning, the Block-Attention model achieves performance comparable to self-attention models (68.4\% vs 67.9\% on Llama3) or even superior performance (62.8\% vs 59.6\% on Mistral). Notably, Block-Attention significantly reduces the time to first token (TTFT) and floating point operations (FLOPs) to a very low level. It only takes 45 ms to output the first token for an input sequence with a total length of 32K. Compared to the self-attention models, the time consumption and corresponding FLOPs are reduced by 98.7\% and 99.8\%, respectively.
♻ ☆ Prompt-SAW: Leveraging Relation-Aware Graphs for Textual Prompt Compression
Large Language Models (LLMs) have shown exceptional abilities for multiple different natural language processing tasks. While prompting is a crucial tool for LLM inference, we observe that there is a significant cost associated with exceedingly lengthy prompts. Existing attempts to compress lengthy prompts lead to substandard results in terms of readability/interpretability of the compressed prompt, with a detrimental impact on prompt utility. To address this, we propose PromptSAW: Prompt compresSion via Relation AWare graphs, an effective strategy for prompt compression over task-agnostic and task-aware prompts. Prompt-SAW uses the prompt's textual information to build a graph and later extracts key information elements in the graph to come up with the compressed prompt. We also propose GSM8K-aug, i.e., an extended version of the existing GSM8K benchmark for task-agnostic prompts in order to provide a comprehensive evaluation platform. Experimental evaluation using benchmark datasets shows that prompts compressed by Prompt-SAW are not only better in terms of readability, but they also outperform the best-performing baseline models by up to 10.1 and 77.1, respectively, for task-agnostic and task-aware settings while compressing the original prompt text by 34.9 and 56.7.
comment: 16 pages
♻ ☆ A Systematic Analysis of Large Language Models as Soft Reasoners: The Case of Syllogistic Inferences EMNLP 2024
The reasoning abilities of Large Language Models (LLMs) are becoming a central focus of study in NLP. In this paper, we consider the case of syllogistic reasoning, an area of deductive reasoning studied extensively in logic and cognitive psychology. Previous research has shown that pre-trained LLMs exhibit reasoning biases, such as $\textit{content effects}$, avoid answering that $\textit{no conclusion follows}$, display human-like difficulties, and struggle with multi-step reasoning. We contribute to this research line by systematically investigating the effects of chain-of-thought reasoning, in-context learning (ICL), and supervised fine-tuning (SFT) on syllogistic reasoning, considering syllogisms with conclusions that support or violate world knowledge, as well as ones with multiple premises. Crucially, we go beyond the standard focus on accuracy, with an in-depth analysis of the conclusions generated by the models. Our results suggest that the behavior of pre-trained LLMs can be explained by heuristics studied in cognitive science and that both ICL and SFT improve model performance on valid inferences, although only the latter mitigates most reasoning biases without harming model consistency.
comment: Accepted to EMNLP 2024 (main conference)
♻ ☆ Unmasking Database Vulnerabilities: Zero-Knowledge Schema Inference Attacks in Text-to-SQL Systems
Text-to-SQL systems empower users to interact with databases using natural language, automatically translating queries into executable SQL code. However, their reliance on database schema information for SQL generation exposes them to significant security vulnerabilities, particularly schema inference attacks that can lead to unauthorized data access or manipulation. In this paper, we introduce a novel zero-knowledge framework for reconstructing the underlying database schema of text-to-SQL models without any prior knowledge of the database. Our approach systematically probes text-to-SQL models with specially crafted questions and leverages a surrogate GPT-4 model to interpret the outputs, effectively uncovering hidden schema elements -- including tables, columns, and data types. We demonstrate that our method achieves high accuracy in reconstructing table names, with F1 scores of up to .99 for generative models and .78 for fine-tuned models, underscoring the severity of schema leakage risks. Furthermore, we propose a simple protection mechanism for generative models and empirically show its limitations in mitigating these attacks.
♻ ☆ BLT: Can Large Language Models Handle Basic Legal Text?
We find that the best publicly available LLMs like GPT-4 and Claude currently perform poorly on basic legal text handling. This motivates the creation of a benchmark consisting of examples that lawyers and paralegals would expect LLMs to handle zero-shot, such as looking up the text at a line of a witness deposition or at a subsection of a contract. LLMs' poor performance on this benchmark casts into doubt their reliability as-is for legal practice. However, fine-tuning on our training set brings even a small model to near-perfect performance. This benchmark will be useful for fine-tuning LLMs for downstream legal tasks, as well as for tracking LLMs' reliability as-is for basic legal tasks.
♻ ☆ Towards Inducing Document-Level Abilities in Standard Multilingual Neural Machine Translation Models
Neural Machine Translation (NMT) models have traditionally used Sinusoidal Positional Embeddings (PEs), which often struggle to capture long-range dependencies and are less efficient for handling extended context or document-level translation tasks. This work addresses the challenge of transitioning pre-trained NMT models from absolute sinusoidal PEs to relative PEs, such as Rotary Positional Embeddings (ROPE) and Attention with Linear Biases (ALIBI), without compromising performance. We demonstrate that parameter-efficient fine-tuning, using only a small amount of high-quality data, can successfully facilitate this transition. Experimental results indicate that switching from sinusoidal to relative PEs results in competitive translation quality on sentence-level evaluation benchmarks. Additionally, models trained with ROPE consistently outperform those using ALIBI and Sinusoidal PEs on document-level benchmarks across both string-based metrics and qualitative evaluations. Moreover, we find that a small amount of long-context data in a few languages is sufficient for cross-lingual length generalization, thereby inducing long-context capabilities.
comment: Under Review
♻ ☆ Granular Privacy Control for Geolocation with Vision Language Models EMNLP 2024
Vision Language Models (VLMs) are rapidly advancing in their capability to answer information-seeking questions. As these models are widely deployed in consumer applications, they could lead to new privacy risks due to emergent abilities to identify people in photos, geolocate images, etc. As we demonstrate, somewhat surprisingly, current open-source and proprietary VLMs are very capable image geolocators, making widespread geolocation with VLMs an immediate privacy risk, rather than merely a theoretical future concern. As a first step to address this challenge, we develop a new benchmark, GPTGeoChat, to test the ability of VLMs to moderate geolocation dialogues with users. We collect a set of 1,000 image geolocation conversations between in-house annotators and GPT-4v, which are annotated with the granularity of location information revealed at each turn. Using this new dataset, we evaluate the ability of various VLMs to moderate GPT-4v geolocation conversations by determining when too much location information has been revealed. We find that custom fine-tuned models perform on par with prompted API-based models when identifying leaked location information at the country or city level; however, fine-tuning on supervised data appears to be needed to accurately moderate finer granularities, such as the name of a restaurant or building.
comment: Accepted to EMNLP 2024 main conference
♻ ☆ Human and LLM Biases in Hate Speech Annotations: A Socio-Demographic Analysis of Annotators and Targets
The rise of online platforms exacerbated the spread of hate speech, demanding scalable and effective detection. However, the accuracy of hate speech detection systems heavily relies on human-labeled data, which is inherently susceptible to biases. While previous work has examined the issue, the interplay between the characteristics of the annotator and those of the target of the hate are still unexplored. We fill this gap by leveraging an extensive dataset with rich socio-demographic information of both annotators and targets, uncovering how human biases manifest in relation to the target's attributes. Our analysis surfaces the presence of widespread biases, which we quantitatively describe and characterize based on their intensity and prevalence, revealing marked differences. Furthermore, we compare human biases with those exhibited by persona-based LLMs. Our findings indicate that while persona-based LLMs do exhibit biases, these differ significantly from those of human annotators. Overall, our work offers new and nuanced results on human biases in hate speech annotations, as well as fresh insights into the design of AI-driven hate speech detection systems.
♻ ☆ Efficient In-Domain Question Answering for Resource-Constrained Environments
Retrieval Augmented Generation (RAG) is a common method for integrating external knowledge into pretrained Large Language Models (LLMs) to enhance accuracy and relevancy in question answering (QA) tasks. However, prompt engineering and resource efficiency remain significant bottlenecks in developing optimal and robust RAG solutions for real-world QA applications. Recent studies have shown success in using fine tuning to address these problems; in particular, Retrieval Augmented Fine Tuning (RAFT) applied to smaller 7B models has demonstrated superior performance compared to RAG setups with much larger models such as GPT-3.5. The combination of RAFT with parameter-efficient fine tuning (PEFT) techniques, such as Low-Rank Adaptation (LoRA), promises an even more efficient solution, yet remains an unexplored area. In this work, we combine RAFT with LoRA to reduce fine tuning and storage requirements and gain faster inference times while maintaining comparable RAG performance. This results in a more compute-efficient RAFT, or CRAFT, which is particularly useful for knowledge-intensive QA tasks in resource-constrained environments where internet access may be restricted and hardware resources limited.
comment: 6 pages, 2 tables
♻ ☆ LayerSkip: Enabling Early Exit Inference and Self-Speculative Decoding ACL 2024
We present LayerSkip, an end-to-end solution to speed-up inference of large language models (LLMs). First, during training we apply layer dropout, with low dropout rates for earlier layers and higher dropout rates for later layers, and an early exit loss where all transformer layers share the same exit. Second, during inference, we show that this training recipe increases the accuracy of early exit at earlier layers, without adding any auxiliary layers or modules to the model. Third, we present a novel self-speculative decoding solution where we exit at early layers and verify and correct with remaining layers of the model. Our proposed self-speculative decoding approach has less memory footprint than other speculative decoding approaches and benefits from shared compute and activations of the draft and verification stages. We run experiments on different Llama model sizes on different types of training: pretraining from scratch, continual pretraining, finetuning on specific data domain, and finetuning on specific task. We implement our inference solution and show speedups of up to 2.16x on summarization for CNN/DM documents, 1.82x on coding, and 2.0x on TOPv2 semantic parsing task. We open source our code and checkpoints at https://github.com/facebookresearch/LayerSkip.
comment: ACL 2024
♻ ☆ Building Better: Avoiding Pitfalls in Developing Language Resources when Data is Scarce
Language is a symbolic capital that affects people's lives in many ways (Bourdieu, 1977, 1991). It is a powerful tool that accounts for identities, cultures, traditions, and societies in general. Hence, data in a given language should be viewed as more than a collection of tokens. Good data collection and labeling practices are key to building more human-centered and socially aware technologies. While there has been a rising interest in mid- to low-resource languages within the NLP community, work in this space has to overcome unique challenges such as data scarcity and access to suitable annotators. In this paper, we collect feedback from those directly involved in and impacted by NLP artefacts for mid- to low-resource languages. We conduct a quantitative and qualitative analysis of the responses and highlight the main issues related to (1) data quality such as linguistic and cultural data suitability; and (2) the ethics of common annotation practices such as the misuse of online community services. Based on these findings, we make several recommendations for the creation of high-quality language artefacts that reflect the cultural milieu of its speakers, while simultaneously respecting the dignity and labor of data workers.
♻ ☆ LLM-based Cognitive Models of Students with Misconceptions
Accurately modeling student cognition is crucial for developing effective AI-driven educational technologies. A key challenge is creating realistic student models that satisfy two essential properties: (1) accurately replicating specific misconceptions, and (2) correctly solving problems where these misconceptions are not applicable. This dual requirement reflects the complex nature of student understanding, where misconceptions coexist with correct knowledge. This paper investigates whether Large Language Models (LLMs) can be instruction-tuned to meet this dual requirement and effectively simulate student thinking in algebra. We introduce MalAlgoPy, a novel Python library that generates datasets reflecting authentic student solution patterns through a graph-based representation of algebraic problem-solving. Utilizing MalAlgoPy, we define and examine Cognitive Student Models (CSMs) - LLMs instruction tuned to faithfully emulate realistic student behavior. Our findings reveal that LLMs trained on misconception examples can efficiently learn to replicate errors. However, the training diminishes the model's ability to solve problems correctly, particularly for problem types where the misconceptions are not applicable, thus failing to satisfy second property of CSMs. We demonstrate that by carefully calibrating the ratio of correct to misconception examples in the training data - sometimes as low as 0.25 - it is possible to develop CSMs that satisfy both properties. Our insights enhance our understanding of AI-based student models and pave the way for effective adaptive learning systems.
♻ ☆ MuJo: Multimodal Joint Feature Space Learning for Human Activity Recognition
Human Activity Recognition (HAR) is a longstanding problem in AI with applications in a broad range of areas, including healthcare, sports and fitness, security, and more. The performance of HAR in real-world settings is strongly dependent on the type and quality of the input signal that can be acquired. Given an unobstructed, high-quality camera view of a scene, computer vision systems, in particular in conjunction with foundation models, can today fairly reliably distinguish complex activities. On the other hand, recognition using modalities such as wearable sensors (which are often more broadly available, e.g., in mobile phones and smartwatches) is a more difficult problem, as the signals often contain less information and labeled training data is more difficult to acquire. To alleviate the need for labeled data, we introduce our comprehensive Fitness Multimodal Activity Dataset (FiMAD) in this work, which can be used with the proposed pre-training method MuJo (Multimodal Joint Feature Space Learning) to enhance HAR performance across various modalities. FiMAD was created using YouTube fitness videos and contains parallel video, language, pose, and simulated IMU sensor data. MuJo utilizes this dataset to learn a joint feature space for these modalities. We show that classifiers pre-trained on FiMAD can increase the performance on real HAR datasets such as MM-Fit, MyoGym, MotionSense, and MHEALTH. For instance, on MM-Fit, we achieve an Macro F1-Score of up to 0.855 when fine-tuning on only 2% of the training data and 0.942 when utilizing the full training set for classification tasks. We have compared our approach to other self-supervised ones and showed that, unlike them, ours can consistently improve on the baseline network performance as well as provide a better data-efficiency.
♻ ☆ Learning to Ask Informative Questions: Enhancing LLMs with Preference Optimization and Expected Information Gain EMNLP 2024
Questions are essential tools for acquiring the necessary information to complete information-seeking tasks. However, large language models (LLMs), especially open-source models, often perform poorly in generating informative questions, as measured by expected information gain (EIG). In this paper, we propose a method to enhance the informativeness of LLM-generated questions in 20-question game dialogues. We sample multiple questions from the same model (LLAMA 2-CHAT 7B) for each game and create pairs of low-EIG and high-EIG questions to apply a Direct Preference Optimization (DPO) algorithm. Our results show that this method produces more effective questions (in terms of EIG), even in domains different from those used to train the DPO model.
comment: Accepted to EMNLP 2024 (Findings)
♻ ☆ Relay Decoding: Concatenating Large Language Models for Machine Translation
Leveraging large language models for machine translation has demonstrated promising results. However, it does require the large language models to possess the capability of handling both the source and target languages in machine translation. When it is challenging to find large models that support the desired languages, resorting to continuous learning methods becomes a costly endeavor. To mitigate these expenses, we propose an innovative approach called RD (Relay Decoding), which entails concatenating two distinct large models that individually support the source and target languages. By incorporating a simple mapping layer to facilitate the connection between these two models and utilizing a limited amount of parallel data for training, we successfully achieve superior results in the machine translation task. Experimental results conducted on the Multi30k and WikiMatrix datasets validate the effectiveness of our proposed method.
comment: Work in progress
♻ ☆ On the Reliability of Large Language Models to Misinformed and Demographically-Informed Prompts AAAI
We investigate and observe the behaviour and performance of Large Language Model (LLM)-backed chatbots in addressing misinformed prompts and questions with demographic information within the domains of Climate Change and Mental Health. Through a combination of quantitative and qualitative methods, we assess the chatbots' ability to discern the veracity of statements, their adherence to facts, and the presence of bias or misinformation in their responses. Our quantitative analysis using True/False questions reveals that these chatbots can be relied on to give the right answers to these close-ended questions. However, the qualitative insights, gathered from domain experts, shows that there are still concerns regarding privacy, ethical implications, and the necessity for chatbots to direct users to professional services. We conclude that while these chatbots hold significant promise, their deployment in sensitive areas necessitates careful consideration, ethical oversight, and rigorous refinement to ensure they serve as a beneficial augmentation to human expertise rather than an autonomous solution.
comment: Study conducted between August and December 2023. Under review at AAAI-AI Magazine. Submitted for archival purposes only
♻ ☆ Beyond Thumbs Up/Down: Untangling Challenges of Fine-Grained Feedback for Text-to-Image Generation
Human feedback plays a critical role in learning and refining reward models for text-to-image generation, but the optimal form the feedback should take for learning an accurate reward function has not been conclusively established. This paper investigates the effectiveness of fine-grained feedback which captures nuanced distinctions in image quality and prompt-alignment, compared to traditional coarse-grained feedback (for example, thumbs up/down or ranking between a set of options). While fine-grained feedback holds promise, particularly for systems catering to diverse societal preferences, we show that demonstrating its superiority to coarse-grained feedback is not automatic. Through experiments on real and synthetic preference data, we surface the complexities of building effective models due to the interplay of model choice, feedback type, and the alignment between human judgment and computational interpretation. We identify key challenges in eliciting and utilizing fine-grained feedback, prompting a reassessment of its assumed benefits and practicality. Our findings -- e.g., that fine-grained feedback can lead to worse models for a fixed budget, in some settings; however, in controlled settings with known attributes, fine grained rewards can indeed be more helpful -- call for careful consideration of feedback attributes and potentially beckon novel modeling approaches to appropriately unlock the potential value of fine-grained feedback in-the-wild.
♻ ☆ InferAct: Inferring Safe Actions for LLM-Based Agents Through Preemptive Evaluation and Human Feedback
A crucial requirement for deploying LLM-based agents in real-life applications is the robustness against risky or even irreversible mistakes. However, the existing research lacks a focus on preemptive evaluation of reasoning trajectories performed by LLM agents, leading to a gap in ensuring safe and reliable operations. To explore better solutions, this paper introduces InferAct, a novel approach that leverages the belief reasoning ability of LLMs, grounded in Theory-of-Mind, to proactively detect potential errors before risky actions are executed (e.g., `buy-now' in automatic online trading or web shopping). InferAct acts as a human proxy, detecting unsafe actions and alerting users for intervention, which helps prevent irreversible risks in time and enhances the actor agent's decision-making process. Experiments on three widely-used tasks demonstrate the effectiveness of InferAct, presenting a novel solution for safely developing LLM agents in environments involving critical decision-making.
♻ ☆ Pyramid-Driven Alignment: Pyramid Principle Guided Integration of Large Language Models and Knowledge Graphs
Large Language Models (LLMs) possess impressive reasoning abilities but are prone to generating incorrect information, often referred to as hallucinations. While incorporating external Knowledge Graphs (KGs) can partially mitigate this issue, existing methods primarily treat KGs as static knowledge repositories, overlooking the critical disparity between KG and LLM knowledge, and failing to fully exploit the reasoning capabilities inherent in KGs. To address these limitations, we propose Pyramid-Driven Alignment (PDA), a novel framework for seamlessly integrating LLMs with KGs. PDA utilizes Pyramid Principle analysis to construct a hierarchical pyramid structure. This structure is designed to reflect the input question and generate more validated deductive knowledge, thereby enhancing the alignment of LLMs and KGs and ensuring more cohesive integration. Furthermore, PDA employs a recursive mechanism to harness the underlying reasoning abilities of KGs, resulting in more accurate knowledge retrieval for question-answering tasks. Our experimental results reveal a substantial performance advantage of PDA over state-of-the-art baselines, with improvements reaching 26.70% and 26.78%.
♻ ☆ Autonomous Agents for Collaborative Task under Information Asymmetry NeurIPS 2024
Large Language Model Multi-Agent Systems (LLM-MAS) have achieved great progress in solving complex tasks. It performs communication among agents within the system to collaboratively solve tasks, under the premise of shared information. However, when agents' collaborations are leveraged to perform multi-person tasks, a new challenge arises due to information asymmetry, since each agent can only access the information of its human user. Previous MAS struggle to complete tasks under this condition. To address this, we propose a new MAS paradigm termed iAgents, which denotes Informative Multi-Agent Systems. In iAgents, the human social network is mirrored in the agent network, where agents proactively exchange human information necessary for task resolution, thereby overcoming information asymmetry. iAgents employs a novel agent reasoning mechanism, InfoNav, to navigate agents' communication toward effective information exchange. Together with InfoNav, iAgents organizes human information in a mixed memory to provide agents with accurate and comprehensive information for exchange. Additionally, we introduce InformativeBench, the first benchmark tailored for evaluating LLM agents' task-solving ability under information asymmetry. Experimental results show that iAgents can collaborate within a social network of 140 individuals and 588 relationships, autonomously communicate over 30 turns, and retrieve information from nearly 70,000 messages to complete tasks within 3 minutes.
comment: 32 pages, 12 figures, 6 tables, accepted by NeurIPS 2024, see detail at https://thinkwee.top/iagents
♻ ☆ MedAide: Towards an Omni Medical Aide via Specialized LLM-based Multi-Agent Collaboration
Large Language Model (LLM)-driven interactive systems currently show potential promise in healthcare domains. Despite their remarkable capabilities, LLMs typically lack personalized recommendations and diagnosis analysis in sophisticated medical applications, causing hallucinations and performance bottlenecks. To address these challenges, this paper proposes MedAide, an LLM-based omni medical multi-agent collaboration framework for specialized healthcare services. Specifically, MedAide first performs query rewriting through retrieval-augmented generation to accomplish accurate medical intent understanding. Immediately, we devise a contextual encoder to obtain intent prototype embeddings, which are used to recognize fine-grained intents by similarity matching. According to the intent relevance, the activated agents collaborate effectively to provide integrated decision analysis. Extensive experiments are conducted on four medical benchmarks with composite intents. Experimental results from automated metrics and expert doctor evaluations show that MedAide outperforms current LLMs and improves their medical proficiency and strategic reasoning.
comment: LLM-based Multi-Agent Collaboration for Medical Applications
♻ ☆ Skeleton: A New Framework for Accelerating Language Models via Task Neuron Localized Prompt Tuning
Prompt tuning methods have shown comparable performance to general training methods as parameter-efficient fine-tuning (PEFT) methods in various natural language understanding tasks. However, existing prompt tuning methods still utilize the entire model architecture even when solving a specific task, which prevents them from accelerating inference speed during the application procedure. In this paper, we propose a novel prompt tuning framework called Skeleton to efficiently utilize a language model in terms of memory and time complexity for solving various tasks, retaining only task-relevant neurons by using an explainability method. From our framework, we can efficiently solve various tasks by using only task-relevant neurons and prepending adequate task-specific prompt tokens with only a single language model. Experiments reveal that our method significantly enhances inference efficiency (at most x 1.73 speed up) for various widely used benchmarks, showing comparable performance to the prompt tuning method. Moreover, our method is applicable across various transformer-based architectures, confirming its practicality and scalability.
comment: 11 pages
♻ ☆ LLoCO: Learning Long Contexts Offline EMNLP 2024
Processing long contexts remains a challenge for large language models (LLMs) due to the quadratic computational and memory overhead of the self-attention mechanism and the substantial KV cache sizes during generation. We propose LLoCO, a novel approach to address this problem by learning contexts offline through context compression and in-domain parameter-efficient finetuning with LoRA. Our method enables an LLM to create a concise representation of the original context and efficiently retrieve relevant information to answer questions accurately. Our approach extends the effective context window of a 4k token LLaMA2-7B model to handle up to 128k tokens. We evaluate our approach on several long-context question-answering datasets, demonstrating that LLoCO significantly outperforms in-context learning while using $30\times$ fewer tokens during inference. LLoCO achieves up to $7.62\times$ speed-up during inference and $11.52\times$ higher throughput during finetuning, substantially reduces the cost of long document question answering. This makes it a promising solution for efficient long context processing. Our code is publicly available on https://github.com/jeffreysijuntan/lloco.
comment: EMNLP 2024. The first two authors contributed equally to this work
♻ ☆ Are Large Language Models Good Classifiers? A Study on Edit Intent Classification in Scientific Document Revisions EMNLP2024
Classification is a core NLP task architecture with many potential applications. While large language models (LLMs) have brought substantial advancements in text generation, their potential for enhancing classification tasks remains underexplored. To address this gap, we propose a framework for thoroughly investigating fine-tuning LLMs for classification, including both generation- and encoding-based approaches. We instantiate this framework in edit intent classification (EIC), a challenging and underexplored classification task. Our extensive experiments and systematic comparisons with various training approaches and a representative selection of LLMs yield new insights into their application for EIC. We investigate the generalizability of these findings on five further classification tasks. To demonstrate the proposed methods and address the data shortage for empirical edit analysis, we use our best-performing EIC model to create Re3-Sci2.0, a new large-scale dataset of 1,780 scientific document revisions with over 94k labeled edits. The quality of the dataset is assessed through human evaluation. The new dataset enables an in-depth empirical study of human editing behavior in academic writing. We make our experimental framework, models and data publicly available.
comment: EMNLP2024 Main
♻ ☆ From Measurement Instruments to Data: Leveraging Theory-Driven Synthetic Training Data for Classifying Social Constructs
Computational text classification is a challenging task, especially for multi-dimensional social constructs. Recently, there has been increasing discussion that synthetic training data could enhance classification by offering examples of how these constructs are represented in texts. In this paper, we systematically examine the potential of theory-driven synthetic training data for improving the measurement of social constructs. In particular, we explore how researchers can transfer established knowledge from measurement instruments in the social sciences, such as survey scales or annotation codebooks, into theory-driven generation of synthetic data. Using two studies on measuring sexism and political topics, we assess the added value of synthetic training data for fine-tuning text classification models. Although the results of the sexism study were less promising, our findings demonstrate that synthetic data can be highly effective in reducing the need for labeled data in political topic classification. With only a minimal drop in performance, synthetic data allows for substituting large amounts of labeled data. Furthermore, theory-driven synthetic data performed markedly better than data generated without conceptual information in mind.
♻ ☆ Pragmatic Competence Evaluation of Large Language Models for the Korean Language
Benchmarks play a significant role in the current evaluation of Large Language Models (LLMs), yet they often overlook the models' abilities to capture the nuances of human language, primarily focusing on evaluating embedded knowledge and technical skills. To address this gap, our study evaluates how well LLMs understand context-dependent expressions from a pragmatic standpoint, specifically in Korean. We use both Multiple-Choice Questions (MCQs) for automatic evaluation and Open-Ended Questions (OEQs) assessed by human experts. Our results show that GPT-4 leads with scores of 81.11 in MCQs and 85.69 in OEQs, closely followed by HyperCLOVA X. Additionally, while few-shot learning generally improves performance, Chain-of-Thought (CoT) prompting tends to encourage literal interpretations, which may limit effective pragmatic inference. Our findings highlight the need for LLMs to better understand and generate language that reflects human communicative norms.
comment: 38th Pacific Asia Conference on Language, Information and Computation
♻ ☆ LightPAL: Lightweight Passage Retrieval for Open Domain Multi-Document Summarization
Open-Domain Multi-Document Summarization (ODMDS) is the task of generating summaries from large document collections in response to user queries. This task is crucial for efficiently addressing diverse information needs from users. Traditional retrieve-then-summarize approaches fall short for open-ended queries in ODMDS tasks. These queries often require broader context than initially retrieved passages provide, making it challenging to retrieve all relevant information in a single search. While iterative retrieval methods has been explored for multi-hop question answering (MQA), it's impractical for ODMDS due to high latency from repeated LLM inference. Accordingly, we propose LightPAL, a lightweight passage retrieval method for ODMDS. LightPAL leverages an LLM to pre-construct a graph representing passage relationships, then employs random walk during retrieval, avoiding iterative LLM inference. Experiments demonstrate that LightPAL outperforms naive sparse and pre-trained dense retrievers in both retrieval and summarization metrics, while achieving higher efficiency compared to iterative MQA approaches.
comment: 15 pages, 7 figures, 6 tables
♻ ☆ SafeGen: Mitigating Sexually Explicit Content Generation in Text-to-Image Models CCS 2024
Text-to-image (T2I) models, such as Stable Diffusion, have exhibited remarkable performance in generating high-quality images from text descriptions in recent years. However, text-to-image models may be tricked into generating not-safe-for-work (NSFW) content, particularly in sexually explicit scenarios. Existing countermeasures mostly focus on filtering inappropriate inputs and outputs, or suppressing improper text embeddings, which can block sexually explicit content (e.g., naked) but may still be vulnerable to adversarial prompts -- inputs that appear innocent but are ill-intended. In this paper, we present SafeGen, a framework to mitigate sexual content generation by text-to-image models in a text-agnostic manner. The key idea is to eliminate explicit visual representations from the model regardless of the text input. In this way, the text-to-image model is resistant to adversarial prompts since such unsafe visual representations are obstructed from within. Extensive experiments conducted on four datasets and large-scale user studies demonstrate SafeGen's effectiveness in mitigating sexually explicit content generation while preserving the high-fidelity of benign images. SafeGen outperforms eight state-of-the-art baseline methods and achieves 99.4% sexual content removal performance. Furthermore, our constructed benchmark of adversarial prompts provides a basis for future development and evaluation of anti-NSFW-generation methods.
comment: Accepted by ACM CCS 2024. Please cite this paper as "Xinfeng Li, Yuchen Yang, Jiangyi Deng, Chen Yan, Yanjiao Chen, Xiaoyu Ji, Wenyuan Xu. SafeGen: Mitigating Sexually Explicit Content Generation in Text-to-Image Models. In Proceedings of ACM Conference on Computer and Communications Security (CCS), 2024."
♻ ☆ SpreadsheetBench: Towards Challenging Real World Spreadsheet Manipulation
We introduce SpreadsheetBench, a challenging spreadsheet manipulation benchmark exclusively derived from real-world scenarios, designed to immerse current large language models (LLMs) in the actual workflow of spreadsheet users. Unlike existing benchmarks that rely on synthesized queries and simplified spreadsheet files, SpreadsheetBench is built from 912 real questions gathered from online Excel forums, which reflect the intricate needs of users. The associated spreadsheets from the forums contain a variety of tabular data such as multiple tables, non-standard relational tables, and abundant non-textual elements. Furthermore, we propose a more reliable evaluation metric akin to online judge platforms, where multiple spreadsheet files are created as test cases for each instruction, ensuring the evaluation of robust solutions capable of handling spreadsheets with varying values. Our comprehensive evaluation of various LLMs under both single-round and multi-round inference settings reveals a substantial gap between the state-of-the-art (SOTA) models and human performance, highlighting the benchmark's difficulty.
comment: Neurips 2024 (Spotlight); Homepage: https://spreadsheetbench.github.io/
♻ ☆ Beyond Instruction Following: Evaluating Inferential Rule Following of Large Language Models
Although Large Language Models (LLMs) have demonstrated strong ability, they are further supposed to be controlled and guided by in real-world scenarios to be safe, accurate, and intelligent. This demands the possession of capability of LLMs. However, no prior work has made a clear evaluation of the inferential rule-following capability of LLMs. Previous studies that try to evaluate the inferential rule-following capability of LLMs fail to distinguish the inferential rule-following scenarios from the instruction-following scenarios. Therefore, this paper first clarifies the concept of inferential rule-following and proposes a comprehensive benchmark, RuleBench, to evaluate a diversified range of inferential rule-following abilities. Our experimental results on a variety of LLMs show that they are still limited in following rules. Our analysis based on the evaluation results provides insights into the improvements for LLMs toward a better inferential rule-following intelligent agent. We further propose Inferential Rule-Following Tuning (IRFT). The experimental results show that through IRFT, LLMs can learn abstract rule-following abilities from purely synthetic data and then generalize to RuleBench. The data and code can be found at: https://anonymous.4open.science/r/llm-rule-following-B3E3/
♻ ☆ Temporally Consistent Factuality Probing for Large Language Models
The prolific use of Large Language Models (LLMs) as an alternate knowledge base requires them to be factually consistent, necessitating both correctness and consistency traits for paraphrased queries. Recently, significant attempts have been made to benchmark datasets and metrics to evaluate LLMs for these traits. However, structural simplicity (subject-relation-object) and contemporary association in their query formulation limit the broader definition of factuality and consistency. In this study, we introduce TeCFaP, a novel Temporally Consistent Factuality Probe task to expand the consistent factuality probe in the temporal dimension. To this end, we propose TEMP-COFAC, a high-quality dataset of prefix-style English query paraphrases. Subsequently, we extend the definitions of existing metrics to represent consistent factuality across temporal dimension. We experiment with a diverse set of LLMs and find most of them performing poorly on TeCFaP. Next, we propose a novel solution CoTSeLF (Consistent-Time-Sensitive Learning Framework) combining multi-task instruction tuning (MT-IT) with consistent-time-sensitive reinforcement learning (CTSRL) to improve temporally consistent factuality in LLMs. Our experiments demonstrate the efficacy of CoTSeLF over several baselines.
♻ ☆ Investigating Chain-of-thought with ChatGPT for Stance Detection on Social Media
Stance detection predicts attitudes towards targets in texts and has gained attention with the rise of social media. Traditional approaches include conventional machine learning, early deep neural networks, and pre-trained fine-tuning models. However, with the evolution of very large pre-trained language models (VLPLMs) like ChatGPT (GPT-3.5), traditional methods face deployment challenges. The parameter-free Chain-of-Thought (CoT) approach, not requiring backpropagation training, has emerged as a promising alternative. This paper examines CoT's effectiveness in stance detection tasks, demonstrating its superior accuracy and discussing associated challenges.
comment: arXiv admin note: text overlap with arXiv:2212.14548
♻ ☆ MedCare: Advancing Medical LLMs through Decoupling Clinical Alignment and Knowledge Aggregation EMNLP2024
Large language models (LLMs) have shown substantial progress in natural language understanding and generation, proving valuable especially in the medical field. Despite advancements, challenges persist due to the complexity and diversity inherent in medical tasks, which can be categorized as knowledge-intensive tasks and alignment-required tasks. Previous approaches either ignore the latter task or focus on a minority of tasks and hence lose generalization. To address these drawbacks, we propose a progressive fine-tuning pipeline. This pipeline employs a Knowledge Aggregator and a Noise aggregator to encode diverse knowledge in the first stage and filter out detrimental information. In the second stage, we drop the Noise Aggregator to avoid the interference of suboptimal representation and leverage an additional alignment module optimized towards an orthogonal direction to the knowledge space to mitigate knowledge forgetting. Based on this two-stage paradigm, we proposed a Medical LLM through decoupling Clinical Alignment and Knowledge Aggregation (MedCare), which is designed to achieve state-of-the-art (SOTA) performance on over 20 medical tasks, as well as SOTA results on specific medical alignment tasks. Various model sizes of MedCare (1.8B, 7B, 14B) all demonstrate significant improvements over existing models with similar model sizes.
comment: EMNLP2024 Findings
♻ ☆ Belief Revision: The Adaptability of Large Language Models Reasoning
The capability to reason from text is crucial for real-world NLP applications. Real-world scenarios often involve incomplete or evolving data. In response, individuals update their beliefs and understandings accordingly. However, most existing evaluations assume that language models (LMs) operate with consistent information. We introduce Belief-R, a new dataset designed to test LMs' belief revision ability when presented with new evidence. Inspired by how humans suppress prior inferences, this task assesses LMs within the newly proposed delta reasoning ($\Delta R$) framework. Belief-R features sequences of premises designed to simulate scenarios where additional information could necessitate prior conclusions drawn by LMs. We evaluate $\sim$30 LMs across diverse prompting strategies and found that LMs generally struggle to appropriately revise their beliefs in response to new information. Further, models adept at updating often underperformed in scenarios without necessary updates, highlighting a critical trade-off. These insights underscore the importance of improving LMs' adaptiveness to changing information, a step toward more reliable AI systems.
♻ ☆ Enabling Natural Zero-Shot Prompting on Encoder Models via Statement-Tuning
While Large Language Models (LLMs) exhibit remarkable capabilities in zero-shot and few-shot scenarios, they often require computationally prohibitive sizes. Conversely, smaller Masked Language Models (MLMs) like BERT and RoBERTa achieve state-of-the-art results through fine-tuning but struggle with extending to few-shot and zero-shot settings due to their architectural constraints. Hence, we propose Statement-Tuning, a technique that models discriminative tasks as a set of finite statements and trains an encoder model to discriminate between the potential statements to determine the label. We do Statement-Tuning on multiple tasks to enable cross-task generalization. Experimental results demonstrate that Statement-Tuning achieves competitive performance compared to state-of-the-art LLMs with significantly fewer parameters. Moreover, the study investigates the impact of several design choices on few-shot and zero-shot generalization, revealing that Statement-Tuning can achieve strong performance with modest training data and benefits from task and statement diversity for unseen task generalizability.
♻ ☆ PrivacyLens: Evaluating Privacy Norm Awareness of Language Models in Action NeurIPS 2024
As language models (LMs) are widely utilized in personalized communication scenarios (e.g., sending emails, writing social media posts) and endowed with a certain level of agency, ensuring they act in accordance with the contextual privacy norms becomes increasingly critical. However, quantifying the privacy norm awareness of LMs and the emerging privacy risk in LM-mediated communication is challenging due to (1) the contextual and long-tailed nature of privacy-sensitive cases, and (2) the lack of evaluation approaches that capture realistic application scenarios. To address these challenges, we propose PrivacyLens, a novel framework designed to extend privacy-sensitive seeds into expressive vignettes and further into agent trajectories, enabling multi-level evaluation of privacy leakage in LM agents' actions. We instantiate PrivacyLens with a collection of privacy norms grounded in privacy literature and crowdsourced seeds. Using this dataset, we reveal a discrepancy between LM performance in answering probing questions and their actual behavior when executing user instructions in an agent setup. State-of-the-art LMs, like GPT-4 and Llama-3-70B, leak sensitive information in 25.68% and 38.69% of cases, even when prompted with privacy-enhancing instructions. We also demonstrate the dynamic nature of PrivacyLens by extending each seed into multiple trajectories to red-team LM privacy leakage risk. Dataset and code are available at https://github.com/SALT-NLP/PrivacyLens.
comment: NeurIPS 2024 Datasets and Benchmarks Track
♻ ☆ Prompt Compression for Large Language Models: A Survey
Leveraging large language models (LLMs) for complex natural language tasks typically requires long-form prompts to convey detailed requirements and information, which results in increased memory usage and inference costs. To mitigate these challenges, multiple efficient methods have been proposed, with prompt compression gaining significant research interest. This survey provides an overview of prompt compression techniques, categorized into hard prompt methods and soft prompt methods. First, the technical approaches of these methods are compared, followed by an exploration of various ways to understand their mechanisms, including the perspectives of attention optimization, Parameter-Efficient Fine-Tuning (PEFT), modality integration, and new synthetic language. We also examine the downstream adaptations of various prompt compression techniques. Finally, the limitations of current prompt compression methods are analyzed, and several future directions are outlined, such as optimizing the compression encoder, combining hard and soft prompts methods, and leveraging insights from multimodality.
♻ ☆ Mixture of In-Context Experts Enhance LLMs' Long Context Awareness
Many studies have revealed that large language models (LLMs) exhibit uneven awareness of different contextual positions. Their limited context awareness can lead to overlooking critical information and subsequent task failures. While several approaches have been proposed to enhance LLMs' context awareness, achieving both effectiveness and efficiency remains challenging. In this paper, for LLMs utilizing RoPE as position embeddings, we introduce a novel method called "Mixture of In-Context Experts" (MoICE) to address this challenge. MoICE comprises two key components: a router integrated into each attention head within LLMs and a lightweight router-only training optimization strategy: (1) MoICE views each RoPE angle as an `in-context' expert, demonstrated to be capable of directing the attention of a head to specific contextual positions. Consequently, each attention head flexibly processes tokens using multiple RoPE angles dynamically selected by the router to attend to the needed positions. This approach mitigates the risk of overlooking essential contextual information. (2) The router-only training strategy entails freezing LLM parameters and exclusively updating routers for only a few steps. When applied to open-source LLMs including Llama and Mistral, MoICE surpasses prior methods across multiple tasks on long context understanding and generation, all while maintaining commendable inference efficiency.
comment: Accepted by Neurips2024
♻ ☆ Cross-modality Information Check for Detecting Jailbreaking in Multimodal Large Language Models EMNLP 2024
Multimodal Large Language Models (MLLMs) extend the capacity of LLMs to understand multimodal information comprehensively, achieving remarkable performance in many vision-centric tasks. Despite that, recent studies have shown that these models are susceptible to jailbreak attacks, which refer to an exploitative technique where malicious users can break the safety alignment of the target model and generate misleading and harmful answers. This potential threat is caused by both the inherent vulnerabilities of LLM and the larger attack scope introduced by vision input. To enhance the security of MLLMs against jailbreak attacks, researchers have developed various defense techniques. However, these methods either require modifications to the model's internal structure or demand significant computational resources during the inference phase. Multimodal information is a double-edged sword. While it increases the risk of attacks, it also provides additional data that can enhance safeguards. Inspired by this, we propose Cross-modality Information DEtectoR (CIDER), a plug-and-play jailbreaking detector designed to identify maliciously perturbed image inputs, utilizing the cross-modal similarity between harmful queries and adversarial images. CIDER is independent of the target MLLMs and requires less computation cost. Extensive experimental results demonstrate the effectiveness and efficiency of CIDER, as well as its transferability to both white-box and black-box MLLMs.
comment: 12 pages, 9 figures, EMNLP 2024 Findings
♻ ☆ Tables as Texts or Images: Evaluating the Table Reasoning Ability of LLMs and MLLMs ACL 2024
In this paper, we investigate the effectiveness of various LLMs in interpreting tabular data through different prompting strategies and data formats. Our analyses extend across six benchmarks for table-related tasks such as question-answering and fact-checking. We introduce for the first time the assessment of LLMs' performance on image-based table representations. Specifically, we compare five text-based and three image-based table representations, demonstrating the role of representation and prompting on LLM performance. Our study provides insights into the effective use of LLMs on table-related tasks.
comment: Accepted to ACL 2024 Findings; Naihao and Zhenjie contributed equally to the project; Data available at: https://github.com/dnaihao/Tables-as-Texts-or-Images
♻ ☆ CREAM: Consistency Regularized Self-Rewarding Language Models
Recent self-rewarding large language models (LLM) have successfully applied LLM-as-a-Judge to iteratively improve the alignment performance without the need of human annotations for preference data. These methods commonly utilize the same LLM to act as both the policy model (which generates responses) and the reward model (which scores and ranks those responses). The ranked responses are then used as preference pairs to train the LLM via direct alignment technologies (e.g. DPO). However, it is noteworthy that throughout this process, there is no guarantee of accuracy in the rewarding and ranking, which is critical for ensuring accurate rewards and high-quality preference data. Empirical results from relatively small LLMs (e.g., 7B parameters) also indicate that improvements from self-rewarding may diminish after several iterations in certain situations, which we hypothesize is due to accumulated bias in the reward system. This bias can lead to unreliable preference data for training the LLM. To address this issue, we first formulate and analyze the generalized iterative preference fine-tuning framework for self-rewarding language model. We then introduce the regularization to this generalized framework to mitigate the overconfident preference labeling in the self-rewarding process. Based on this theoretical insight, we propose a Consistency Regularized sElf-rewarding lAnguage Model (CREAM) that leverages the rewarding consistency across different iterations to regularize the self-rewarding training, helping the model to learn from more reliable preference data. With this explicit regularization, our empirical results demonstrate the superiority of CREAM in improving both reward consistency and alignment performance. The code is publicly available at https://github.com/Raibows/CREAM.
♻ ☆ What Matters in Transformers? Not All Attention is Needed
While scaling Transformer-based large language models (LLMs) has demonstrated promising performance across various tasks, it also introduces redundant architectures, posing efficiency challenges for real-world deployment. Despite some recognition of redundancy in LLMs, the variability of redundancy across different architectures in transformers, such as MLP and Attention layers, is under-explored. In this work, we investigate redundancy across different modules within Transformers, including Blocks, MLP, and Attention layers, using a similarity-based metric. Surprisingly, despite the critical role of attention layers in distinguishing transformers from other architectures, we found that a large portion of these layers exhibit excessively high similarity and can be pruned without degrading performance. For instance, Llama-2-70B achieved a 48.4\% speedup with only a 2.4\% performance drop by pruning half of the attention layers. Furthermore, by tracing model checkpoints throughout the training process, we observed that attention layer redundancy is inherent and consistent across training stages. Additionally, we further propose a method that jointly drops Attention and MLP layers, allowing us to more aggressively drop additional layers. For instance, when dropping 31 layers (Attention + MLP), Llama-2-13B still retains 90\% of the performance on the MMLU task. Our work provides valuable insights for future network architecture design. The code is released at: \url{https://github.com/Shwai-He/LLM-Drop}.
comment: 15 pages, 13 figures, 6 tables
♻ ☆ LLM-based Translation Inference with Iterative Bilingual Understanding
The remarkable understanding and generation capabilities of large language models (LLMs) have greatly improved translation performance. However, incorrect understanding of the sentence to be translated can degrade translation quality. To address this issue, we proposed a novel Iterative Bilingual Understanding Translation (IBUT) method based on the cross-lingual capabilities of LLMs and the dual characteristics of translation tasks. The cross-lingual capability of LLMs enables the generation of contextual understanding for both the source and target languages separately. Furthermore, the dual characteristics allow IBUT to generate effective cross-lingual feedback, iteratively refining contextual understanding, thereby reducing errors and improving translation performance. Experimental results showed that the proposed IBUT outperforms several strong comparison methods, especially being generalized to multiple domains (e.g., news, commonsense, and cultural translation benchmarks).
comment: Work in progress
♻ ☆ ActiveRAG: Autonomously Knowledge Assimilation and Accommodation through Retrieval-Augmented Agents
Retrieval-Augmented Generation (RAG) enables Large Language Models (LLMs) to leverage external knowledge, enhancing their performance on knowledge-intensive tasks. However, existing RAG models often treat LLMs as passive recipients of information, which can lead to interference from noisy retrieved content. In this paper, we introduce ActiveRAG, a multi-agent framework that mimics human learning behavior to help LLMs actively engage with and learn from retrieved evidence. ActiveRAG designs a knowledge assimilation agent to form the knowledge understanding by associating external knowledge with the parametric memory of LLMs. Then our model employs the thought accommodation agent to calibrate the internal thought of LLMs for response refinement. Our experiments show that ActiveRAG achieves a 10\% improvement over vanilla RAG on various question-answering benchmarks. Further analysis reveals that ActiveRAG mitigates the impact of noisy retrievals, alleviates conflicts between external knowledge and parametric memory and improves the self-consistency of LLMs in answering the question. All data and codes are available at https://github.com/OpenMatch/ActiveRAG.
♻ ☆ RAGEval: Scenario Specific RAG Evaluation Dataset Generation Framework
Retrieval-Augmented Generation (RAG) is a powerful approach that enables large language models (LLMs) to incorporate external knowledge. However, evaluating the effectiveness of RAG systems in specialized scenarios remains challenging due to the high costs of data construction and the lack of suitable evaluation metrics. This paper introduces RAGEval, a framework designed to assess RAG systems across diverse scenarios by generating high-quality documents, questions, answers, and references through a schema-based pipeline. With a focus on factual accuracy, we propose three novel metrics Completeness, Hallucination, and Irrelevance to rigorously evaluate LLM-generated responses. Experimental results show that RAGEval outperforms zero-shot and one-shot methods in terms of clarity, safety, conformity, and richness of generated samples. Furthermore, the use of LLMs for scoring the proposed metrics demonstrates a high level of consistency with human evaluations. RAGEval establishes a new paradigm for evaluating RAG systems in real-world applications.
comment: https://github.com/OpenBMB/RAGEval
♻ ☆ Benchmarking LLMs for Translating Classical Chinese Poetry:Evaluating Adequacy, Fluency, and Elegance
Large language models (LLMs) have shown remarkable performance in translation tasks. However, the increasing demand for high-quality translations that are not only adequate but also fluent and elegant. To evaluate the extent to which current LLMs can meet these demands, we introduce a suitable benchmark (PoetMT) for translating classical Chinese poetry into English. This task requires not only adequacy in translating culturally and historically significant content but also a strict adherence to linguistic fluency and poetic elegance. To overcome the limitations of traditional evaluation metrics, we propose an automatic evaluation metric based on GPT-4, which better evaluates translation quality in terms of adequacy, fluency, and elegance. Our evaluation study reveals that existing large language models fall short in this task. To evaluate these issues, we propose RAT, a Retrieval-Augmented machine Translation method that enhances the translation process by incorporating knowledge related to classical poetry. Our dataset and code will be made available.
comment: Work in progress
♻ ☆ A Theory for Token-Level Harmonization in Retrieval-Augmented Generation
Retrieval-augmented generation (RAG) utilizes retrieved texts to enhance large language models (LLMs). Studies show that while RAG provides valuable external information (benefit), it may also mislead LLMs (detriment) with noisy or incorrect retrieved texts. Although many existing methods attempt to preserve benefit and avoid detriment, they lack a theoretical explanation for RAG. The benefit and detriment in the next token prediction of RAG remain a black box that cannot be quantified or compared in an explainable manner, so existing methods are data-driven, need additional utility evaluators or post-hoc. This paper takes the first step towards providing a theory to explain and trade off the benefit and detriment in RAG. First, we model RAG as the fusion between distribution of LLMs knowledge and distribution of retrieved texts. Then, we formalize the trade-off between the value of external knowledge (benefit) and its potential risk of misleading LLMs (detriment) in next token prediction of RAG by distribution difference in this fusion. Finally, we prove that the actual effect of RAG on the token, which is the comparison between benefit and detriment, can be predicted without any training or accessing the utility of retrieval. Based on our theory, we propose a practical novel method, Tok-RAG, which achieves collaborative generation between the pure LLM and RAG at token level to preserve benefit and avoid detriment. Experiments in real-world tasks using LLMs such as OPT, LLaMA-2, and Mistral show the effectiveness of our method and support our theoretical findings.
comment: 25 pages
♻ ☆ Instruction Matters: A Simple yet Effective Task Selection for Optimized Instruction Tuning of Specific Tasks EMNLP 2024
Instruction tuning has been proven effective in enhancing zero-shot generalization across various tasks and in improving the performance of specific tasks. For task-specific improvements, strategically selecting and training on related tasks that provide meaningful supervision is crucial, as this approach enhances efficiency and prevents performance degradation from learning irrelevant tasks. In this light, we introduce a simple yet effective task selection method that leverages instruction information alone to identify relevant tasks, optimizing instruction tuning for specific tasks. Our method is significantly more efficient than traditional approaches, which require complex measurements of pairwise transferability between tasks or the creation of data samples for the target task. Additionally, by aligning the model with the unique instructional template style of the meta-dataset, we enhance its ability to granularly discern relevant tasks, leading to improved overall performance. Experimental results demonstrate that training on a small set of tasks, chosen solely based on the instructions, results in substantial improvements in performance on benchmarks such as P3, Big-Bench, NIV2, and Big-Bench Hard. Significantly, these improvements surpass those achieved by prior task selection methods, highlighting the superiority of our approach.
comment: EMNLP 2024 (Camera-ready), by Janghoon Han and Changho Lee, with equal contribution
♻ ☆ RULE: Reliable Multimodal RAG for Factuality in Medical Vision Language Models EMNLP 2024
The recent emergence of Medical Large Vision Language Models (Med-LVLMs) has enhanced medical diagnosis. However, current Med-LVLMs frequently encounter factual issues, often generating responses that do not align with established medical facts. Retrieval-Augmented Generation (RAG), which utilizes external knowledge, can improve the factual accuracy of these models but introduces two major challenges. First, limited retrieved contexts might not cover all necessary information, while excessive retrieval can introduce irrelevant and inaccurate references, interfering with the model's generation. Second, in cases where the model originally responds correctly, applying RAG can lead to an over-reliance on retrieved contexts, resulting in incorrect answers. To address these issues, we propose RULE, which consists of two components. First, we introduce a provably effective strategy for controlling factuality risk through the calibrated selection of the number of retrieved contexts. Second, based on samples where over-reliance on retrieved contexts led to errors, we curate a preference dataset to fine-tune the model, balancing its dependence on inherent knowledge and retrieved contexts for generation. We demonstrate the effectiveness of RULE on medical VQA and report generation tasks across three datasets, achieving an average improvement of 47.4% in factual accuracy. We publicly release our benchmark and code in https://github.com/richard-peng-xia/RULE.
comment: EMNLP 2024 main
♻ ☆ TAIA: Large Language Models are Out-of-Distribution Data Learners NeurIPS
Fine-tuning on task-specific question-answer pairs is a predominant method for enhancing the performance of instruction-tuned large language models (LLMs) on downstream tasks. However, in certain specialized domains, such as healthcare or harmless content generation, it is nearly impossible to obtain a large volume of high-quality data that matches the downstream distribution. To improve the performance of LLMs in data-scarce domains with domain-mismatched data, we re-evaluated the Transformer architecture and discovered that not all parameter updates during fine-tuning contribute positively to downstream performance. Our analysis reveals that within the self-attention and feed-forward networks, only the fine-tuned attention parameters are particularly beneficial when the training set's distribution does not fully align with the test set. Based on this insight, we propose an effective inference-time intervention method: Training All parameters but Inferring with only Attention (\trainallInfAttn). We empirically validate \trainallInfAttn using two general instruction-tuning datasets and evaluate it on seven downstream tasks involving math, reasoning, and knowledge understanding across LLMs of different parameter sizes and fine-tuning techniques. Our comprehensive experiments demonstrate that \trainallInfAttn achieves superior improvements compared to both the fully fine-tuned model and the base model in most scenarios, with significant performance gains. The high tolerance of \trainallInfAttn to data mismatches makes it resistant to jailbreaking tuning and enhances specialized tasks using general data. Code is available in \url{https://github.com/pixas/TAIA_LLM}.
comment: 29 pages. Accepted as a 2024 NeurIPS paper
♻ ☆ From Redundancy to Relevance: Information Flow in LVLMs Across Reasoning Tasks
Large Vision Language Models (LVLMs) achieve great performance on visual-language reasoning tasks, however, the black-box nature of LVLMs hinders in-depth research on the reasoning mechanism. As all images need to be converted into image tokens to fit the input format of large language models (LLMs) along with natural language prompts, sequential visual representation is essential to the performance of LVLMs, and the information flow analysis approach can be an effective tool for determining interactions between these representations. In this paper, we propose integrating attention analysis with LLaVA-CAM, concretely, attention scores highlight relevant regions during forward propagation, while LLaVA-CAM captures gradient changes through backward propagation, revealing key image features. By exploring the information flow from the perspective of visual representation contribution, we observe that it tends to converge in shallow layers but diversify in deeper layers. To validate our analysis, we conduct comprehensive experiments with truncation strategies across various LVLMs for visual question answering and image captioning tasks, and experimental results not only verify our hypothesis but also reveal a consistent pattern of information flow convergence in the corresponding layers, and the information flow cliff layer will be different due to different contexts. The paper's source code can be accessed from \url{https://github.com/zhangbaijin/From-Redundancy-to-Relevance}
♻ ☆ Avoiding Copyright Infringement via Large Language Model Unlearning
Pre-trained Large Language Models (LLMs) have demonstrated remarkable capabilities but also pose risks by learning and generating copyrighted material, leading to significant legal and ethical concerns. In real-world scenarios, model owners need to continuously address copyright infringement as new requests for content removal emerge at different time points. This leads to the need for sequential unlearning, where copyrighted content is removed sequentially as new requests arise. Despite its practical relevance, sequential unlearning in the context of copyright infringement has not been rigorously explored in existing literature. To address this gap, we propose Stable Sequential Unlearning (SSU), a novel framework designed to unlearn copyrighted content from LLMs over multiple time steps. Our approach works by identifying and removing specific weight updates in the model's parameters that correspond to copyrighted content. We improve unlearning efficacy by introducing random labeling loss and ensuring the model retains its general-purpose knowledge by adjusting targeted parameters. Experimental results show that SSU achieves an effective trade-off between unlearning efficacy and general-purpose language abilities, outperforming existing baselines.
♻ ☆ REAL: Response Embedding-based Alignment for LLMs
Aligning large language models (LLMs) to human preferences is a crucial step in building helpful and safe AI tools, which usually involve training on supervised datasets. Popular algorithms such as Direct Preference Optimization rely on pairs of AI-generated responses ranked according to human feedback. The labeling process is the most labor-intensive and costly part of the alignment pipeline, and improving its efficiency would have a meaningful impact on AI development. We propose a strategy for sampling a high-quality training dataset that focuses on acquiring the most informative response pairs for labeling out of a set of AI-generated responses. Experimental results on synthetic HH-RLHF benchmarks indicate that choosing dissimilar response pairs enhances the direct alignment of LLMs while reducing inherited labeling errors. We also applied our method to the real-world dataset SHP2, selecting optimal pairs from multiple responses. The model aligned on dissimilar response pairs obtained the best win rate on the dialogue task. Our findings suggest that focusing on less similar pairs can improve the efficiency of LLM alignment, saving up to 65% of annotators' work.
♻ ☆ Experimental Contexts Can Facilitate Robust Semantic Property Inference in Language Models, but Inconsistently EMNLP 2024
Recent zero-shot evaluations have highlighted important limitations in the abilities of language models (LMs) to perform meaning extraction. However, it is now well known that LMs can demonstrate radical improvements in the presence of experimental contexts such as in-context examples and instructions. How well does this translate to previously studied meaning-sensitive tasks? We present a case-study on the extent to which experimental contexts can improve LMs' robustness in performing property inheritance -- predicting semantic properties of novel concepts, a task that they have been previously shown to fail on. Upon carefully controlling the nature of the in-context examples and the instructions, our work reveals that they can indeed lead to non-trivial property inheritance behavior in LMs. However, this ability is inconsistent: with a minimal reformulation of the task, some LMs were found to pick up on shallow, non-semantic heuristics from their inputs, suggesting that the computational principles of semantic property inference are yet to be mastered by LMs.
comment: EMNLP 2024 (main) camera-ready
♻ ☆ Comparative Analysis of Extrinsic Factors for NER in French
Named entity recognition (NER) is a crucial task that aims to identify structured information, which is often replete with complex, technical terms and a high degree of variability. Accurate and reliable NER can facilitate the extraction and analysis of important information. However, NER for other than English is challenging due to limited data availability, as the high expertise, time, and expenses are required to annotate its data. In this paper, by using the limited data, we explore various factors including model structure, corpus annotation scheme and data augmentation techniques to improve the performance of a NER model for French. Our experiments demonstrate that these approaches can significantly improve the model's F1 score from original CRF score of 62.41 to 79.39. Our findings suggest that considering different extrinsic factors and combining these techniques is a promising approach for improving NER performance where the size of data is limited.
♻ ☆ Improving Bilingual Lexicon Induction with Cross-Encoder Reranking EMNLP 2022
Bilingual lexicon induction (BLI) with limited bilingual supervision is a crucial yet challenging task in multilingual NLP. Current state-of-the-art BLI methods rely on the induction of cross-lingual word embeddings (CLWEs) to capture cross-lingual word similarities; such CLWEs are obtained 1) via traditional static models (e.g., VecMap), or 2) by extracting type-level CLWEs from multilingual pretrained language models (mPLMs), or 3) through combining the former two options. In this work, we propose a novel semi-supervised post-hoc reranking method termed BLICEr (BLI with Cross-Encoder Reranking), applicable to any precalculated CLWE space, which improves their BLI capability. The key idea is to 'extract' cross-lingual lexical knowledge from mPLMs, and then combine it with the original CLWEs. This crucial step is done via 1) creating a word similarity dataset, comprising positive word pairs (i.e., true translations) and hard negative pairs induced from the original CLWE space, and then 2) fine-tuning an mPLM (e.g., mBERT or XLM-R) in a cross-encoder manner to predict the similarity scores. At inference, we 3) combine the similarity score from the original CLWE space with the score from the BLI-tuned cross-encoder. BLICEr establishes new state-of-the-art results on two standard BLI benchmarks spanning a wide spectrum of diverse languages: it substantially outperforms a series of strong baselines across the board. We also validate the robustness of BLICEr with different CLWEs.
comment: Findings of EMNLP 2022
♻ ☆ Scaling laws for post-training quantized large language models
Generalization abilities of well-trained large language models (LLMs) are known to scale predictably as a function of model size. In contrast to the existence of practical scaling laws governing pre-training, the quality of LLMs after post-training compression remains highly unpredictable, often requiring case-by-case validation in practice. In this work, we attempted to close this gap for post-training weight quantization of LLMs by conducting a systematic empirical study on multiple LLM families quantized to numerous low-precision tensor data types using popular weight quantization techniques. We identified key scaling factors pertaining to characteristics of the local loss landscape, based on which the performance of quantized LLMs can be reasonably well predicted by a statistical model.
♻ ☆ MACAROON: Training Vision-Language Models To Be Your Engaged Partners
Large vision-language models (LVLMs), while proficient in following instructions and responding to diverse questions, invariably generate detailed responses even when questions are ambiguous or unanswerable, leading to hallucinations and bias issues. Thus, it is essential for LVLMs to proactively engage with humans to ask for clarifications or additional information for better responses. In this study, we aim to shift LVLMs from passive answer providers to proactive engaged partners. We begin by establishing a three-tiered hierarchy for questions of invalid, ambiguous, and personalizable nature to measure the proactive engagement capabilities of LVLMs. Utilizing this hierarchy, we create PIE, (ProactIve Engagement Evaluation) through GPT-4o and human annotators, consisting of 853 questions across six distinct, fine-grained question types that are verified by human annotators and accompanied with well-defined metrics. Our evaluations on \benchmark indicate poor performance of existing LVLMs, with the best-performing open-weights model only achieving an Aggregate Align Rate (AAR) of 0.28. In response, we introduce MACAROON, self-iMaginAtion for ContrAstive pReference OptimizatiON, which instructs LVLMs to autonomously generate contrastive response pairs for unlabeled questions given the task description and human-crafted criteria. Then, the self-imagined data is formatted for conditional reinforcement learning. Experimental results show MACAROON effectively improves LVLMs' capabilities to be proactively engaged (0.84 AAR) while maintaining comparable performance on general tasks.
comment: The code will be made public at https://github.com/ShujinWu-0814/MACAROON
♻ ☆ Improving Word Translation via Two-Stage Contrastive Learning ACL 2022
Word translation or bilingual lexicon induction (BLI) is a key cross-lingual task, aiming to bridge the lexical gap between different languages. In this work, we propose a robust and effective two-stage contrastive learning framework for the BLI task. At Stage C1, we propose to refine standard cross-lingual linear maps between static word embeddings (WEs) via a contrastive learning objective; we also show how to integrate it into the self-learning procedure for even more refined cross-lingual maps. In Stage C2, we conduct BLI-oriented contrastive fine-tuning of mBERT, unlocking its word translation capability. We also show that static WEs induced from the `C2-tuned' mBERT complement static WEs from Stage C1. Comprehensive experiments on standard BLI datasets for diverse languages and different experimental setups demonstrate substantial gains achieved by our framework. While the BLI method from Stage C1 already yields substantial gains over all state-of-the-art BLI methods in our comparison, even stronger improvements are met with the full two-stage framework: e.g., we report gains for 112/112 BLI setups, spanning 28 language pairs.
comment: ACL 2022 Main
♻ ☆ UnSeenTimeQA: Time-Sensitive Question-Answering Beyond LLMs' Memorization
This paper introduces UnSeenTimeQA, a novel data contamination free time-sensitive question-answering (TSQA) benchmark. It differs from existing TSQA benchmarks by avoiding web-searchable queries grounded in the real-world. We present a series of time-sensitive event scenarios based on synthetically generated facts. It requires large language models (LLMs) to engage in genuine temporal reasoning without depending on the factual knowledge acquired during the pre-training phase. We designed three types of time-sensitive questions to test LLMs' temporal reasoning abilities over sequential and parallel event occurrences. Our evaluation of five LLMs shows that their performance on synthetic fact-based TSQA is inferior as compared to their performance on real-world fact-based TSQA. Further analysis of LLM-generated reasoning chains indicates difficulty in capturing long-range event dependencies and the effect of interlinked events in synthetic scenarios.
♻ ☆ Into the Unknown Unknowns: Engaged Human Learning through Participation in Language Model Agent Conversations EMNLP 2024
While language model (LM)-powered chatbots and generative search engines excel at answering concrete queries, discovering information in the terrain of unknown unknowns remains challenging for users. To emulate the common educational scenario where children/students learn by listening to and participating in conversations of their parents/teachers, we create Collaborative STORM (Co-STORM). Unlike QA systems that require users to ask all the questions, Co-STORM lets users observe and occasionally steer the discourse among several LM agents. The agents ask questions on the user's behalf, allowing the user to discover unknown unknowns serendipitously. To facilitate user interaction, Co-STORM assists users in tracking the discourse by organizing the uncovered information into a dynamic mind map, ultimately generating a comprehensive report as takeaways. For automatic evaluation, we construct the WildSeek dataset by collecting real information-seeking records with user goals. Co-STORM outperforms baseline methods on both discourse trace and report quality. In a further human evaluation, 70% of participants prefer Co-STORM over a search engine, and 78% favor it over a RAG chatbot.
comment: EMNLP 2024 Main
♻ ☆ GUS-Net: Social Bias Classification in Text with Generalizations, Unfairness, and Stereotypes
The detection of bias in natural language processing (NLP) is a critical challenge, particularly with the increasing use of large language models (LLMs) in various domains. This paper introduces GUS-Net, an innovative approach to bias detection that focuses on three key types of biases: (G)eneralizations, (U)nfairness, and (S)tereotypes. GUS-Net leverages generative AI and automated agents to create a comprehensive synthetic dataset, enabling robust multi-label token classification. Our methodology enhances traditional bias detection methods by incorporating the contextual encodings of pre-trained models, resulting in improved accuracy and depth in identifying biased entities. Through extensive experiments, we demonstrate that GUS-Net outperforms state-of-the-art techniques, achieving superior performance in terms of accuracy, F1-score, and Hamming Loss. The findings highlight GUS-Net's effectiveness in capturing a wide range of biases across diverse contexts, making it a valuable tool for social bias detection in text. This study contributes to the ongoing efforts in NLP to address implicit bias, providing a pathway for future research and applications in various fields. The Jupyter notebooks used to create the dataset and model are available at: https://github.com/Ethical-Spectacle/fair-ly/tree/main/resources. Warning: This paper contains examples of harmful language, and reader discretion is recommended.
♻ ☆ Modeling offensive content detection for TikTok
The advent of social media transformed interpersonal communication and information consumption processes. This digital landscape accommodates user intentions, also resulting in an increase of offensive language and harmful behavior. Concurrently, social media platforms collect vast datasets comprising user-generated content and behavioral information. These datasets are instrumental for platforms deploying machine learning and data-driven strategies, facilitating customer insights and countermeasures against social manipulation mechanisms like disinformation and offensive content. Nevertheless, the availability of such datasets, along with the application of various machine learning techniques, to researchers and practitioners, for specific social media platforms regarding particular events, is limited. In particular for TikTok, which offers unique tools for personalized content creation and sharing, the existing body of knowledge would benefit from having diverse comprehensive datasets and associated data analytics solutions on offensive content. While efforts from social media platforms, research, and practitioner communities are seen on this behalf, such content continues to proliferate. This translates to an essential need to make datasets publicly available and build corresponding intelligent solutions. On this behalf, this research undertakes the collection and analysis of TikTok data containing offensive content, building a series of machine learning and deep learning models for offensive content detection. This is done aiming at answering the following research question: "How to develop a series of computational models to detect offensive content on TikTok?". To this end, a Data Science methodological approach is considered, 120.423 TikTok comments are collected, and on a balanced, binary classification approach, F1 score performance results of 0.863 is obtained.
comment: Accepted as a conference paper at DPSH 2024, 8 pages
♻ ☆ WinoQueer: A Community-in-the-Loop Benchmark for Anti-LGBTQ+ Bias in Large Language Models ACL 2023
We present WinoQueer: a benchmark specifically designed to measure whether large language models (LLMs) encode biases that are harmful to the LGBTQ+ community. The benchmark is community-sourced, via application of a novel method that generates a bias benchmark from a community survey. We apply our benchmark to several popular LLMs and find that off-the-shelf models generally do exhibit considerable anti-queer bias. Finally, we show that LLM bias against a marginalized community can be somewhat mitigated by finetuning on data written about or by members of that community, and that social media text written by community members is more effective than news text written about the community by non-members. Our method for community-in-the-loop benchmark development provides a blueprint for future researchers to develop community-driven, harms-grounded LLM benchmarks for other marginalized communities. Note: This version corrects a bug found in evaluation code after publication. General findings have not changed, but tables 5 and 6 and figure 1 have been corrected.
comment: Accepted to ACL 2023 (main conference). This version corrects a bug found in evaluation code after publication. General findings have not changed, but tables 5 and 6 and figure 1 have been corrected
♻ ☆ PrE-Text: Training Language Models on Private Federated Data in the Age of LLMs ICML 2024
On-device training is currently the most common approach for training machine learning (ML) models on private, distributed user data. Despite this, on-device training has several drawbacks: (1) most user devices are too small to train large models on-device, (2) on-device training is communication- and computation-intensive, and (3) on-device training can be difficult to debug and deploy. To address these problems, we propose Private Evolution-Text (PrE-Text), a method for generating differentially private (DP) synthetic textual data. First, we show that across multiple datasets, training small models (models that fit on user devices) with PrE-Text synthetic data outperforms small models trained on-device under practical privacy regimes ($\epsilon=1.29$, $\epsilon=7.58$). We achieve these results while using 9$\times$ fewer rounds, 6$\times$ less client computation per round, and 100$\times$ less communication per round. Second, finetuning large models on PrE-Text's DP synthetic data improves large language model (LLM) performance on private data across the same range of privacy budgets. Altogether, these results suggest that training on DP synthetic data can be a better option than training a model on-device on private distributed data. Code is available at https://github.com/houcharlie/PrE-Text.
comment: ICML 2024 (Oral). Latest revision corrects a discussion on concurrent work arXiv:2403.01749. We described their work as reliant on using closed-sourced models when in reality they also evaluate and use open source models. This has been corrected in this version
♻ ☆ Evaluating Contextualized Representations of (Spanish) Ambiguous Words: A New Lexical Resource and Empirical Analysis NAACL 2025
Lexical ambiguity -- where a single wordform takes on distinct, context-dependent meanings -- serves as a useful tool to compare across different language models' (LMs') ability to form distinct, contextualized representations of the same stimulus. Few studies have systematically compared LMs' contextualized word embeddings for languages beyond English. Here, we evaluate semantic representations of Spanish ambiguous nouns in context in a suite of Spanish-language monolingual and multilingual BERT-based models. We develop a novel dataset of minimal-pair sentences evoking the same or different sense for a target ambiguous noun. In a pre-registered study, we collect contextualized human relatedness judgments for each sentence pair. We find that various BERT-based LMs' contextualized semantic representations capture some variance in human judgments but fall short of the human benchmark. In exploratory work, we find that performance scales with model size. We also identify stereotyped trajectories of target noun disambiguation as a proportion of traversal through a given LM family's architecture, which we partially replicate in English. We contribute (1) a dataset of controlled, Spanish sentence stimuli with human relatedness norms, and (2) to our evolving understanding of the impact that LM specification (architectures, training protocols) exerts on contextualized embeddings.
comment: 17 pages, 12 figures, submitted to conference (NAACL 2025)
♻ ☆ ZeQR: Zero-shot Query Reformulation for Conversational Search SIGIR
As the popularity of voice assistants continues to surge, conversational search has gained increased attention in Information Retrieval. However, data sparsity issues in conversational search significantly hinder the progress of supervised conversational search methods. Consequently, researchers are focusing more on zero-shot conversational search approaches. Nevertheless, existing zero-shot methods face three primary limitations: they are not universally applicable to all retrievers, their effectiveness lacks sufficient explainability, and they struggle to resolve common conversational ambiguities caused by omission. To address these limitations, we introduce a novel Zero-shot Query Reformulation (or Query Rewriting) (ZeQR) framework that reformulates queries based on previous dialogue contexts without requiring supervision from conversational search data. Specifically, our framework utilizes language models designed for machine reading comprehension tasks to explicitly resolve two common ambiguities: coreference and omission, in raw queries. In comparison to existing zero-shot methods, our approach is universally applicable to any retriever without additional adaptation or indexing. It also provides greater explainability and effectively enhances query intent understanding because ambiguities are explicitly and proactively resolved. Through extensive experiments on four TREC conversational datasets, we demonstrate the effectiveness of our method, which consistently outperforms state-of-the-art baselines.
comment: Accepted by the 9th ACM SIGIR International Conference on the Theory of Information Retrieval
♻ ☆ "Vorbeşti Româneşte?" A Recipe to Train Powerful Romanian LLMs with English Instructions
In recent years, Large Language Models (LLMs) have achieved almost human-like performance on various tasks. While some LLMs have been trained on multilingual data, most of the training data is in English; hence, their performance in English greatly exceeds other languages. To our knowledge, we are the first to collect and translate a large collection of texts, instructions, and benchmarks and train, evaluate, and release open-source LLMs tailored for Romanian. We evaluate our methods on four different categories, including academic benchmarks, MT-Bench (manually translated), and a professionally built historical, cultural, and social benchmark adapted to Romanian. We argue for the usefulness and high performance of RoLLMs by obtaining state-of-the-art results across the board. We publicly release all resources (i.e., data, training and evaluation code, models) to support and encourage research on Romanian LLMs while concurrently creating a generalizable recipe, adequate for other low or less-resourced languages.
comment: arXiv admin note: text overlap with arXiv:2405.07703
♻ ☆ Multi-Conditional Ranking with Large Language Models
Utilizing large language models (LLMs) to rank a set of items has become a common approach in recommendation and retrieval systems. Typically, these systems focus on ordering a substantial number of documents in a monotonic order based on a given query. However, real-world scenarios often present a different challenge: ranking a comparatively smaller set of items, but according to a variety of diverse and occasionally conflicting conditions. In this paper, we define and explore the task of multi-conditional ranking by introducing MCRank, a benchmark tailored for assessing multi-conditional ranking across various item types and conditions. Our analysis of LLMs using MCRank indicates a significant decrease in performance as the number and complexity of items and conditions grow. To overcome this limitation, we propose a novel decomposed reasoning method, consisting of EXtracting and Sorting the conditions, and then Iteratively Ranking the items (EXSIR). Our extensive experiments show that this decomposed reasoning method enhances LLMs' performance significantly, achieving up to a 14.4% improvement over existing LLMs. We also provide a detailed analysis of LLMs performance across various condition categories, and examine the effectiveness of decomposition step. Furthermore, we compare our method with existing approaches such as Chain-of-Thought and existing ranking models, demonstrating the superiority of our approach and complexity of MCR task. We released our dataset and code.
♻ ☆ DALK: Dynamic Co-Augmentation of LLMs and KG to answer Alzheimer's Disease Questions with Scientific Literature EMNLP 2024
Recent advancements in large language models (LLMs) have achieved promising performances across various applications. Nonetheless, the ongoing challenge of integrating long-tail knowledge continues to impede the seamless adoption of LLMs in specialized domains. In this work, we introduce DALK, a.k.a. Dynamic Co-Augmentation of LLMs and KG, to address this limitation and demonstrate its ability on studying Alzheimer's Disease (AD), a specialized sub-field in biomedicine and a global health priority. With a synergized framework of LLM and KG mutually enhancing each other, we first leverage LLM to construct an evolving AD-specific knowledge graph (KG) sourced from AD-related scientific literature, and then we utilize a coarse-to-fine sampling method with a novel self-aware knowledge retrieval approach to select appropriate knowledge from the KG to augment LLM inference capabilities. The experimental results, conducted on our constructed AD question answering (ADQA) benchmark, underscore the efficacy of DALK. Additionally, we perform a series of detailed analyses that can offer valuable insights and guidelines for the emerging topic of mutually enhancing KG and LLM. We will release the code and data at https://github.com/David-Li0406/DALK.
comment: Accepted by EMNLP 2024 Findings; revise format problem
♻ ☆ Framing in the Presence of Supporting Data: A Case Study in U.S. Economic News ACL 2024
The mainstream media has much leeway in what it chooses to cover and how it covers it. These choices have real-world consequences on what people know and their subsequent behaviors. However, the lack of objective measures to evaluate editorial choices makes research in this area particularly difficult. In this paper, we argue that there are newsworthy topics where objective measures exist in the form of supporting data and propose a computational framework to analyze editorial choices in this setup. We focus on the economy because the reporting of economic indicators presents us with a relatively easy way to determine both the selection and framing of various publications. Their values provide a ground truth of how the economy is doing relative to how the publications choose to cover it. To do this, we define frame prediction as a set of interdependent tasks. At the article level, we learn to identify the reported stance towards the general state of the economy. Then, for every numerical quantity reported in the article, we learn to identify whether it corresponds to an economic indicator and whether it is being reported in a positive or negative way. To perform our analysis, we track six American publishers and each article that appeared in the top 10 slots of their landing page between 2015 and 2023.
comment: published in ACL 2024; total pages: 19; main body pages: 8; total figures: 19
Computer Vision and Pattern Recognition 172
Fluid: Scaling Autoregressive Text-to-image Generative Models with Continuous Tokens
Scaling up autoregressive models in vision has not proven as beneficial as in large language models. In this work, we investigate this scaling problem in the context of text-to-image generation, focusing on two critical factors: whether models use discrete or continuous tokens, and whether tokens are generated in a random or fixed raster order using BERT- or GPT-like transformer architectures. Our empirical results show that, while all models scale effectively in terms of validation loss, their evaluation performance -- measured by FID, GenEval score, and visual quality -- follows different trends. Models based on continuous tokens achieve significantly better visual quality than those using discrete tokens. Furthermore, the generation order and attention mechanisms significantly affect the GenEval score: random-order models achieve notably better GenEval scores compared to raster-order models. Inspired by these findings, we train Fluid, a random-order autoregressive model on continuous tokens. Fluid 10.5B model achieves a new state-of-the-art zero-shot FID of 6.16 on MS-COCO 30K, and 0.69 overall score on the GenEval benchmark. We hope our findings and results will encourage future efforts to further bridge the scaling gap between vision and language models.
comment: Tech report
☆ UniDrive: Towards Universal Driving Perception Across Camera Configurations
Vision-centric autonomous driving has demonstrated excellent performance with economical sensors. As the fundamental step, 3D perception aims to infer 3D information from 2D images based on 3D-2D projection. This makes driving perception models susceptible to sensor configuration (e.g., camera intrinsics and extrinsics) variations. However, generalizing across camera configurations is important for deploying autonomous driving models on different car models. In this paper, we present UniDrive, a novel framework for vision-centric autonomous driving to achieve universal perception across camera configurations. We deploy a set of unified virtual cameras and propose a ground-aware projection method to effectively transform the original images into these unified virtual views. We further propose a virtual configuration optimization method by minimizing the expected projection error between original cameras and virtual cameras. The proposed virtual camera projection can be applied to existing 3D perception methods as a plug-and-play module to mitigate the challenges posed by camera parameter variability, resulting in more adaptable and reliable driving perception models. To evaluate the effectiveness of our framework, we collect a dataset on Carla by driving the same routes while only modifying the camera configurations. Experimental results demonstrate that our method trained on one specific camera configuration can generalize to varying configurations with minor performance degradation.
comment: Preprint; 14 pages, 5 figures, 2 tables; Code at https://github.com/ywyeli/UniDrive
☆ DepthSplat: Connecting Gaussian Splatting and Depth
Gaussian splatting and single/multi-view depth estimation are typically studied in isolation. In this paper, we present DepthSplat to connect Gaussian splatting and depth estimation and study their interactions. More specifically, we first contribute a robust multi-view depth model by leveraging pre-trained monocular depth features, leading to high-quality feed-forward 3D Gaussian splatting reconstructions. We also show that Gaussian splatting can serve as an unsupervised pre-training objective for learning powerful depth models from large-scale unlabelled datasets. We validate the synergy between Gaussian splatting and depth estimation through extensive ablation and cross-task transfer experiments. Our DepthSplat achieves state-of-the-art performance on ScanNet, RealEstate10K and DL3DV datasets in terms of both depth estimation and novel view synthesis, demonstrating the mutual benefits of connecting both tasks. Our code, models, and video results are available at https://haofeixu.github.io/depthsplat/.
comment: Project page: https://haofeixu.github.io/depthsplat/
☆ PUMA: Empowering Unified MLLM with Multi-granular Visual Generation
Recent advancements in multimodal foundation models have yielded significant progress in vision-language understanding. Initial attempts have also explored the potential of multimodal large language models (MLLMs) for visual content generation. However, existing works have insufficiently addressed the varying granularity demands of different image generation tasks within a unified MLLM paradigm - from the diversity required in text-to-image generation to the precise controllability needed in image manipulation. In this work, we propose PUMA, emPowering Unified MLLM with Multi-grAnular visual generation. PUMA unifies multi-granular visual features as both inputs and outputs of MLLMs, elegantly addressing the different granularity requirements of various image generation tasks within a unified MLLM framework. Following multimodal pretraining and task-specific instruction tuning, PUMA demonstrates proficiency in a wide range of multimodal tasks. This work represents a significant step towards a truly unified MLLM capable of adapting to the granularity demands of various visual tasks. The code and model will be released in https://github.com/rongyaofang/PUMA.
comment: Project page: https://rongyaofang.github.io/puma/
☆ VLM-Grounder: A VLM Agent for Zero-Shot 3D Visual Grounding
3D visual grounding is crucial for robots, requiring integration of natural language and 3D scene understanding. Traditional methods depending on supervised learning with 3D point clouds are limited by scarce datasets. Recently zero-shot methods leveraging LLMs have been proposed to address the data issue. While effective, these methods only use object-centric information, limiting their ability to handle complex queries. In this work, we present VLM-Grounder, a novel framework using vision-language models (VLMs) for zero-shot 3D visual grounding based solely on 2D images. VLM-Grounder dynamically stitches image sequences, employs a grounding and feedback scheme to find the target object, and uses a multi-view ensemble projection to accurately estimate 3D bounding boxes. Experiments on ScanRefer and Nr3D datasets show VLM-Grounder outperforms previous zero-shot methods, achieving 51.6% Acc@0.25 on ScanRefer and 48.0% Acc on Nr3D, without relying on 3D geometry or object priors. Codes are available at https://github.com/OpenRobotLab/VLM-Grounder .
comment: CoRL 2024 Camera Ready. 25 pages. A novel zero-shot 3D visual grounding framework based solely on 2D images
☆ $γ-$MoD: Exploring Mixture-of-Depth Adaptation for Multimodal Large Language Models
Despite the significant progress in multimodal large language models (MLLMs), their high computational cost remains a barrier to real-world deployment. Inspired by the mixture of depths (MoDs) in natural language processing, we aim to address this limitation from the perspective of ``activated tokens''. Our key insight is that if most tokens are redundant for the layer computation, then can be skipped directly via the MoD layer. However, directly converting the dense layers of MLLMs to MoD layers leads to substantial performance degradation. To address this issue, we propose an innovative MoD adaptation strategy for existing MLLMs called $\gamma$-MoD. In $\gamma$-MoD, a novel metric is proposed to guide the deployment of MoDs in the MLLM, namely rank of attention maps (ARank). Through ARank, we can effectively identify which layer is redundant and should be replaced with the MoD layer. Based on ARank, we further propose two novel designs to maximize the computational sparsity of MLLM while maintaining its performance, namely shared vision-language router and masked routing learning. With these designs, more than 90% dense layers of the MLLM can be effectively converted to the MoD ones. To validate our method, we apply it to three popular MLLMs, and conduct extensive experiments on 9 benchmark datasets. Experimental results not only validate the significant efficiency benefit of $\gamma$-MoD to existing MLLMs but also confirm its generalization ability on various MLLMs. For example, with a minor performance drop, i.e., -1.5%, $\gamma$-MoD can reduce the training and inference time of LLaVA-HR by 31.0% and 53.2%, respectively.
☆ Can MLLMs Understand the Deep Implication Behind Chinese Images?
As the capabilities of Multimodal Large Language Models (MLLMs) continue to improve, the need for higher-order capability evaluation of MLLMs is increasing. However, there is a lack of work evaluating MLLM for higher-order perception and understanding of Chinese visual content. To fill the gap, we introduce the **C**hinese **I**mage **I**mplication understanding **Bench**mark, **CII-Bench**, which aims to assess the higher-order perception and understanding capabilities of MLLMs for Chinese images. CII-Bench stands out in several ways compared to existing benchmarks. Firstly, to ensure the authenticity of the Chinese context, images in CII-Bench are sourced from the Chinese Internet and manually reviewed, with corresponding answers also manually crafted. Additionally, CII-Bench incorporates images that represent Chinese traditional culture, such as famous Chinese traditional paintings, which can deeply reflect the model's understanding of Chinese traditional culture. Through extensive experiments on CII-Bench across multiple MLLMs, we have made significant findings. Initially, a substantial gap is observed between the performance of MLLMs and humans on CII-Bench. The highest accuracy of MLLMs attains 64.4%, where as human accuracy averages 78.2%, peaking at an impressive 81.0%. Subsequently, MLLMs perform worse on Chinese traditional culture images, suggesting limitations in their ability to understand high-level semantics and lack a deep knowledge base of Chinese traditional culture. Finally, it is observed that most models exhibit enhanced accuracy when image emotion hints are incorporated into the prompts. We believe that CII-Bench will enable MLLMs to gain a better understanding of Chinese semantics and Chinese-specific images, advancing the journey towards expert artificial general intelligence (AGI). Our project is publicly available at https://cii-bench.github.io/.
comment: 32 pages,18 figures. Project Page: https://cii-bench.github.io/ Code: https://github.com/MING_X/CII-Bench Dataset: https://huggingface.co/datasets/m-a-p/CII-Bench
☆ Retrospective Learning from Interactions
Multi-turn interactions between large language models (LLMs) and users naturally include implicit feedback signals. If an LLM responds in an unexpected way to an instruction, the user is likely to signal it by rephrasing the request, expressing frustration, or pivoting to an alternative task. Such signals are task-independent and occupy a relatively constrained subspace of language, allowing the LLM to identify them even if it fails on the actual task. This creates an avenue for continually learning from interactions without additional annotations. We introduce ReSpect, a method to learn from such signals in past interactions via retrospection. We deploy ReSpect in a new multimodal interaction scenario, where humans instruct an LLM to solve an abstract reasoning task with a combinatorial solution space. Through thousands of interactions with humans, we show how ReSpect gradually improves task completion rate from 31% to 82%, all without any external annotation.
☆ Differentiable Robot Rendering
Vision foundation models trained on massive amounts of visual data have shown unprecedented reasoning and planning skills in open-world settings. A key challenge in applying them to robotic tasks is the modality gap between visual data and action data. We introduce differentiable robot rendering, a method allowing the visual appearance of a robot body to be directly differentiable with respect to its control parameters. Our model integrates a kinematics-aware deformable model and Gaussians Splatting and is compatible with any robot form factors and degrees of freedom. We demonstrate its capability and usage in applications including reconstruction of robot poses from images and controlling robots through vision language models. Quantitative and qualitative results show that our differentiable rendering model provides effective gradients for robotic control directly from pixels, setting the foundation for the future applications of vision foundation models in robotics.
comment: Project Page: https://drrobot.cs.columbia.edu/
☆ Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generation
In this paper, we introduce Janus, an autoregressive framework that unifies multimodal understanding and generation. Prior research often relies on a single visual encoder for both tasks, such as Chameleon. However, due to the differing levels of information granularity required by multimodal understanding and generation, this approach can lead to suboptimal performance, particularly in multimodal understanding. To address this issue, we decouple visual encoding into separate pathways, while still leveraging a single, unified transformer architecture for processing. The decoupling not only alleviates the conflict between the visual encoder's roles in understanding and generation, but also enhances the framework's flexibility. For instance, both the multimodal understanding and generation components can independently select their most suitable encoding methods. Experiments show that Janus surpasses previous unified model and matches or exceeds the performance of task-specific models. The simplicity, high flexibility, and effectiveness of Janus make it a strong candidate for next-generation unified multimodal models.
comment: Technical Report
☆ D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement
We introduce D-FINE, a powerful real-time object detector that achieves outstanding localization precision by redefining the bounding box regression task in DETR models. D-FINE comprises two key components: Fine-grained Distribution Refinement (FDR) and Global Optimal Localization Self-Distillation (GO-LSD). FDR transforms the regression process from predicting fixed coordinates to iteratively refining probability distributions, providing a fine-grained intermediate representation that significantly enhances localization accuracy. GO-LSD is a bidirectional optimization strategy that transfers localization knowledge from refined distributions to shallower layers through self-distillation, while also simplifying the residual prediction tasks for deeper layers. Additionally, D-FINE incorporates lightweight optimizations in computationally intensive modules and operations, achieving a better balance between speed and accuracy. Specifically, D-FINE-L / X achieves 54.0% / 55.8% AP on the COCO dataset at 124 / 78 FPS on an NVIDIA T4 GPU. When pretrained on Objects365, D-FINE-L / X attains 57.1% / 59.3% AP, surpassing all existing real-time detectors. Furthermore, our method significantly enhances the performance of a wide range of DETR models by up to 5.3% AP with negligible extra parameters and training costs. Our code and pretrained models: https://github.com/Peterande/D-FINE.
☆ VidPanos: Generative Panoramic Videos from Casual Panning Videos SIGGRAPH
Panoramic image stitching provides a unified, wide-angle view of a scene that extends beyond the camera's field of view. Stitching frames of a panning video into a panoramic photograph is a well-understood problem for stationary scenes, but when objects are moving, a still panorama cannot capture the scene. We present a method for synthesizing a panoramic video from a casually-captured panning video, as if the original video were captured with a wide-angle camera. We pose panorama synthesis as a space-time outpainting problem, where we aim to create a full panoramic video of the same length as the input video. Consistent completion of the space-time volume requires a powerful, realistic prior over video content and motion, for which we adapt generative video models. Existing generative models do not, however, immediately extend to panorama completion, as we show. We instead apply video generation as a component of our panorama synthesis system, and demonstrate how to exploit the strengths of the models while minimizing their limitations. Our system can create video panoramas for a range of in-the-wild scenes including people, vehicles, and flowing water, as well as stationary background features.
comment: Project page at https://vidpanos.github.io/. To appear at SIGGRAPH Asia 2024 (conference track)
☆ DreamVideo-2: Zero-Shot Subject-Driven Video Customization with Precise Motion Control
Recent advances in customized video generation have enabled users to create videos tailored to both specific subjects and motion trajectories. However, existing methods often require complicated test-time fine-tuning and struggle with balancing subject learning and motion control, limiting their real-world applications. In this paper, we present DreamVideo-2, a zero-shot video customization framework capable of generating videos with a specific subject and motion trajectory, guided by a single image and a bounding box sequence, respectively, and without the need for test-time fine-tuning. Specifically, we introduce reference attention, which leverages the model's inherent capabilities for subject learning, and devise a mask-guided motion module to achieve precise motion control by fully utilizing the robust motion signal of box masks derived from bounding boxes. While these two components achieve their intended functions, we empirically observe that motion control tends to dominate over subject learning. To address this, we propose two key designs: 1) the masked reference attention, which integrates a blended latent mask modeling scheme into reference attention to enhance subject representations at the desired positions, and 2) a reweighted diffusion loss, which differentiates the contributions of regions inside and outside the bounding boxes to ensure a balance between subject and motion control. Extensive experimental results on a newly curated dataset demonstrate that DreamVideo-2 outperforms state-of-the-art methods in both subject customization and motion control. The dataset, code, and models will be made publicly available.
comment: Project page: https://dreamvideo2.github.io/
☆ Unearthing Skill-Level Insights for Understanding Trade-Offs of Foundation Models
With models getting stronger, evaluations have grown more complex, testing multiple skills in one benchmark and even in the same instance at once. However, skill-wise performance is obscured when inspecting aggregate accuracy, under-utilizing the rich signal modern benchmarks contain. We propose an automatic approach to recover the underlying skills relevant for any evaluation instance, by way of inspecting model-generated rationales. After validating the relevance of rationale-parsed skills and inferring skills for $46$k instances over $12$ benchmarks, we observe many skills to be common across benchmarks, resulting in the curation of hundreds of skill-slices (i.e. sets of instances testing a common skill). Inspecting accuracy over these slices yields novel insights on model trade-offs: e.g., compared to GPT-4o and Claude 3.5 Sonnet, on average, Gemini 1.5 Pro is $18\%$ more accurate in "computing molar mass", but $19\%$ less accurate in "applying constitutional law", despite the overall accuracies of the three models differing by a mere $0.4\%$. Furthermore, we demonstrate the practical utility of our approach by showing that insights derived from skill slice analysis can generalize to held-out instances: when routing each instance to the model strongest on the relevant skills, we see a $3\%$ accuracy improvement over our $12$ dataset corpus. Our skill-slices and framework open a new avenue in model evaluation, leveraging skill-specific analyses to unlock a more granular and actionable understanding of model capabilities.
comment: Code at: github.com/microsoft/skill-slice-insights
☆ Harnessing Webpage UIs for Text-Rich Visual Understanding
Text-rich visual understanding-the ability to process environments where dense textual content is integrated with visuals-is crucial for multimodal large language models (MLLMs) to interact effectively with structured environments. To enhance this capability, we propose synthesizing general multimodal instructions from webpage UIs using text-based large language models (LLMs). Despite lacking direct visual input, text-based LLMs are able to process structured text representations from webpage accessibility trees. These instructions are then paired with UI screenshots to train multimodal models. We introduce MultiUI, a dataset containing 7.3 million samples from 1 million websites, covering diverse multimodal tasks and UI layouts. Models trained on MultiUI not only excel in web UI tasks-achieving up to a 48\% improvement on VisualWebBench and a 19.1\% boost in action accuracy on a web agent dataset Mind2Web-but also generalize surprisingly well to non-web UI tasks and even to non-UI domains, such as document understanding, OCR, and chart interpretation. These results highlight the broad applicability of web UI data for advancing text-rich visual understanding across various scenarios.
☆ Deep Generative Models Unveil Patterns in Medical Images Through Vision-Language Conditioning NeurIPS2024
Deep generative models have significantly advanced medical imaging analysis by enhancing dataset size and quality. Beyond mere data augmentation, our research in this paper highlights an additional, significant capacity of deep generative models: their ability to reveal and demonstrate patterns in medical images. We employ a generative structure with hybrid conditions, combining clinical data and segmentation masks to guide the image synthesis process. Furthermore, we innovatively transformed the tabular clinical data into textual descriptions. This approach simplifies the handling of missing values and also enables us to leverage large pre-trained vision-language models that investigate the relations between independent clinical entries and comprehend general terms, such as gender and smoking status. Our approach differs from and presents a more challenging task than traditional medical report-guided synthesis due to the less visual correlation of our clinical information with the images. To overcome this, we introduce a text-visual embedding mechanism that strengthens the conditions, ensuring the network effectively utilizes the provided information. Our pipeline is generalizable to both GAN-based and diffusion models. Experiments on chest CT, particularly focusing on the smoking status, demonstrated a consistent intensity shift in the lungs which is in agreement with clinical observations, indicating the effectiveness of our method in capturing and visualizing the impact of specific attributes on medical image patterns. Our methods offer a new avenue for the early detection and precise visualization of complex clinical conditions with deep generative models. All codes are https://github.com/junzhin/DGM-VLC.
comment: Accepted by AIM-FM Workshop of NeurIPS2024
☆ Multi-style conversion for semantic segmentation of lesions in fundus images by adversarial attacks
The diagnosis of diabetic retinopathy, which relies on fundus images, faces challenges in achieving transparency and interpretability when using a global classification approach. However, segmentation-based databases are significantly more expensive to acquire and combining them is often problematic. This paper introduces a novel method, termed adversarial style conversion, to address the lack of standardization in annotation styles across diverse databases. By training a single architecture on combined databases, the model spontaneously modifies its segmentation style depending on the input, demonstrating the ability to convert among different labeling styles. The proposed methodology adds a linear probe to detect dataset origin based on encoder features and employs adversarial attacks to condition the model's segmentation style. Results indicate significant qualitative and quantitative through dataset combination, offering avenues for improved model generalization, uncertainty estimation and continuous interpolation between annotation styles. Our approach enables training a segmentation model with diverse databases while controlling and leveraging annotation styles for improved retinopathy diagnosis.
comment: preprint
☆ ConsisSR: Delving Deep into Consistency in Diffusion-based Image Super-Resolution
Real-world image super-resolution (Real-ISR) aims at restoring high-quality (HQ) images from low-quality (LQ) inputs corrupted by unknown and complex degradations. In particular, pretrained text-to-image (T2I) diffusion models provide strong generative priors to reconstruct credible and intricate details. However, T2I generation focuses on semantic consistency while Real-ISR emphasizes pixel-level reconstruction, which hinders existing methods from fully exploiting diffusion priors. To address this challenge, we introduce ConsisSR to handle both semantic and pixel-level consistency. Specifically, compared to coarse-grained text prompts, we exploit the more powerful CLIP image embedding and effectively leverage both modalities through our Hybrid Prompt Adapter (HPA) for semantic guidance. Secondly, we introduce Time-aware Latent Augmentation (TALA) to mitigate the inherent gap between T2I generation and Real-ISR consistency requirements. By randomly mixing LQ and HQ latent inputs, our model not only handle timestep-specific diffusion noise but also refine the accumulated latent representations. Last but not least, our GAN-Embedding strategy employs the pretrained Real-ESRGAN model to refine the diffusion start point. This accelerates the inference process to 10 steps while preserving sampling quality, in a training-free manner. Our method demonstrates state-of-the-art performance among both full-scale and accelerated models. The code will be made publicly available.
☆ MotionBank: A Large-scale Video Motion Benchmark with Disentangled Rule-based Annotations
In this paper, we tackle the problem of how to build and benchmark a large motion model (LMM). The ultimate goal of LMM is to serve as a foundation model for versatile motion-related tasks, e.g., human motion generation, with interpretability and generalizability. Though advanced, recent LMM-related works are still limited by small-scale motion data and costly text descriptions. Besides, previous motion benchmarks primarily focus on pure body movements, neglecting the ubiquitous motions in context, i.e., humans interacting with humans, objects, and scenes. To address these limitations, we consolidate large-scale video action datasets as knowledge banks to build MotionBank, which comprises 13 video action datasets, 1.24M motion sequences, and 132.9M frames of natural and diverse human motions. Different from laboratory-captured motions, in-the-wild human-centric videos contain abundant motions in context. To facilitate better motion text alignment, we also meticulously devise a motion caption generation algorithm to automatically produce rule-based, unbiased, and disentangled text descriptions via the kinematic characteristics for each motion. Extensive experiments show that our MotionBank is beneficial for general motion-related tasks of human motion generation, motion in-context generation, and motion understanding. Video motions together with the rule-based text annotations could serve as an efficient alternative for larger LMMs. Our dataset, codes, and benchmark will be publicly available at https://github.com/liangxuy/MotionBank.
☆ Emphasizing Semantic Consistency of Salient Posture for Speech-Driven Gesture Generation
Speech-driven gesture generation aims at synthesizing a gesture sequence synchronized with the input speech signal. Previous methods leverage neural networks to directly map a compact audio representation to the gesture sequence, ignoring the semantic association of different modalities and failing to deal with salient gestures. In this paper, we propose a novel speech-driven gesture generation method by emphasizing the semantic consistency of salient posture. Specifically, we first learn a joint manifold space for the individual representation of audio and body pose to exploit the inherent semantic association between two modalities, and propose to enforce semantic consistency via a consistency loss. Furthermore, we emphasize the semantic consistency of salient postures by introducing a weakly-supervised detector to identify salient postures, and reweighting the consistency loss to focus more on learning the correspondence between salient postures and the high-level semantics of speech content. In addition, we propose to extract audio features dedicated to facial expression and body gesture separately, and design separate branches for face and body gesture synthesis. Extensive experimental results demonstrate the superiority of our method over the state-of-the-art approaches.
☆ Representing Model Weights with Language using Tree Experts
The increasing availability of public models begs the question: can we train neural networks that use other networks as input? This paper learns to represent models within a joint space that embeds both model weights and language. However, machine learning on model weights is challenging as model weights often exhibit significant variation unrelated to the models' semantic properties (nuisance variation). We identify a key property of real-world models: most public models belong to a small set of Model Trees, where all models within a tree are fine-tuned from a common ancestor (e.g., a foundation model). Importantly, we find that within each tree there is less nuisance variation between models. For example, while classifying models according to their training dataset generally requires complex architectures, in our case, even a linear classifier trained on a single layer is often effective. While effective, linear layers are computationally expensive as model weights are very high dimensional. To address this, we introduce Probing Experts (ProbeX), a theoretically motivated, lightweight probing method. Notably, ProbeX is the first probing method designed to learn from the weights of just a single model layer. We also construct and release a dataset that simulates the structure of public model repositories. Our results show that ProbeX can effectively map the weights of large models into a shared weight-language embedding space. Furthermore, we demonstrate the impressive generalization of our method, achieving zero-shot model classification and retrieval.
☆ Eyelid Fold Consistency in Facial Modeling
Eyelid shape is integral to identity and likeness in human facial modeling. Human eyelids are diverse in appearance with varied skin fold and epicanthal fold morphology between individuals. Existing parametric face models express eyelid shape variation to an extent, but do not preserve sufficient likeness across a diverse range of individuals. We propose a new definition of eyelid fold consistency and implement geometric processing techniques to model diverse eyelid shapes in a unified topology. Using this method we reprocess data used to train a parametric face model and demonstrate significant improvements in face-related machine learning tasks.
☆ Improving Multi-modal Large Language Model through Boosting Vision Capabilities
We focus on improving the visual understanding capability for boosting the vision-language models. We propose \textbf{Arcana}, a multiModal language model, which introduces two crucial techniques. First, we present Multimodal LoRA (MM-LoRA), a module designed to enhance the decoder. Unlike traditional language-driven decoders, MM-LoRA consists of two parallel LoRAs -- one for vision and one for language -- each with its own parameters. This disentangled parameters design allows for more specialized learning in each modality and better integration of multimodal information. Second, we introduce the Query Ladder adapter (QLadder) to improve the visual encoder. QLadder employs a learnable ``\textit{ladder}'' structure to deeply aggregates the intermediate representations from the frozen pretrained visual encoder (e.g., CLIP image encoder). This enables the model to learn new and informative visual features, as well as remaining the powerful capabilities of the pretrained visual encoder. These techniques collectively enhance Arcana's visual perception power, enabling it to leverage improved visual information for more accurate and contextually relevant outputs across various multimodal scenarios. Extensive experiments and ablation studies demonstrate the effectiveness and generalization capability of our Arcana. The code and re-annotated data are available at \url{https://arcana-project-page.github.io}.
☆ DAWN: Dynamic Frame Avatar with Non-autoregressive Diffusion Framework for Talking Head Video Generation
Talking head generation intends to produce vivid and realistic talking head videos from a single portrait and speech audio clip. Although significant progress has been made in diffusion-based talking head generation, almost all methods rely on autoregressive strategies, which suffer from limited context utilization beyond the current generation step, error accumulation, and slower generation speed. To address these challenges, we present DAWN (Dynamic frame Avatar With Non-autoregressive diffusion), a framework that enables all-at-once generation of dynamic-length video sequences. Specifically, it consists of two main components: (1) audio-driven holistic facial dynamics generation in the latent motion space, and (2) audio-driven head pose and blink generation. Extensive experiments demonstrate that our method generates authentic and vivid videos with precise lip motions, and natural pose/blink movements. Additionally, with a high generation speed, DAWN possesses strong extrapolation capabilities, ensuring the stable production of high-quality long videos. These results highlight the considerable promise and potential impact of DAWN in the field of talking head video generation. Furthermore, we hope that DAWN sparks further exploration of non-autoregressive approaches in diffusion models. Our code will be publicly at https://github.com/Hanbo-Cheng/DAWN-pytorch.
☆ Movie Gen: A Cast of Media Foundation Models
We present Movie Gen, a cast of foundation models that generates high-quality, 1080p HD videos with different aspect ratios and synchronized audio. We also show additional capabilities such as precise instruction-based video editing and generation of personalized videos based on a user's image. Our models set a new state-of-the-art on multiple tasks: text-to-video synthesis, video personalization, video editing, video-to-audio generation, and text-to-audio generation. Our largest video generation model is a 30B parameter transformer trained with a maximum context length of 73K video tokens, corresponding to a generated video of 16 seconds at 16 frames-per-second. We show multiple technical innovations and simplifications on the architecture, latent spaces, training objectives and recipes, data curation, evaluation protocols, parallelization techniques, and inference optimizations that allow us to reap the benefits of scaling pre-training data, model size, and training compute for training large scale media generation models. We hope this paper helps the research community to accelerate progress and innovation in media generation models. All videos from this paper are available at https://go.fb.me/MovieGenResearchVideos.
☆ Exploring the Design Space of Visual Context Representation in Video MLLMs
Video Multimodal Large Language Models (MLLMs) have shown remarkable capability of understanding the video semantics on various downstream tasks. Despite the advancements, there is still a lack of systematic research on visual context representation, which refers to the scheme to select frames from a video and further select the tokens from a frame. In this paper, we explore the design space for visual context representation, and aim to improve the performance of video MLLMs by finding more effective representation schemes. Firstly, we formulate the task of visual context representation as a constrained optimization problem, and model the language modeling loss as a function of the number of frames and the number of embeddings (or tokens) per frame, given the maximum visual context window size. Then, we explore the scaling effects in frame selection and token selection respectively, and fit the corresponding function curve by conducting extensive empirical experiments. We examine the effectiveness of typical selection strategies and present empirical findings to determine the two factors. Furthermore, we study the joint effect of frame selection and token selection, and derive the optimal formula for determining the two factors. We demonstrate that the derived optimal settings show alignment with the best-performed results of empirical experiments. Our code and model are available at: https://github.com/RUCAIBox/Opt-Visor.
comment: Long Video MLLM; work in progress
☆ Label-free prediction of fluorescence markers in bovine satellite cells using deep learning
Assessing the quality of bovine satellite cells (BSCs) is essential for the cultivated meat industry, which aims to address global food sustainability challenges. This study aims to develop a label-free method for predicting fluorescence markers in isolated BSCs using deep learning. We employed a U-Net-based CNN model to predict multiple fluorescence signals from a single bright-field microscopy image of cell culture. Two key biomarkers, DAPI and Pax7, were used to determine the abundance and quality of BSCs. The image pre-processing pipeline included fluorescence denoising to improve prediction performance and consistency. A total of 48 biological replicates were used, with statistical performance metrics such as Pearson correlation coefficient and SSIM employed for model evaluation. The model exhibited better performance with DAPI predictions due to uniform staining. Pax7 predictions were more variable, reflecting biological heterogeneity. Enhanced visualization techniques, including color mapping and image overlay, improved the interpretability of the predictions by providing better contextual and perceptual information. The findings highlight the importance of data pre-processing and demonstrate the potential of deep learning to advance non-invasive, label-free assessment techniques in the cultivated meat industry, paving the way for reliable and actionable AI-driven evaluations.
comment: 11 pages, 4 figures
☆ Pose-Based Sign Language Appearance Transfer
We introduce a method for transferring the signer's appearance in sign language skeletal poses while preserving the sign content. Using estimated poses, we transfer the appearance of one signer to another, maintaining natural movements and transitions. This approach improves pose-based rendering and sign stitching while obfuscating identity. Our experiments show that while the method reduces signer identification accuracy, it slightly harms sign recognition performance, highlighting a tradeoff between privacy and utility. Our code is available at \url{https://github.com/sign-language-processing/pose-anonymization}.
☆ Diffusion Curriculum: Synthetic-to-Real Generative Curriculum Learning via Image-Guided Diffusion
Low-quality or scarce data has posed significant challenges for training deep neural networks in practice. While classical data augmentation cannot contribute very different new data, diffusion models opens up a new door to build self-evolving AI by generating high-quality and diverse synthetic data through text-guided prompts. However, text-only guidance cannot control synthetic images' proximity to the original images, resulting in out-of-distribution data detrimental to the model performance. To overcome the limitation, we study image guidance to achieve a spectrum of interpolations between synthetic and real images. With stronger image guidance, the generated images are similar to the training data but hard to learn. While with weaker image guidance, the synthetic images will be easier for model but contribute to a larger distribution gap with the original data. The generated full spectrum of data enables us to build a novel "Diffusion Curriculum (DisCL)". DisCL adjusts the image guidance level of image synthesis for each training stage: It identifies and focuses on hard samples for the model and assesses the most effective guidance level of synthetic images to improve hard data learning. We apply DisCL to two challenging tasks: long-tail (LT) classification and learning from low-quality data. It focuses on lower-guidance images of high-quality to learn prototypical features as a warm-up of learning higher-guidance images that might be weak on diversity or quality. Extensive experiments showcase a gain of 2.7% and 2.1% in OOD and ID macro-accuracy when applying DisCL to iWildCam dataset. On ImageNet-LT, DisCL improves the base model's tail-class accuracy from 4.4% to 23.64% and leads to a 4.02% improvement in all-class accuracy.
☆ VL-GLUE: A Suite of Fundamental yet Challenging Visuo-Linguistic Reasoning Tasks
Deriving inference from heterogeneous inputs (such as images, text, and audio) is an important skill for humans to perform day-to-day tasks. A similar ability is desirable for the development of advanced Artificial Intelligence (AI) systems. While state-of-the-art models are rapidly closing the gap with human-level performance on diverse computer vision and NLP tasks separately, they struggle to solve tasks that require joint reasoning over visual and textual modalities. Inspired by GLUE (Wang et. al., 2018)- a multitask benchmark for natural language understanding, we propose VL-GLUE in this paper. VL-GLUE consists of over 100k samples spanned across seven different tasks, which at their core require visuo-linguistic reasoning. Moreover, our benchmark comprises of diverse image types (from synthetically rendered figures, and day-to-day scenes to charts and complex diagrams) and includes a broad variety of domain-specific text (from cooking, politics, and sports to high-school curricula), demonstrating the need for multi-modal understanding in the real-world. We show that this benchmark is quite challenging for existing large-scale vision-language models and encourage development of systems that possess robust visuo-linguistic reasoning capabilities.
comment: 18 pages, 7 figures
☆ DiRecNetV2: A Transformer-Enhanced Network for Aerial Disaster Recognition
The integration of Unmanned Aerial Vehicles (UAVs) with artificial intelligence (AI) models for aerial imagery processing in disaster assessment, necessitates models that demonstrate exceptional accuracy, computational efficiency, and real-time processing capabilities. Traditionally Convolutional Neural Networks (CNNs), demonstrate efficiency in local feature extraction but are limited by their potential for global context interpretation. On the other hand, Vision Transformers (ViTs) show promise for improved global context interpretation through the use of attention mechanisms, although they still remain underinvestigated in UAV-based disaster response applications. Bridging this research gap, we introduce DiRecNetV2, an improved hybrid model that utilizes convolutional and transformer layers. It merges the inductive biases of CNNs for robust feature extraction with the global context understanding of Transformers, maintaining a low computational load ideal for UAV applications. Additionally, we introduce a new, compact multi-label dataset of disasters, to set an initial benchmark for future research, exploring how models trained on single-label data perform in a multi-label test set. The study assesses lightweight CNNs and ViTs on the AIDERSv2 dataset, based on the frames per second (FPS) for efficiency and the weighted F1 scores for classification performance. DiRecNetV2 not only achieves a weighted F1 score of 0.964 on a single-label test set but also demonstrates adaptability, with a score of 0.614 on a complex multi-label test set, while functioning at 176.13 FPS on the Nvidia Orin Jetson device.
comment: 23 pages
☆ ActionCOMET: A Zero-shot Approach to Learn Image-specific Commonsense Concepts about Actions
Humans observe various actions being performed by other humans (physically or in videos/images) and can draw a wide range of inferences about it beyond what they can visually perceive. Such inferences include determining the aspects of the world that make action execution possible (e.g. liquid objects can undergo pouring), predicting how the world will change as a result of the action (e.g. potatoes being golden and crispy after frying), high-level goals associated with the action (e.g. beat the eggs to make an omelet) and reasoning about actions that possibly precede or follow the current action (e.g. crack eggs before whisking or draining pasta after boiling). Similar reasoning ability is highly desirable in autonomous systems that would assist us in performing everyday tasks. To that end, we propose a multi-modal task to learn aforementioned concepts about actions being performed in images. We develop a dataset consisting of 8.5k images and 59.3k inferences about actions grounded in those images, collected from an annotated cooking-video dataset. We propose ActionCOMET, a zero-shot framework to discern knowledge present in language models specific to the provided visual input. We present baseline results of ActionCOMET over the collected dataset and compare them with the performance of the best existing VQA approaches.
comment: 15 pages, 3 figures. arXiv admin note: text overlap with arXiv:2004.10796 by other authors
☆ Help Me Identify: Is an LLM+VQA System All We Need to Identify Visual Concepts?
An ability to learn about new objects from a small amount of visual data and produce convincing linguistic justification about the presence/absence of certain concepts (that collectively compose the object) in novel scenarios is an important characteristic of human cognition. This is possible due to abstraction of attributes/properties that an object is composed of e.g. an object `bird' can be identified by the presence of a beak, feathers, legs, wings, etc. Inspired by this aspect of human reasoning, in this work, we present a zero-shot framework for fine-grained visual concept learning by leveraging large language model and Visual Question Answering (VQA) system. Specifically, we prompt GPT-3 to obtain a rich linguistic description of visual objects in the dataset. We convert the obtained concept descriptions into a set of binary questions. We pose these questions along with the query image to a VQA system and aggregate the answers to determine the presence or absence of an object in the test images. Our experiments demonstrate comparable performance with existing zero-shot visual classification methods and few-shot concept learning approaches, without substantial computational overhead, yet being fully explainable from the reasoning perspective.
comment: 14 pages, 7 figures
☆ Enhanced Prompt-leveraged Weakly Supervised Cancer Segmentation based on Segment Anything
This work proposes a novel approach beyond supervised learning for effective pathological image analysis, addressing the challenge of limited robust labeled data. Pathological diagnosis of diseases like cancer has conventionally relied on the evaluation of morphological features by physicians and pathologists. However, recent advancements in compute-aided diagnosis (CAD) systems are gaining significant attention as diagnostic support tools. Although the advancement of deep learning has improved CAD significantly, segmentation models typically require large pixel-level annotated dataset, and such labeling is expensive. Existing studies not based on supervised approaches still struggle with limited generalization, and no practical approach has emerged yet. To address this issue, we present a weakly supervised semantic segmentation (WSSS) model by combining class activation map and Segment Anything Model (SAM)-based pseudo-labeling. For effective pretraining, we adopt the SAM-a foundation model that is pretrained on large datasets and operates in zero-shot configurations using only coarse prompts. The proposed approach transfer enhanced Attention Dropout Layer's knowledge to SAM, thereby generating pseudo-labels. To demonstrate the superiority of the proposed method, experimental studies are conducted on histopathological breast cancer datasets. The proposed method outperformed other WSSS methods across three datasets, demonstrating its efficiency by achieving this with only 12GB of GPU memory during training. Our code is available at : https://github.com/QI-NemoSong/EPLC-SAM
comment: 10 pages, 7 figures
☆ LoLDU: Low-Rank Adaptation via Lower-Diag-Upper Decomposition for Parameter-Efficient Fine-Tuning
The rapid growth of model scale has necessitated substantial computational resources for fine-tuning. Existing approach such as Low-Rank Adaptation (LoRA) has sought to address the problem of handling the large updated parameters in full fine-tuning. However, LoRA utilize random initialization and optimization of low-rank matrices to approximate updated weights, which can result in suboptimal convergence and an accuracy gap compared to full fine-tuning. To address these issues, we propose LoLDU, a Parameter-Efficient Fine-Tuning (PEFT) approach that significantly reduces trainable parameters by 2600 times compared to regular PEFT methods while maintaining comparable performance. LoLDU leverages Lower-Diag-Upper Decomposition (LDU) to initialize low-rank matrices for faster convergence and orthogonality. We focus on optimizing the diagonal matrix for scaling transformations. To the best of our knowledge, LoLDU has the fewest parameters among all PEFT approaches. We conducted extensive experiments across 4 instruction-following datasets, 6 natural language understanding (NLU) datasets, 8 image classification datasets, and image generation datasets with multiple model types (LLaMA2, RoBERTa, ViT, and Stable Diffusion), providing a comprehensive and detailed analysis. Our open-source code can be accessed at \href{https://github.com/SKDDJ/LoLDU}{https://github.com/SKDDJ/LoLDU}.
comment: 13 pages, 7 figures
☆ Spatiotemporal Object Detection for Improved Aerial Vehicle Detection in Traffic Monitoring
This work presents advancements in multi-class vehicle detection using UAV cameras through the development of spatiotemporal object detection models. The study introduces a Spatio-Temporal Vehicle Detection Dataset (STVD) containing 6, 600 annotated sequential frame images captured by UAVs, enabling comprehensive training and evaluation of algorithms for holistic spatiotemporal perception. A YOLO-based object detection algorithm is enhanced to incorporate temporal dynamics, resulting in improved performance over single frame models. The integration of attention mechanisms into spatiotemporal models is shown to further enhance performance. Experimental validation demonstrates significant progress, with the best spatiotemporal model exhibiting a 16.22% improvement over single frame models, while it is demonstrated that attention mechanisms hold the potential for additional performance gains.
comment: 13 pages
☆ Material Fingerprinting: Identifying and Predicting Perceptual Attributes of Material Appearance
The world is abundant with diverse materials, each possessing unique surface appearances that play a crucial role in our daily perception and understanding of their properties. Despite advancements in technology enabling the capture and realistic reproduction of material appearances for visualization and quality control, the interoperability of material property information across various measurement representations and software platforms remains a complex challenge. A key to overcoming this challenge lies in the automatic identification of materials' perceptual features, enabling intuitive differentiation of properties stored in disparate material data representations. We reasoned that for many practical purposes, a compact representation of the perceptual appearance is more useful than an exhaustive physical description.This paper introduces a novel approach to material identification by encoding perceptual features obtained from dynamic visual stimuli. We conducted a psychophysical experiment to select and validate 16 particularly significant perceptual attributes obtained from videos of 347 materials. We then gathered attribute ratings from over twenty participants for each material, creating a 'material fingerprint' that encodes the unique perceptual properties of each material. Finally, we trained a multi-layer perceptron model to predict the relationship between statistical and deep learning image features and their corresponding perceptual properties. We demonstrate the model's performance in material retrieval and filtering according to individual attributes. This model represents a significant step towards simplifying the sharing and understanding of material properties in diverse digital environments regardless of their digital representation, enhancing both the accuracy and efficiency of material identification.
comment: 14 pages, 12 figures, 3 tables
☆ MEGA: Memory-Efficient 4D Gaussian Splatting for Dynamic Scenes
4D Gaussian Splatting (4DGS) has recently emerged as a promising technique for capturing complex dynamic 3D scenes with high fidelity. It utilizes a 4D Gaussian representation and a GPU-friendly rasterizer, enabling rapid rendering speeds. Despite its advantages, 4DGS faces significant challenges, notably the requirement of millions of 4D Gaussians, each with extensive associated attributes, leading to substantial memory and storage cost. This paper introduces a memory-efficient framework for 4DGS. We streamline the color attribute by decomposing it into a per-Gaussian direct color component with only 3 parameters and a shared lightweight alternating current color predictor. This approach eliminates the need for spherical harmonics coefficients, which typically involve up to 144 parameters in classic 4DGS, thereby creating a memory-efficient 4D Gaussian representation. Furthermore, we introduce an entropy-constrained Gaussian deformation technique that uses a deformation field to expand the action range of each Gaussian and integrates an opacity-based entropy loss to limit the number of Gaussians, thus forcing our model to use as few Gaussians as possible to fit a dynamic scene well. With simple half-precision storage and zip compression, our framework achieves a storage reduction by approximately 190$\times$ and 125$\times$ on the Technicolor and Neural 3D Video datasets, respectively, compared to the original 4DGS. Meanwhile, it maintains comparable rendering speeds and scene representation quality, setting a new standard in the field.
☆ H2OVL-Mississippi Vision Language Models Technical Report
Smaller vision-language models (VLMs) are becoming increasingly important for privacy-focused, on-device applications due to their ability to run efficiently on consumer hardware for processing enterprise commercial documents and images. These models require strong language understanding and visual capabilities to enhance human-machine interaction. To address this need, we present H2OVL-Mississippi, a pair of small VLMs trained on 37 million image-text pairs using 240 hours of compute on 8 x H100 GPUs. H2OVL-Mississippi-0.8B is a tiny model with 0.8 billion parameters that specializes in text recognition, achieving state of the art performance on the Text Recognition portion of OCRBench and surpassing much larger models in this area. Additionally, we are releasing H2OVL-Mississippi-2B, a 2 billion parameter model for general use cases, exhibiting highly competitive metrics across various academic benchmarks. Both models build upon our prior work with H2O-Danube language models, extending their capabilities into the visual domain. We release them under the Apache 2.0 license, making VLMs accessible to everyone, democratizing document AI and visual LLMs.
☆ DN-4DGS: Denoised Deformable Network with Temporal-Spatial Aggregation for Dynamic Scene Rendering NeurIPS 2024
Dynamic scenes rendering is an intriguing yet challenging problem. Although current methods based on NeRF have achieved satisfactory performance, they still can not reach real-time levels. Recently, 3D Gaussian Splatting (3DGS) has gar?nered researchers attention due to their outstanding rendering quality and real?time speed. Therefore, a new paradigm has been proposed: defining a canonical 3D gaussians and deforming it to individual frames in deformable fields. How?ever, since the coordinates of canonical 3D gaussians are filled with noise, which can transfer noise into the deformable fields, and there is currently no method that adequately considers the aggregation of 4D information. Therefore, we pro?pose Denoised Deformable Network with Temporal-Spatial Aggregation for Dy?namic Scene Rendering (DN-4DGS). Specifically, a Noise Suppression Strategy is introduced to change the distribution of the coordinates of the canonical 3D gaussians and suppress noise. Additionally, a Decoupled Temporal-Spatial Ag?gregation Module is designed to aggregate information from adjacent points and frames. Extensive experiments on various real-world datasets demonstrate that our method achieves state-of-the-art rendering quality under a real-time level.
comment: Accepted by NeurIPS 2024
☆ Let Me Finish My Sentence: Video Temporal Grounding with Holistic Text Understanding
Video Temporal Grounding (VTG) aims to identify visual frames in a video clip that match text queries. Recent studies in VTG employ cross-attention to correlate visual frames and text queries as individual token sequences. However, these approaches overlook a crucial aspect of the problem: a holistic understanding of the query sentence. A model may capture correlations between individual word tokens and arbitrary visual frames while possibly missing out on the global meaning. To address this, we introduce two primary contributions: (1) a visual frame-level gate mechanism that incorporates holistic textual information, (2) cross-modal alignment loss to learn the fine-grained correlation between query and relevant frames. As a result, we regularize the effect of individual word tokens and suppress irrelevant visual frames. We demonstrate that our method outperforms state-of-the-art approaches in VTG benchmarks, indicating that holistic text understanding guides the model to focus on the semantically important parts within the video.
comment: Accepted by ACMMM 24
☆ Deep-learning recognition and tracking of individual nanotubes in low-contrast microscopy videos
This study addresses the challenge of analyzing the growth kinetics of carbon nanotubes using in-situ homodyne polarization microscopy (HPM) by developing an automated deep learning (DL) approach. A Mask-RCNN architecture, enhanced with a ResNet-50 backbone, was employed to recognize and track individual nanotubes in microscopy videos, significantly improving the efficiency and reproducibility of kinetic data extraction. The method involves a series of video processing steps to enhance contrast and used differential treatment techniques to manage low signal and fast kinetics. The DL model demonstrates consistency with manual measurements and increased throughput, laying the foundation for statistical studies of nanotube growth. The approach can be adapted for other types of in-situ microscopy studies, emphasizing the importance of automation in high-throughput data acquisition for research on individual nano-objects.
comment: 13 pages, 5 Figures, No supporting information included
☆ Pseudo Dataset Generation for Out-of-Domain Multi-Camera View Recommendation
Multi-camera systems are indispensable in movies, TV shows, and other media. Selecting the appropriate camera at every timestamp has a decisive impact on production quality and audience preferences. Learning-based view recommendation frameworks can assist professionals in decision-making. However, they often struggle outside of their training domains. The scarcity of labeled multi-camera view recommendation datasets exacerbates the issue. Based on the insight that many videos are edited from the original multi-camera videos, we propose transforming regular videos into pseudo-labeled multi-camera view recommendation datasets. Promisingly, by training the model on pseudo-labeled datasets stemming from videos in the target domain, we achieve a 68% relative improvement in the model's accuracy in the target domain and bridge the accuracy gap between in-domain and never-before-seen domains.
comment: Accepted to VCIP 2024. Project page: https://eric11220.github.io/publication/VCIP24/
☆ Co-Segmentation without any Pixel-level Supervision with Application to Large-Scale Sketch Classification ACCV 2024
This work proposes a novel method for object co-segmentation, i.e. pixel-level localization of a common object in a set of images, that uses no pixel-level supervision for training. Two pre-trained Vision Transformer (ViT) models are exploited: ImageNet classification-trained ViT, whose features are used to estimate rough object localization through intra-class token relevance, and a self-supervised DINO-ViT for intra-image token relevance. On recent challenging benchmarks, the method achieves state-of-the-art performance among methods trained with the same level of supervision (image labels) while being competitive with methods trained with pixel-level supervision (binary masks). The benefits of the proposed co-segmentation method are further demonstrated in the task of large-scale sketch recognition, that is, the classification of sketches into a wide range of categories. The limited amount of hand-drawn sketch training data is leveraged by exploiting readily available image-level-annotated datasets of natural images containing a large number of classes. To bridge the domain gap, the classifier is trained on a sketch-like proxy domain derived from edges detected on natural images. We show that sketch recognition significantly benefits when the classifier is trained on sketch-like structures extracted from the co-segmented area rather than from the full image. Code: https://github.com/nikosips/CBNC .
comment: ACCV 2024 Main Paper + Supplementary (Appendix)
☆ DriveDreamer4D: World Models Are Effective Data Machines for 4D Driving Scene Representation
Closed-loop simulation is essential for advancing end-to-end autonomous driving systems. Contemporary sensor simulation methods, such as NeRF and 3DGS, rely predominantly on conditions closely aligned with training data distributions, which are largely confined to forward-driving scenarios. Consequently, these methods face limitations when rendering complex maneuvers (e.g., lane change, acceleration, deceleration). Recent advancements in autonomous-driving world models have demonstrated the potential to generate diverse driving videos. However, these approaches remain constrained to 2D video generation, inherently lacking the spatiotemporal coherence required to capture intricacies of dynamic driving environments. In this paper, we introduce \textit{DriveDreamer4D}, which enhances 4D driving scene representation leveraging world model priors. Specifically, we utilize the world model as a data machine to synthesize novel trajectory videos based on real-world driving data. Notably, we explicitly leverage structured conditions to control the spatial-temporal consistency of foreground and background elements, thus the generated data adheres closely to traffic constraints. To our knowledge, \textit{DriveDreamer4D} is the first to utilize video generation models for improving 4D reconstruction in driving scenarios. Experimental results reveal that \textit{DriveDreamer4D} significantly enhances generation quality under novel trajectory views, achieving a relative improvement in FID by 24.5\%, 39.0\%, and 10.5\% compared to PVG, $\text{S}^3$Gaussian, and Deformable-GS. Moreover, \textit{DriveDreamer4D} markedly enhances the spatiotemporal coherence of driving agents, which is verified by a comprehensive user study and the relative increases of 20.3\%, 42.0\%, and 13.7\% in the NTA-IoU metric.
comment: https://drivedreamer4d.github.io
☆ RGB to Hyperspectral: Spectral Reconstruction for Enhanced Surgical Imaging
This study investigates the reconstruction of hyperspectral signatures from RGB data to enhance surgical imaging, utilizing the publicly available HeiPorSPECTRAL dataset from porcine surgery and an in-house neurosurgery dataset. Various architectures based on convolutional neural networks (CNNs) and transformer models are evaluated using comprehensive metrics. Transformer models exhibit superior performance in terms of RMSE, SAM, PSNR and SSIM by effectively integrating spatial information to predict accurate spectral profiles, encompassing both visible and extended spectral ranges. Qualitative assessments demonstrate the capability to predict spectral profiles critical for informed surgical decision-making during procedures. Challenges associated with capturing both the visible and extended hyperspectral ranges are highlighted using the MAE, emphasizing the complexities involved. The findings open up the new research direction of hyperspectral reconstruction for surgical applications and clinical use cases in real-time surgical environments.
comment: 10 pages, 4 figures, 3 tables
☆ CCUP: A Controllable Synthetic Data Generation Pipeline for Pretraining Cloth-Changing Person Re-Identification Models
Cloth-changing person re-identification (CC-ReID), also known as Long-Term Person Re-Identification (LT-ReID) is a critical and challenging research topic in computer vision that has recently garnered significant attention. However, due to the high cost of constructing CC-ReID data, the existing data-driven models are hard to train efficiently on limited data, causing overfitting issue. To address this challenge, we propose a low-cost and efficient pipeline for generating controllable and high-quality synthetic data simulating the surveillance of real scenarios specific to the CC-ReID task. Particularly, we construct a new self-annotated CC-ReID dataset named Cloth-Changing Unreal Person (CCUP), containing 6,000 IDs, 1,179,976 images, 100 cameras, and 26.5 outfits per individual. Based on this large-scale dataset, we introduce an effective and scalable pretrain-finetune framework for enhancing the generalization capabilities of the traditional CC-ReID models. The extensive experiments demonstrate that two typical models namely TransReID and FIRe^2, when integrated into our framework, outperform other state-of-the-art models after pretraining on CCUP and finetuning on the benchmarks such as PRCC, VC-Clothes and NKUP. The CCUP is available at: https://github.com/yjzhao1019/CCUP.
☆ 360U-Former: HDR Illumination Estimation with Panoramic Adapted Vision Transformers ECCV 2024
Recent illumination estimation methods have focused on enhancing the resolution and improving the quality and diversity of the generated textures. However, few have explored tailoring the neural network architecture to the Equirectangular Panorama (ERP) format utilised in image-based lighting. Consequently, high dynamic range images (HDRI) results usually exhibit a seam at the side borders and textures or objects that are warped at the poles. To address this shortcoming we propose a novel architecture, 360U-Former, based on a U-Net style Vision-Transformer which leverages the work of PanoSWIN, an adapted shifted window attention tailored to the ERP format. To the best of our knowledge, this is the first purely Vision-Transformer model used in the field of illumination estimation. We train 360U-Former as a GAN to generate HDRI from a limited field of view low dynamic range image (LDRI). We evaluate our method using current illumination estimation evaluation protocols and datasets, demonstrating that our approach outperforms existing and state-of-the-art methods without the artefacts typically associated with the use of the ERP format.
comment: Accepted at AIM Workshop 2024 at ECCV 2024, 18 pages, 6 figures
☆ Generative Location Modeling for Spatially Aware Object Insertion
Generative models have become a powerful tool for image editing tasks, including object insertion. However, these methods often lack spatial awareness, generating objects with unrealistic locations and scales, or unintentionally altering the scene background. A key challenge lies in maintaining visual coherence, which requires both a geometrically suitable object location and a high-quality image edit. In this paper, we focus on the former, creating a location model dedicated to identifying realistic object locations. Specifically, we train an autoregressive model that generates bounding box coordinates, conditioned on the background image and the desired object class. This formulation allows to effectively handle sparse placement annotations and to incorporate implausible locations into a preference dataset by performing direct preference optimization. Our extensive experiments demonstrate that our generative location model, when paired with an inpainting method, substantially outperforms state-of-the-art instruction-tuned models and location modeling baselines in object insertion tasks, delivering accurate and visually coherent results.
☆ RemoteDet-Mamba: A Hybrid Mamba-CNN Network for Multi-modal Object Detection in Remote Sensing Images
Unmanned aerial vehicle (UAV) remote sensing is widely applied in fields such as emergency response, owing to its advantages of rapid information acquisition and low cost. However, due to the effects of shooting distance and imaging mechanisms, the objects in the images present challenges such as small size, dense distribution, and low inter-class differentiation. To this end, we propose a multimodal remote sensing detection network that employs a quad-directional selective scanning fusion strategy called RemoteDet-Mamba. RemoteDet-Mamba simultaneously facilitates the learning of single-modal local features and the integration of patch-level global features across modalities, enhancing the distinguishability for small objects and utilizing local information to improve discrimination between different classes. Additionally, the use of Mamba's serial processing significantly increases detection speed. Experimental results on the DroneVehicle dataset demonstrate the effectiveness of RemoteDet-Mamba, which achieves superior detection accuracy compared to state-of-the-art methods while maintaining computational efficiency and parameter count.
☆ L3DG: Latent 3D Gaussian Diffusion SIGGRAPH
We propose L3DG, the first approach for generative 3D modeling of 3D Gaussians through a latent 3D Gaussian diffusion formulation. This enables effective generative 3D modeling, scaling to generation of entire room-scale scenes which can be very efficiently rendered. To enable effective synthesis of 3D Gaussians, we propose a latent diffusion formulation, operating in a compressed latent space of 3D Gaussians. This compressed latent space is learned by a vector-quantized variational autoencoder (VQ-VAE), for which we employ a sparse convolutional architecture to efficiently operate on room-scale scenes. This way, the complexity of the costly generation process via diffusion is substantially reduced, allowing higher detail on object-level generation, as well as scalability to large scenes. By leveraging the 3D Gaussian representation, the generated scenes can be rendered from arbitrary viewpoints in real-time. We demonstrate that our approach significantly improves visual quality over prior work on unconditional object-level radiance field synthesis and showcase its applicability to room-scale scene generation.
comment: SIGGRAPH Asia 2024, project page: https://barbararoessle.github.io/l3dg , video: https://youtu.be/UHEEiXCYeLU
☆ Generative Adversarial Synthesis of Radar Point Cloud Scenes
For the validation and verification of automotive radars, datasets of realistic traffic scenarios are required, which, how ever, are laborious to acquire. In this paper, we introduce radar scene synthesis using GANs as an alternative to the real dataset acquisition and simulation-based approaches. We train a PointNet++ based GAN model to generate realistic radar point cloud scenes and use a binary classifier to evaluate the performance of scenes generated using this model against a test set of real scenes. We demonstrate that our GAN model achieves similar performance (~87%) to the real scenes test set.
comment: ICMIM 2024; 7th IEEE MTT Conference
☆ Can Medical Vision-Language Pre-training Succeed with Purely Synthetic Data?
Medical Vision-Language Pre-training (MedVLP) has made significant progress in enabling zero-shot tasks for medical image understanding. However, training MedVLP models typically requires large-scale datasets with paired, high-quality image-text data, which are scarce in the medical domain. Recent advancements in Large Language Models (LLMs) and diffusion models have made it possible to generate large-scale synthetic image-text pairs. This raises the question: *Can MedVLP succeed using purely synthetic data?* To address this, we use off-the-shelf generative models to create synthetic radiology reports and paired Chest X-ray (CXR) images, and propose an automated pipeline to build a diverse, high-quality synthetic dataset, enabling a rigorous study that isolates model and training settings, focusing entirely from the data perspective. Our results show that MedVLP models trained *exclusively on synthetic data* outperform those trained on real data by **3.8%** in averaged AUC on zero-shot classification. Moreover, using a combination of synthetic and real data leads to a further improvement of **9.07%**. Additionally, MedVLP models trained on synthetic or mixed data consistently outperform those trained on real data in zero-shot grounding, as well as in fine-tuned classification and segmentation tasks. Our analysis suggests MedVLP trained on well-designed synthetic data can outperform models trained on real datasets, which may be limited by low-quality samples and long-tailed distributions.
comment: Under Review
☆ GeoCoder: Solving Geometry Problems by Generating Modular Code through Vision-Language Models
Geometry problem-solving demands advanced reasoning abilities to process multimodal inputs and employ mathematical knowledge effectively. Vision-language models (VLMs) have made significant progress in various multimodal tasks. Yet, they still struggle with geometry problems and are significantly limited by their inability to perform mathematical operations not seen during pre-training, such as calculating the cosine of an arbitrary angle, and by difficulties in correctly applying relevant geometry formulas. To overcome these challenges, we present GeoCoder, which leverages modular code-finetuning to generate and execute code using a predefined geometry function library. By executing the code, we achieve accurate and deterministic calculations, contrasting the stochastic nature of autoregressive token prediction, while the function library minimizes errors in formula usage. We also propose a multimodal retrieval-augmented variant of GeoCoder, named RAG-GeoCoder, which incorporates a non-parametric memory module for retrieving functions from the geometry library, thereby reducing reliance on parametric memory. Our modular code-finetuning approach enhances the geometric reasoning capabilities of VLMs, yielding an average improvement of over 16% across various question complexities on the GeomVerse dataset compared to other finetuning methods.
☆ SAda-Net: A Self-Supervised Adaptive Stereo Estimation CNN For Remote Sensing Image Data ICPR2024
Stereo estimation has made many advancements in recent years with the introduction of deep-learning. However the traditional supervised approach to deep-learning requires the creation of accurate and plentiful ground-truth data, which is expensive to create and not available in many situations. This is especially true for remote sensing applications, where there is an excess of available data without proper ground truth. To tackle this problem, we propose a self-supervised CNN with self-improving adaptive abilities. In the first iteration, the created disparity map is inaccurate and noisy. Leveraging the left-right consistency check, we get a sparse but more accurate disparity map which is used as an initial pseudo ground-truth. This pseudo ground-truth is then adapted and updated after every epoch in the training step of the network. We use the sum of inconsistent points in order to track the network convergence. The code for our method is publicly available at: https://github.com/thedodo/SAda-Net}{https://github.com/thedodo/SAda-Net
comment: Will be presented at ICPR2024 in December 2024 in Kolkata, India
☆ SemSim: Revisiting Weak-to-Strong Consistency from a Semantic Similarity Perspective for Semi-supervised Medical Image Segmentation
Semi-supervised learning (SSL) for medical image segmentation is a challenging yet highly practical task, which reduces reliance on large-scale labeled dataset by leveraging unlabeled samples. Among SSL techniques, the weak-to-strong consistency framework, popularized by FixMatch, has emerged as a state-of-the-art method in classification tasks. Notably, such a simple pipeline has also shown competitive performance in medical image segmentation. However, two key limitations still persist, impeding its efficient adaptation: (1) the neglect of contextual dependencies results in inconsistent predictions for similar semantic features, leading to incomplete object segmentation; (2) the lack of exploitation of semantic similarity between labeled and unlabeled data induces considerable class-distribution discrepancy. To address these limitations, we propose a novel semi-supervised framework based on FixMatch, named SemSim, powered by two appealing designs from semantic similarity perspective: (1) rectifying pixel-wise prediction by reasoning about the intra-image pair-wise affinity map, thus integrating contextual dependencies explicitly into the final prediction; (2) bridging labeled and unlabeled data via a feature querying mechanism for compact class representation learning, which fully considers cross-image anatomical similarities. As the reliable semantic similarity extraction depends on robust features, we further introduce an effective spatial-aware fusion module (SFM) to explore distinctive information from multiple scales. Extensive experiments show that SemSim yields consistent improvements over the state-of-the-art methods across three public segmentation benchmarks.
☆ Day-Night Adaptation: An Innovative Source-free Adaptation Framework for Medical Image Segmentation
Distribution shifts widely exist in medical images acquired from different medical centers, hindering the deployment of semantic segmentation models trained on data from one center (source domain) to another (target domain). While unsupervised domain adaptation (UDA) has shown significant promise in mitigating these shifts, it poses privacy risks due to sharing data between centers. To facilitate adaptation while preserving data privacy, source-free domain adaptation (SFDA) and test-time adaptation (TTA) have emerged as effective paradigms, relying solely on target domain data. However, the scenarios currently addressed by SFDA and TTA are limited, making them less suitable for clinical applications. In a more realistic clinical scenario, the pre-trained model is deployed in a medical centre to assist with clinical tasks during the day and rest at night. During the daytime process, TTA can be employed to enhance inference performance. During the nighttime process, after collecting the test data from the day, the model can be fine-tuned utilizing SFDA to further adapt to the target domain. With above insights, we propose a novel adaptation framework called Day-Night Adaptation (DyNA). This framework adapts the model to the target domain through day-night loops without requiring access to source data. Specifically, we implement distinct adaptation strategies for daytime and nighttime to better meet the demands of clinical settings. During the daytime, model parameters are frozen, and a specific low-frequency prompt is trained for each test sample. Additionally, we construct a memory bank for prompt initialization and develop a warm-up mechanism to enhance prompt training. During nighttime, we integrate a global student model into the traditional teacher-student self-training paradigm to fine-tune the model while ensuring training stability...
comment: 10 pages, 4 figures, 6 tables
☆ SiamSeg: Self-Training with Contrastive Learning for Unsupervised Domain Adaptation in Remote Sensing
Semantic segmentation of remote sensing (RS) images is a challenging task with significant potential across various applications. Deep learning, especially supervised learning with large-scale labeled datasets, has greatly advanced this field. However, acquiring high-quality labeled data is expensive and time-consuming. Moreover, variations in ground sampling distance (GSD), imaging equipment, and geographic diversity contribute to domain shifts between datasets, which pose significant challenges to models trained solely on source domain data, leading to poor cross-domain performance. Domain shift is well-known for undermining a model's generalization ability in the target domain. To address this, unsupervised domain adaptation (UDA) has emerged as a promising solution, enabling models to learn from unlabeled target domain data while training on labeled source domain data. Recent advancements, particularly in self-supervised learning via pseudo-label generation, have shown potential in mitigating domain discrepancies. Strategies combining source and target domain images with their true and pseudo labels for self-supervised training have been effective in addressing domain bias. Despite progress in computer vision, the application of pseudo-labeling methods to RS image segmentation remains underexplored.
☆ Object Pose Estimation Using Implicit Representation For Transparent Objects
Object pose estimation is a prominent task in computer vision. The object pose gives the orientation and translation of the object in real-world space, which allows various applications such as manipulation, augmented reality, etc. Various objects exhibit different properties with light, such as reflections, absorption, etc. This makes it challenging to understand the object's structure in RGB and depth channels. Recent research has been moving toward learning-based methods, which provide a more flexible and generalizable approach to object pose estimation utilizing deep learning. One such approach is the render-and-compare method, which renders the object from multiple views and compares it against the given 2D image, which often requires an object representation in the form of a CAD model. We reason that the synthetic texture of the CAD model may not be ideal for rendering and comparing operations. We showed that if the object is represented as an implicit (neural) representation in the form of Neural Radiance Field (NeRF), it exhibits a more realistic rendering of the actual scene and retains the crucial spatial features, which makes the comparison more versatile. We evaluated our NeRF implementation of the render-and-compare method on transparent datasets and found that it surpassed the current state-of-the-art results.
☆ Augmentation Policy Generation for Image Classification Using Large Language Models ISCA
Automated data augmentation methods have significantly improved the performance and generalization capability of deep learning models in image classification. Yet, most state-of-the-art methods are optimized on common benchmark datasets, limiting their applicability to more diverse or domain-specific data, such as medical datasets. In this paper, we propose a strategy that uses large language models to automatically generate efficient augmentation policies, customized to fit the specific characteristics of any dataset and model architecture. The proposed method iteratively interacts with an LLM to obtain and refine the augmentation policies on model performance feedback, creating a dataset-agnostic data augmentation pipeline. The proposed method was evaluated on medical imaging datasets, showing a clear improvement over state-of-the-art methods. The proposed approach offers an adaptive and scalable solution. Although it increases computational cost, it significantly boosts model robustness, automates the process, and minimizes the need for human involvement during model development.
comment: 5 pages, 2 figures, 4 tables, submitted for consideration to the International Workshop on Computational Intelligence for Multimedia Understanding (IWCIM), ISCAS 2025
☆ Similarity-Dissimilarity Loss with Supervised Contrastive Learning for Multi-label Classification
Supervised contrastive learning has been explored in making use of label information for multi-label classification, but determining positive samples in multi-label scenario remains challenging. Previous studies have examined strategies for identifying positive samples, considering label overlap proportion between anchors and samples. However, they ignore various relations between given anchors and samples, as well as how to dynamically adjust the weights in contrastive loss functions based on different relations, leading to great ambiguity. In this paper, we introduce five distinct relations between multi-label samples and propose a Similarity-Dissimilarity Loss with contrastive learning for multi-label classification. Our loss function re-weights the loss by computing the similarity and dissimilarity between positive samples and a given anchor based on the introduced relations. We mainly conduct experiments for multi-label text classification on MIMIC datasets, then further extend the evaluation on MS-COCO. The Experimental results show that our proposed loss effectively improves the performance on all encoders under supervised contrastive learning paradigm, demonstrating its effectiveness and robustness.
☆ Temporal-Enhanced Multimodal Transformer for Referring Multi-Object Tracking and Segmentation
Referring multi-object tracking (RMOT) is an emerging cross-modal task that aims to locate an arbitrary number of target objects and maintain their identities referred by a language expression in a video. This intricate task involves the reasoning of linguistic and visual modalities, along with the temporal association of target objects. However, the seminal work employs only loose feature fusion and overlooks the utilization of long-term information on tracked objects. In this study, we introduce a compact Transformer-based method, termed TenRMOT. We conduct feature fusion at both encoding and decoding stages to fully exploit the advantages of Transformer architecture. Specifically, we incrementally perform cross-modal fusion layer-by-layer during the encoding phase. In the decoding phase, we utilize language-guided queries to probe memory features for accurate prediction of the desired objects. Moreover, we introduce a query update module that explicitly leverages temporal prior information of the tracked objects to enhance the consistency of their trajectories. In addition, we introduce a novel task called Referring Multi-Object Tracking and Segmentation (RMOTS) and construct a new dataset named Ref-KITTI Segmentation. Our dataset consists of 18 videos with 818 expressions, and each expression averages 10.7 masks, which poses a greater challenge compared to the typical single mask in most existing referring video segmentation datasets. TenRMOT demonstrates superior performance on both the referring multi-object tracking and the segmentation tasks.
☆ Unsupervised Skull Segmentation via Contrastive MR-to-CT Modality Translation ACCV 2024
The skull segmentation from CT scans can be seen as an already solved problem. However, in MR this task has a significantly greater complexity due to the presence of soft tissues rather than bones. Capturing the bone structures from MR images of the head, where the main visualization objective is the brain, is very demanding. The attempts that make use of skull stripping seem to not be well suited for this task and fail to work in many cases. On the other hand, supervised approaches require costly and time-consuming skull annotations. To overcome the difficulties we propose a fully unsupervised approach, where we do not perform the segmentation directly on MR images, but we rather perform a synthetic CT data generation via MR-to-CT translation and perform the segmentation there. We address many issues associated with unsupervised skull segmentation including the unpaired nature of MR and CT datasets (contrastive learning), low resolution and poor quality (super-resolution), and generalization capabilities. The research has a significant value for downstream tasks requiring skull segmentation from MR volumes such as craniectomy or surgery planning and can be seen as an important step towards the utilization of synthetic data in medical imaging.
comment: 16 pages, 5 figures, ACCV 2024 - GAISynMeD Workshop
☆ Performance of Gaussian Mixture Model Classifiers on Embedded Feature Spaces
Data embeddings with CLIP and ImageBind provide powerful features for the analysis of multimedia and/or multimodal data. We assess their performance here for classification using a Gaussian Mixture models (GMMs) based layer as an alternative to the standard Softmax layer. GMMs based classifiers have recently been shown to have interesting performances as part of deep learning pipelines trained end-to-end. Our first contribution is to investigate GMM based classification performance taking advantage of the embedded spaces CLIP and ImageBind. Our second contribution is in proposing our own GMM based classifier with a lower parameters count than previously proposed. Our findings are, that in most cases, on these tested embedded spaces, one gaussian component in the GMMs is often enough for capturing each class, and we hypothesize that this may be due to the contrastive loss used for training these embedded spaces that naturally concentrates features together for each class. We also observed that ImageBind often provides better performance than CLIP for classification of image datasets even when these embedded spaces are compressed using PCA.
comment: 8 pages
☆ RescueADI: Adaptive Disaster Interpretation in Remote Sensing Images with Autonomous Agents
Current methods for disaster scene interpretation in remote sensing images (RSIs) mostly focus on isolated tasks such as segmentation, detection, or visual question-answering (VQA). However, current interpretation methods often fail at tasks that require the combination of multiple perception methods and specialized tools. To fill this gap, this paper introduces Adaptive Disaster Interpretation (ADI), a novel task designed to solve requests by planning and executing multiple sequentially correlative interpretation tasks to provide a comprehensive analysis of disaster scenes. To facilitate research and application in this area, we present a new dataset named RescueADI, which contains high-resolution RSIs with annotations for three connected aspects: planning, perception, and recognition. The dataset includes 4,044 RSIs, 16,949 semantic masks, 14,483 object bounding boxes, and 13,424 interpretation requests across nine challenging request types. Moreover, we propose a new disaster interpretation method employing autonomous agents driven by large language models (LLMs) for task planning and execution, proving its efficacy in handling complex disaster interpretations. The proposed agent-based method solves various complex interpretation requests such as counting, area calculation, and path-finding without human intervention, which traditional single-task approaches cannot handle effectively. Experimental results on RescueADI demonstrate the feasibility of the proposed task and show that our method achieves an accuracy 9% higher than existing VQA methods, highlighting its advantages over conventional disaster interpretation approaches. The dataset will be publicly available.
☆ Railway LiDAR semantic segmentation based on intelligent semi-automated data annotation
Automated vehicles rely on an accurate and robust perception of the environment. Similarly to automated cars, highly automated trains require an environmental perception. Although there is a lot of research based on either camera or LiDAR sensors in the automotive domain, very few contributions for this task exist yet for automated trains. Additionally, no public dataset or described approach for a 3D LiDAR semantic segmentation in the railway environment exists yet. Thus, we propose an approach for a point-wise 3D semantic segmentation based on the 2DPass network architecture using scans and images jointly. In addition, we present a semi-automated intelligent data annotation approach, which we use to efficiently and accurately label the required dataset recorded on a railway track in Germany. To improve performance despite a still small number of labeled scans, we apply an active learning approach to intelligently select scans for the training dataset. Our contributions are threefold: We annotate rail data including camera and LiDAR data from the railway environment, transfer label the raw LiDAR point clouds using an image segmentation network, and train a state-of-the-art 3D LiDAR semantic segmentation network efficiently leveraging active learning. The trained network achieves good segmentation results with a mean IoU of 71.48% of 9 classes.
comment: This article has been accepted for publication in the IEEE VTC Fall 2024
☆ Accurate Checkerboard Corner Detection under Defoucs
Camera calibration is a critical process in 3D vision, im pacting applications in autonomous driving, robotics, ar chitecture, and so on. This paper focuses on enhancing feature extraction for chessboard corner detection, a key step in calibration. We analyze existing methods, high lighting their limitations and propose a novel sub-pixel refinement approach based on symmetry, which signifi cantly improves accuracy for visible light cameras. Un like prior symmetry based method that assume a contin uous physical pattern, our approach accounts for abrupt changes in visible light camera images and defocus ef fects. We introduce a simplified objective function that reduces computation time and mitigates overfitting risks. Furthermore, we derive an explicit expression for the pixel value of a blurred edge, providing insights into the relationship between pixel value and center intensity. Our method demonstrates superior performance, achiev ing substantial accuracy improvements over existing tech niques, particularly in the context of visible light cam era calibration. Our code is available from https: //github.com/spdfghi/Accurate-Checkerboard Corner-Detection-under-Defoucs.git.
☆ MagicTailor: Component-Controllable Personalization in Text-to-Image Diffusion Models
Recent advancements in text-to-image (T2I) diffusion models have enabled the creation of high-quality images from text prompts, but they still struggle to generate images with precise control over specific visual concepts. Existing approaches can replicate a given concept by learning from reference images, yet they lack the flexibility for fine-grained customization of the individual component within the concept. In this paper, we introduce component-controllable personalization, a novel task that pushes the boundaries of T2I models by allowing users to reconfigure specific components when personalizing visual concepts. This task is particularly challenging due to two primary obstacles: semantic pollution, where unwanted visual elements corrupt the personalized concept, and semantic imbalance, which causes disproportionate learning of the concept and component. To overcome these challenges, we design MagicTailor, an innovative framework that leverages Dynamic Masked Degradation (DM-Deg) to dynamically perturb undesired visual semantics and Dual-Stream Balancing (DS-Bal) to establish a balanced learning paradigm for desired visual semantics. Extensive comparisons, ablations, and analyses demonstrate that MagicTailor not only excels in this challenging task but also holds significant promise for practical applications, paving the way for more nuanced and creative image generation.
comment: Project page: https://correr-zhou.github.io/MagicTailor
☆ Remember, Retrieve and Generate: Understanding Infinite Visual Concepts as Your Personalized Assistant
The development of large language models (LLMs) has significantly enhanced the capabilities of multimodal LLMs (MLLMs) as general assistants. However, lack of user-specific knowledge still restricts their application in human's daily life. In this paper, we introduce the Retrieval Augmented Personalization (RAP) framework for MLLMs' personalization. Starting from a general MLLM, we turn it into a personalized assistant in three steps. (a) Remember: We design a key-value database to store user-related information, e.g., user's name, avatar and other attributes. (b) Retrieve: When the user initiates a conversation, RAP will retrieve relevant information from the database using a multimodal retriever. (c) Generate: The input query and retrieved concepts' information are fed into MLLMs to generate personalized, knowledge-augmented responses. Unlike previous methods, RAP allows real-time concept editing via updating the external database. To further improve generation quality and alignment with user-specific information, we design a pipeline for data collection and create a specialized dataset for personalized training of MLLMs. Based on the dataset, we train a series of MLLMs as personalized multimodal assistants. By pretraining on large-scale dataset, RAP-MLLMs can generalize to infinite visual concepts without additional finetuning. Our models demonstrate outstanding flexibility and generation quality across a variety of tasks, such as personalized image captioning, question answering and visual recognition. The code, data and models are available at https://github.com/Hoar012/RAP-MLLM.
Self-Supervised Scene Flow Estimation with Point-Voxel Fusion and Surface Representation ICASSP 2025
Scene flow estimation aims to generate the 3D motion field of points between two consecutive frames of point clouds, which has wide applications in various fields. Existing point-based methods ignore the irregularity of point clouds and have difficulty capturing long-range dependencies due to the inefficiency of point-level computation. Voxel-based methods suffer from the loss of detail information. In this paper, we propose a point-voxel fusion method, where we utilize a voxel branch based on sparse grid attention and the shifted window strategy to capture long-range dependencies and a point branch to capture fine-grained features to compensate for the information loss in the voxel branch. In addition, since xyz coordinates are difficult to describe the geometric structure of complex 3D objects in the scene, we explicitly encode the local surface information of the point cloud through the umbrella surface feature extraction (USFE) module. We verify the effectiveness of our method by conducting experiments on the Flyingthings3D and KITTI datasets. Our method outperforms all other self-supervised methods and achieves highly competitive results compared to fully supervised methods. We achieve improvements in all metrics, especially EPE, which is reduced by 8.51% and 10.52% on the KITTIo and KITTIs datasets, respectively.
comment: The paper is under consideration at 2025 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2025)
☆ GlossyGS: Inverse Rendering of Glossy Objects with 3D Gaussian Splatting
Reconstructing objects from posed images is a crucial and complex task in computer graphics and computer vision. While NeRF-based neural reconstruction methods have exhibited impressive reconstruction ability, they tend to be time-comsuming. Recent strategies have adopted 3D Gaussian Splatting (3D-GS) for inverse rendering, which have led to quick and effective outcomes. However, these techniques generally have difficulty in producing believable geometries and materials for glossy objects, a challenge that stems from the inherent ambiguities of inverse rendering. To address this, we introduce GlossyGS, an innovative 3D-GS-based inverse rendering framework that aims to precisely reconstruct the geometry and materials of glossy objects by integrating material priors. The key idea is the use of micro-facet geometry segmentation prior, which helps to reduce the intrinsic ambiguities and improve the decomposition of geometries and materials. Additionally, we introduce a normal map prefiltering strategy to more accurately simulate the normal distribution of reflective surfaces. These strategies are integrated into a hybrid geometry and material representation that employs both explicit and implicit methods to depict glossy objects. We demonstrate through quantitative analysis and qualitative visualization that the proposed method is effective to reconstruct high-fidelity geometries and materials of glossy objects, and performs favorably against state-of-the-arts.
☆ Mitigating Hallucinations in Large Vision-Language Models via Summary-Guided Decoding
Large Vision-Language Models (LVLMs) demonstrate impressive capabilities in generating detailed and coherent responses from visual inputs. However, they are prone to generate hallucinations due to an over-reliance on language priors. To address this issue, we investigate the language priors in LVLMs and make two key observations: (1) Even when predicting the tokens associated with image-related part-of-speech (POS), models increasingly rely on linguistic priors as the token sequences grow, thereby amplifying hallucinations. (2) Methods that directly calibrate LVLM's output distribution to mitigate language priors can lead to a degradation in text quality or even exacerbate hallucinations. Based on these findings, we propose a novel method, Summary-Guided Decoding (SGD). This method naturally encourages the model to focus more on image information by reducing the text context through summaries, while controlling only the image-related POS tokens to maintain text quality. Through experiments, we demonstrate that SGD achieves state-of-the-art performance on object hallucination benchmarks. Furthermore, in terms of the trade-off between precision and recall, SGD achieves Pareto optimality among the existing methods. Lastly, we observe that although existing methods struggle to balance the reduction of object hallucinations with maintaining text quality, SGD demonstrates robustness in handling this challenge.
☆ Inadequate contrast ratio of road markings as an indicator for ADAS failure
Road markings were reported as critical road safety features, equally needed for both human drivers and for machine vision technologies utilised by advanced driver assistance systems (ADAS) and in driving automation. Visibility of road markings is achieved because of their colour contrasting with the roadway surface. During recent testing of an open-source camera-based ADAS under several visibility conditions (day, night, rain, glare), significant failures in trajectory planning were recorded and quantified. Consistently, better ADAS reliability under poor visibility conditions was achieved with Type II road markings (i.e. structured markings, facilitating moisture drainage) as compared to Type I road marking (i.e. flat lines). To further understand these failures, analysis of contrast ratio of road markings, which the tested ADAS was detecting for traffic lane recognition, was performed. The highest contrast ratio (greater than 0.5, calculated per Michelson equation) was measured at night in the absence of confounding factors, with statistically significant difference of 0.1 in favour of Type II road markings over Type I. Under daylight conditions, contrast ratio was reduced, with slightly higher values measured with Type I. The presence of rain or wet roads caused the deterioration of the contrast ratio, with Type II road markings exhibiting significantly higher contrast ratio than Type I, even though the values were low (less than 0.1). These findings matched the output of the ADAS related to traffic lane detection and underlined the importance of road marking visibility. Inadequate lane recognition by ADAS was associated with very low contrast ratio of road markings indeed. Importantly, specific minimum contrast ratio value could not be found, which was due to the complexity of ADAS algorithms...
comment: IRF World Congress 2024
☆ Precipitation Nowcasting Using Diffusion Transformer with Causal Attention
Short-term precipitation forecasting remains challenging due to the difficulty in capturing long-term spatiotemporal dependencies. Current deep learning methods fall short in establishing effective dependencies between conditions and forecast results, while also lacking interpretability. To address this issue, we propose a Precipitation Nowcasting Using Diffusion Transformer with Causal Attention model. Our model leverages Transformer and combines causal attention mechanisms to establish spatiotemporal queries between conditional information (causes) and forecast results (results). This design enables the model to effectively capture long-term dependencies, allowing forecast results to maintain strong causal relationships with input conditions over a wide range of time and space. We explore four variants of spatiotemporal information interactions for DTCA, demonstrating that global spatiotemporal labeling interactions yield the best performance. In addition, we introduce a Channel-To-Batch shift operation to further enhance the model's ability to represent complex rainfall dynamics. We conducted experiments on two datasets. Compared to state-of-the-art U-Net-based methods, our approach improved the CSI (Critical Success Index) for predicting heavy precipitation by approximately 15% and 8% respectively, achieving state-of-the-art performance.
☆ Enhancing Dataset Distillation via Label Inconsistency Elimination and Learning Pattern Refinement ECCV 2024
Dataset Distillation (DD) seeks to create a condensed dataset that, when used to train a model, enables the model to achieve performance similar to that of a model trained on the entire original dataset. It relieves the model training from processing massive data and thus reduces the computation resources, storage, and time costs. This paper illustrates our solution that ranks 1st in the ECCV-2024 Data Distillation Challenge (track 1). Our solution, Modified Difficulty-Aligned Trajectory Matching (M-DATM), introduces two key modifications to the original state-of-the-art method DATM: (1) the soft labels learned by DATM do not achieve one-to-one correspondence with the counterparts generated by the official evaluation script, so we remove the soft labels technique to alleviate such inconsistency; (2) since the removal of soft labels makes it harder for the synthetic dataset to learn late trajectory information, particularly on Tiny ImageNet, we reduce the matching range, allowing the synthetic data to concentrate more on the easier patterns. In the final evaluation, our M-DATM achieved accuracies of 0.4061 and 0.1831 on the CIFAR-100 and Tiny ImageNet datasets, ranking 1st in the Fixed Images Per Class (IPC) Track.
comment: ECCV 2024 Dataset Distillation Challenge
☆ Reference-Based Post-OCR Processing with LLM for Diacritic Languages
Extracting fine-grained OCR text from aged documents in diacritic languages remains challenging due to unexpected artifacts, time-induced degradation, and lack of datasets. While standalone spell correction approaches have been proposed, they show limited performance for historical documents due to numerous possible OCR error combinations and differences between modern and classical corpus distributions. We propose a method utilizing available content-focused ebooks as a reference base to correct imperfect OCR-generated text, supported by large language models. This technique generates high-precision pseudo-page-to-page labels for diacritic languages, where small strokes pose significant challenges in historical conditions. The pipeline eliminates various types of noise from aged documents and addresses issues such as missing characters, words, and disordered sequences. Our post-processing method, which generated a large OCR dataset of classical Vietnamese books, achieved a mean grading score of 8.72 on a 10-point scale. This outperformed the state-of-the-art transformer-based Vietnamese spell correction model, which scored 7.03 when evaluated on a sampled subset of the dataset. We also trained a baseline OCR model to assess and compare it with well-known engines. Experimental results demonstrate the strength of our baseline model compared to widely used open-source solutions. The resulting dataset will be released publicly to support future studies.
☆ PiLocNet: Physics-informed neural network on 3D localization with rotating point spread function
For the 3D localization problem using point spread function (PSF) engineering, we propose a novel enhancement of our previously introduced localization neural network, LocNet. The improved network is a physics-informed neural network (PINN) that we call PiLocNet. Previous works on the localization problem may be categorized separately into model-based optimization and neural network approaches. Our PiLocNet combines the unique strengths of both approaches by incorporating forward-model-based information into the network via a data-fitting loss term that constrains the neural network to yield results that are physically sensible. We additionally incorporate certain regularization terms from the variational method, which further improves the robustness of the network in the presence of image noise, as we show for the Poisson and Gaussian noise models. This framework accords interpretability to the neural network, and the results we obtain show its superiority. Although the paper focuses on the use of single-lobe rotating PSF to encode the full 3D source location, we expect the method to be widely applicable to other PSFs and imaging problems that are constrained by known forward processes.
comment: 25 pages, 4 figures
☆ LESS: Label-Efficient and Single-Stage Referring 3D Segmentation
Referring 3D Segmentation is a visual-language task that segments all points of the specified object from a 3D point cloud described by a sentence of query. Previous works perform a two-stage paradigm, first conducting language-agnostic instance segmentation then matching with given text query. However, the semantic concepts from text query and visual cues are separately interacted during the training, and both instance and semantic labels for each object are required, which is time consuming and human-labor intensive. To mitigate these issues, we propose a novel Referring 3D Segmentation pipeline, Label-Efficient and Single-Stage, dubbed LESS, which is only under the supervision of efficient binary mask. Specifically, we design a Point-Word Cross-Modal Alignment module for aligning the fine-grained features of points and textual embedding. Query Mask Predictor module and Query-Sentence Alignment module are introduced for coarse-grained alignment between masks and query. Furthermore, we propose an area regularization loss, which coarsely reduces irrelevant background predictions on a large scale. Besides, a point-to-point contrastive loss is proposed concentrating on distinguishing points with subtly similar features. Through extensive experiments, we achieve state-of-the-art performance on ScanRefer dataset by surpassing the previous methods about 3.7% mIoU using only binary labels.
☆ Composing Novel Classes: A Concept-Driven Approach to Generalized Category Discovery
We tackle the generalized category discovery (GCD) problem, which aims to discover novel classes in unlabeled datasets by leveraging the knowledge of known classes. Previous works utilize the known class knowledge through shared representation spaces. Despite their progress, our analysis experiments show that novel classes can achieve impressive clustering results on the feature space of a known class pre-trained model, suggesting that existing methods may not fully utilize known class knowledge. To address it, we introduce a novel concept learning framework for GCD, named ConceptGCD, that categorizes concepts into two types: derivable and underivable from known class concepts, and adopts a stage-wise learning strategy to learn them separately. Specifically, our framework first extracts known class concepts by a known class pre-trained model and then produces derivable concepts from them by a generator layer with a covariance-augmented loss. Subsequently, we expand the generator layer to learn underivable concepts in a balanced manner ensured by a concept score normalization strategy and integrate a contrastive loss to preserve previously learned concepts. Extensive experiments on various benchmark datasets demonstrate the superiority of our approach over the previous state-of-the-art methods. Code will be available soon.
comment: Underreview. The first two authors contribute equally
☆ Hybrid bundle-adjusting 3D Gaussians for view consistent rendering with pose optimization
Novel view synthesis has made significant progress in the field of 3D computer vision. However, the rendering of view-consistent novel views from imperfect camera poses remains challenging. In this paper, we introduce a hybrid bundle-adjusting 3D Gaussians model that enables view-consistent rendering with pose optimization. This model jointly extract image-based and neural 3D representations to simultaneously generate view-consistent images and camera poses within forward-facing scenes. The effective of our model is demonstrated through extensive experiments conducted on both real and synthetic datasets. These experiments clearly illustrate that our model can effectively optimize neural scene representations while simultaneously resolving significant camera pose misalignments. The source code is available at https://github.com/Bistu3DV/hybridBA.
comment: Photonics Asia 2024
☆ Inductive Gradient Adjustment For Spectral Bias In Implicit Neural Representations
Implicit Neural Representations (INRs), as a versatile representation paradigm, have achieved success in various computer vision tasks. Due to the spectral bias of the vanilla multi-layer perceptrons (MLPs), existing methods focus on designing MLPs with sophisticated architectures or repurposing training techniques for highly accurate INRs. In this paper, we delve into the linear dynamics model of MLPs and theoretically identify the empirical Neural Tangent Kernel (eNTK) matrix as a reliable link between spectral bias and training dynamics. Based on eNTK matrix, we propose a practical inductive gradient adjustment method, which could purposefully improve the spectral bias via inductive generalization of eNTK-based gradient transformation matrix. We evaluate our method on different INRs tasks with various INR architectures and compare to existing training techniques. The superior representation performance clearly validates the advantage of our proposed method. Armed with our gradient adjustment method, better INRs with more enhanced texture details and sharpened edges can be learned from data by tailored improvements on spectral bias.
comment: 28 pages, 12 figures
☆ Fundus to Fluorescein Angiography Video Generation as a Retinal Generative Foundation Model
Fundus fluorescein angiography (FFA) is crucial for diagnosing and monitoring retinal vascular issues but is limited by its invasive nature and restricted accessibility compared to color fundus (CF) imaging. Existing methods that convert CF images to FFA are confined to static image generation, missing the dynamic lesional changes. We introduce Fundus2Video, an autoregressive generative adversarial network (GAN) model that generates dynamic FFA videos from single CF images. Fundus2Video excels in video generation, achieving an FVD of 1497.12 and a PSNR of 11.77. Clinical experts have validated the fidelity of the generated videos. Additionally, the model's generator demonstrates remarkable downstream transferability across ten external public datasets, including blood vessel segmentation, retinal disease diagnosis, systemic disease prediction, and multimodal retrieval, showcasing impressive zero-shot and few-shot capabilities. These findings position Fundus2Video as a powerful, non-invasive alternative to FFA exams and a versatile retinal generative foundation model that captures both static and temporal retinal features, enabling the representation of complex inter-modality relationships.
☆ Latent Image and Video Resolution Prediction using Convolutional Neural Networks ICIP
This paper introduces a Video Quality Assessment (VQA) problem that has received little attention in the literature, called the latent resolution prediction problem. The problem arises when images or videos are upscaled from their native resolution and are reported as having a higher resolution than their native resolution. This paper formulates the problem, constructs a dataset for training and evaluation, and introduces several machine learning algorithms, including two Convolutional Neural Networks (CNNs), to address this problem. Experiments indicate that some proposed methods can predict the latent video resolution with about 95% accuracy.
comment: Submitted in ICIP conference
☆ UniG: Modelling Unitary 3D Gaussians for View-consistent 3D Reconstruction
In this work, we present UniG, a view-consistent 3D reconstruction and novel view synthesis model that generates a high-fidelity representation of 3D Gaussians from sparse images. Existing 3D Gaussians-based methods usually regress Gaussians per-pixel of each view, create 3D Gaussians per view separately, and merge them through point concatenation. Such a view-independent reconstruction approach often results in a view inconsistency issue, where the predicted positions of the same 3D point from different views may have discrepancies. To address this problem, we develop a DETR (DEtection TRansformer)-like framework, which treats 3D Gaussians as decoder queries and updates their parameters layer by layer by performing multi-view cross-attention (MVDFA) over multiple input images. In this way, multiple views naturally contribute to modeling a unitary representation of 3D Gaussians, thereby making 3D reconstruction more view-consistent. Moreover, as the number of 3D Gaussians used as decoder queries is irrespective of the number of input views, allow an arbitrary number of input images without causing memory explosion. Extensive experiments validate the advantages of our approach, showcasing superior performance over existing methods quantitatively (improving PSNR by 4.2 dB when trained on Objaverse and tested on the GSO benchmark) and qualitatively.
☆ Golyadkin's Torment: Doppelgängers and Adversarial Vulnerability
Many machine learning (ML) classifiers are claimed to outperform humans, but they still make mistakes that humans do not. The most notorious examples of such mistakes are adversarial visual metamers. This paper aims to define and investigate the phenomenon of adversarial Doppelgangers (AD), which includes adversarial visual metamers, and to compare the performance and robustness of ML classifiers to human performance. We find that AD are inputs that are close to each other with respect to a perceptual metric defined in this paper. AD are qualitatively different from the usual adversarial examples. The vast majority of classifiers are vulnerable to AD and robustness-accuracy trade-offs may not improve them. Some classification problems may not admit any AD robust classifiers because the underlying classes are ambiguous. We provide criteria that can be used to determine whether a classification problem is well defined or not; describe the structure and attributes of an AD-robust classifier; introduce and explore the notions of conceptual entropy and regions of conceptual ambiguity for classifiers that are vulnerable to AD attacks, along with methods to bound the AD fooling rate of an attack. We define the notion of classifiers that exhibit hypersensitive behavior, that is, classifiers whose only mistakes are adversarial Doppelgangers. Improving the AD robustness of hyper-sensitive classifiers is equivalent to improving accuracy. We identify conditions guaranteeing that all classifiers with sufficiently high accuracy are hyper-sensitive. Our findings are aimed at significant improvements in the reliability and security of machine learning systems.
☆ Scalable Drift Monitoring in Medical Imaging AI
The integration of artificial intelligence (AI) into medical imaging has advanced clinical diagnostics but poses challenges in managing model drift and ensuring long-term reliability. To address these challenges, we develop MMC+, an enhanced framework for scalable drift monitoring, building upon the CheXstray framework that introduced real-time drift detection for medical imaging AI models using multi-modal data concordance. This work extends the original framework's methodologies, providing a more scalable and adaptable solution for real-world healthcare settings and offers a reliable and cost-effective alternative to continuous performance monitoring addressing limitations of both continuous and periodic monitoring methods. MMC+ introduces critical improvements to the original framework, including more robust handling of diverse data streams, improved scalability with the integration of foundation models like MedImageInsight for high-dimensional image embeddings without site-specific training, and the introduction of uncertainty bounds to better capture drift in dynamic clinical environments. Validated with real-world data from Massachusetts General Hospital during the COVID-19 pandemic, MMC+ effectively detects significant data shifts and correlates them with model performance changes. While not directly predicting performance degradation, MMC+ serves as an early warning system, indicating when AI systems may deviate from acceptable performance bounds and enabling timely interventions. By emphasizing the importance of monitoring diverse data streams and evaluating data shifts alongside model performance, this work contributes to the broader adoption and integration of AI solutions in clinical settings.
☆ FAMSeC: A Few-shot-sample-based General AI-generated Image Detection Method
The explosive growth of generative AI has saturated the internet with AI-generated images, raising security concerns and increasing the need for reliable detection methods. The primary requirement for such detection is generalizability, typically achieved by training on numerous fake images from various models. However, practical limitations, such as closed-source models and restricted access, often result in limited training samples. Therefore, training a general detector with few-shot samples is essential for modern detection mechanisms. To address this challenge, we propose FAMSeC, a general AI-generated image detection method based on LoRA-based Forgery Awareness Module and Semantic feature-guided Contrastive learning strategy. To effectively learn from limited samples and prevent overfitting, we developed a Forgery Awareness Module (FAM) based on LoRA, maintaining the generalization of pre-trained features. Additionally, to cooperate with FAM, we designed a Semantic feature-guided Contrastive learning strategy (SeC), making the FAM focus more on the differences between real/fake image than on the features of the samples themselves. Experiments show that FAMSeC outperforms state-of-the-art method, enhancing classification accuracy by 14.55% with just 0.56% of the training samples.
☆ Utilizing Large Language Models in An Iterative Paradigm with Domain Feedback for Molecule Optimization
Molecule optimization is a critical task in drug discovery to optimize desired properties of a given molecule through chemical modification. Despite Large Language Models (LLMs) holding the potential to efficiently simulate this task by using natural language to direct the optimization, straightforwardly utilizing shows limited performance. In this work, we facilitate utilizing LLMs in an iterative paradigm by proposing a simple yet highly effective domain feedback provider, namely $\text{Re}^2$DF. In detail, $\text{Re}^2$DF harnesses an external toolkit, RDKit, to handle the molecule hallucination, if the modified molecule is chemically invalid. Otherwise, its desired properties are computed and compared to the original one, establishing reliable domain feedback with correct direction and distance towards the objective, followed by a retrieved example, to explicitly guide the LLM to refine the modified molecule. We conduct experiments across both single- and multi-property objectives with 2 thresholds, where $\text{Re}^2$DF shows significant improvements. Particularly, for 20 single-property objectives, $\text{Re}^2$DF enhances the Hit ratio by 16.95\% and 20.76\% under loose and strict thresholds, respectively. For 32 multi-property objectives, $\text{Re}^2$DF enhances the Hit ratio by 6.04\% and 5.25\%.
☆ Mapping Bias in Vision Language Models: Signposts, Pitfalls, and the Road Ahead NAACL 2025
As Vision Language Models (VLMs) gain widespread use, their fairness remains under-explored. In this paper, we analyze demographic biases across five models and six datasets. We find that portrait datasets like UTKFace and CelebA are the best tools for bias detection, finding gaps in performance and fairness between LLaVa and CLIP models. However, scene based datasets like PATA, VLStereoSet fail to be useful benchmarks for bias due to their construction. As for pronoun based datasets like VisoGender, we receive mixed signals as only some subsets of the data are useful in providing insights. To alleviate this problem, we introduce a more difficult version of VisoGender to serve as a more rigorous evaluation. Based on these results, we call for more effective and carefully designed datasets to ensure VLMs are both fair and reliable.
comment: Under Review at NAACL 2025
☆ See Behind Walls in Real-time Using Aerial Drones and Augmented Reality
This work presents ARD2, a framework that enables real-time through-wall surveillance using two aerial drones and an augmented reality (AR) device. ARD2 consists of two main steps: target direction estimation and contour reconstruction. In the first stage, ARD2 leverages geometric relationships between the drones, the user, and the target to project the target's direction onto the user's AR display. In the second stage, images from the drones are synthesized to reconstruct the target's contour, allowing the user to visualize the target behind walls. Experimental results demonstrate the system's accuracy in both direction estimation and contour reconstruction.
comment: 6 pages
☆ Unlocking the Capabilities of Masked Generative Models for Image Synthesis via Self-Guidance NeurIPS 2024
Masked generative models (MGMs) have shown impressive generative ability while providing an order of magnitude efficient sampling steps compared to continuous diffusion models. However, MGMs still underperform in image synthesis compared to recent well-developed continuous diffusion models with similar size in terms of quality and diversity of generated samples. A key factor in the performance of continuous diffusion models stems from the guidance methods, which enhance the sample quality at the expense of diversity. In this paper, we extend these guidance methods to generalized guidance formulation for MGMs and propose a self-guidance sampling method, which leads to better generation quality. The proposed approach leverages an auxiliary task for semantic smoothing in vector-quantized token space, analogous to the Gaussian blur in continuous pixel space. Equipped with the parameter-efficient fine-tuning method and high-temperature sampling, MGMs with the proposed self-guidance achieve a superior quality-diversity trade-off, outperforming existing sampling methods in MGMs with more efficient training and sampling costs. Extensive experiments with the various sampling hyperparameters confirm the effectiveness of the proposed self-guidance.
comment: NeurIPS 2024. Code is available at: https://github.com/JiwanHur/UnlockMGM
☆ Boosting Imperceptibility of Stable Diffusion-based Adversarial Examples Generation with Momentum
We propose a novel framework, Stable Diffusion-based Momentum Integrated Adversarial Examples (SD-MIAE), for generating adversarial examples that can effectively mislead neural network classifiers while maintaining visual imperceptibility and preserving the semantic similarity to the original class label. Our method leverages the text-to-image generation capabilities of the Stable Diffusion model by manipulating token embeddings corresponding to the specified class in its latent space. These token embeddings guide the generation of adversarial images that maintain high visual fidelity. The SD-MIAE framework consists of two phases: (1) an initial adversarial optimization phase that modifies token embeddings to produce misclassified yet natural-looking images and (2) a momentum-based optimization phase that refines the adversarial perturbations. By introducing momentum, our approach stabilizes the optimization of perturbations across iterations, enhancing both the misclassification rate and visual fidelity of the generated adversarial examples. Experimental results demonstrate that SD-MIAE achieves a high misclassification rate of 79%, improving by 35% over the state-of-the-art method while preserving the imperceptibility of adversarial perturbations and the semantic similarity to the original class label, making it a practical method for robust adversarial evaluation.
comment: 10 pages, 12 figures. To be published in IEEE TPS 2024 Proceedings. Code available on GitHub: https://github.com/nashrahhaque/SD-MIAE
☆ Trust but Verify: Programmatic VLM Evaluation in the Wild
Vision-Language Models (VLMs) often generate plausible but incorrect responses to visual queries. However, reliably quantifying the effect of such hallucinations in free-form responses to open-ended queries is challenging as it requires visually verifying each claim within the response. We propose Programmatic VLM Evaluation (PROVE), a new benchmarking paradigm for evaluating VLM responses to open-ended queries. To construct PROVE, we provide a large language model (LLM) with a high-fidelity scene-graph representation constructed from a hyper-detailed image caption, and prompt it to generate diverse question-answer (QA) pairs, as well as programs that can be executed over the scene graph object to verify each QA pair. We thus construct a benchmark of 10.5k challenging but visually grounded QA pairs. Next, to evaluate free-form model responses to queries in PROVE, we propose a programmatic evaluation strategy that measures both the helpfulness and truthfulness of a response within a unified scene graph-based framework. We benchmark the helpfulness-truthfulness trade-offs of a range of VLMs on PROVE, finding that very few are in-fact able to achieve a good balance between the two. Project page: \url{https://prove-explorer.netlify.app/}.
☆ Adversarial Neural Networks in Medical Imaging Advancements and Challenges in Semantic Segmentation
Recent advancements in artificial intelligence (AI) have precipitated a paradigm shift in medical imaging, particularly revolutionizing the domain of brain imaging. This paper systematically investigates the integration of deep learning -- a principal branch of AI -- into the semantic segmentation of brain images. Semantic segmentation serves as an indispensable technique for the delineation of discrete anatomical structures and the identification of pathological markers, essential for the diagnosis of complex neurological disorders. Historically, the reliance on manual interpretation by radiologists, while noteworthy for its accuracy, is plagued by inherent subjectivity and inter-observer variability. This limitation becomes more pronounced with the exponential increase in imaging data, which traditional methods struggle to process efficiently and effectively. In response to these challenges, this study introduces the application of adversarial neural networks, a novel AI approach that not only automates but also refines the semantic segmentation process. By leveraging these advanced neural networks, our approach enhances the precision of diagnostic outputs, reducing human error and increasing the throughput of imaging data analysis. The paper provides a detailed discussion on how adversarial neural networks facilitate a more robust, objective, and scalable solution, thereby significantly improving diagnostic accuracies in neurological evaluations. This exploration highlights the transformative impact of AI on medical imaging, setting a new benchmark for future research and clinical practice in neurology.
☆ MMAD-Purify: A Precision-Optimized Framework for Efficient and Scalable Multi-Modal Attacks
Neural networks have achieved remarkable performance across a wide range of tasks, yet they remain susceptible to adversarial perturbations, which pose significant risks in safety-critical applications. With the rise of multimodality, diffusion models have emerged as powerful tools not only for generative tasks but also for various applications such as image editing, inpainting, and super-resolution. However, these models still lack robustness due to limited research on attacking them to enhance their resilience. Traditional attack techniques, such as gradient-based adversarial attacks and diffusion model-based methods, are hindered by computational inefficiencies and scalability issues due to their iterative nature. To address these challenges, we introduce an innovative framework that leverages the distilled backbone of diffusion models and incorporates a precision-optimized noise predictor to enhance the effectiveness of our attack framework. This approach not only enhances the attack's potency but also significantly reduces computational costs. Our framework provides a cutting-edge solution for multi-modal adversarial attacks, ensuring reduced latency and the generation of high-fidelity adversarial examples with superior success rates. Furthermore, we demonstrate that our framework achieves outstanding transferability and robustness against purification defenses, outperforming existing gradient-based attack models in both effectiveness and efficiency.
☆ Your Interest, Your Summaries: Query-Focused Long Video Summarization
Generating a concise and informative video summary from a long video is important, yet subjective due to varying scene importance. Users' ability to specify scene importance through text queries enhances the relevance of such summaries. This paper introduces an approach for query-focused video summarization, aiming to align video summaries closely with user queries. To this end, we propose the Fully Convolutional Sequence Network with Attention (FCSNA-QFVS), a novel approach designed for this task. Leveraging temporal convolutional and attention mechanisms, our model effectively extracts and highlights relevant content based on user-specified queries. Experimental validation on a benchmark dataset for query-focused video summarization demonstrates the effectiveness of our approach.
comment: To appear at the 18th International Conference on Control, Automation, Robotics and Vision (ICARCV), December 2024, Dubai, UAE
☆ Self Supervised Deep Learning for Robot Grasping
Learning Based Robot Grasping currently involves the use of labeled data. This approach has two major disadvantages. Firstly, labeling data for grasp points and angles is a strenuous process, so the dataset remains limited. Secondly, human labeling is prone to bias due to semantics. In order to solve these problems we propose a simpler self-supervised robotic setup, that will train a Convolutional Neural Network (CNN). The robot will label and collect the data during the training process. The idea is to make a robot that is less costly, small and easily maintainable in a lab setup. The robot will be trained on a large data set for several hundred hours and then the trained Neural Network can be mapped onto a larger grasping robot.
☆ SAMReg: SAM-enabled Image Registration with ROI-based Correspondence
This paper describes a new spatial correspondence representation based on paired regions-of-interest (ROIs), for medical image registration. The distinct properties of the proposed ROI-based correspondence are discussed, in the context of potential benefits in clinical applications following image registration, compared with alternative correspondence-representing approaches, such as those based on sampled displacements and spatial transformation functions. These benefits include a clear connection between learning-based image registration and segmentation, which in turn motivates two cases of image registration approaches using (pre-)trained segmentation networks. Based on the segment anything model (SAM), a vision foundation model for segmentation, we develop a new registration algorithm SAMReg, which does not require any training (or training data), gradient-based fine-tuning or prompt engineering. The proposed SAMReg models are evaluated across five real-world applications, including intra-subject registration tasks with cardiac MR and lung CT, challenging inter-subject registration scenarios with prostate MR and retinal imaging, and an additional evaluation with a non-clinical example with aerial image registration. The proposed methods outperform both intensity-based iterative algorithms and DDF-predicting learning-based networks across tested metrics including Dice and target registration errors on anatomical structures, and further demonstrates competitive performance compared to weakly-supervised registration approaches that rely on fully-segmented training data. Open source code and examples are available at: https://github.com/sqhuang0103/SAMReg.git.
☆ Efficient Vision-Language Models by Summarizing Visual Tokens into Compact Registers
Recent advancements in vision-language models (VLMs) have expanded their potential for real-world applications, enabling these models to perform complex reasoning on images. In the widely used fully autoregressive transformer-based models like LLaVA, projected visual tokens are prepended to textual tokens. Oftentimes, visual tokens are significantly more than prompt tokens, resulting in increased computational overhead during both training and inference. In this paper, we propose Visual Compact Token Registers (Victor), a method that reduces the number of visual tokens by summarizing them into a smaller set of register tokens. Victor adds a few learnable register tokens after the visual tokens and summarizes the visual information into these registers using the first few layers in the language tower of VLMs. After these few layers, all visual tokens are discarded, significantly improving computational efficiency for both training and inference. Notably, our method is easy to implement and requires a small number of new trainable parameters with minimal impact on model performance. In our experiment, with merely 8 visual registers--about 1% of the original tokens--Victor shows less than a 4% accuracy drop while reducing the total training time by 43% and boosting the inference throughput by 3.3X.
☆ FaceSaliencyAug: Mitigating Geographic, Gender and Stereotypical Biases via Saliency-Based Data Augmentation
Geographical, gender and stereotypical biases in computer vision models pose significant challenges to their performance and fairness. {In this study, we present an approach named FaceSaliencyAug aimed at addressing the gender bias in} {Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs). Leveraging the salient regions} { of faces detected by saliency, the propose approach mitigates geographical and stereotypical biases } {in the datasets. FaceSaliencyAug} randomly selects masks from a predefined search space and applies them to the salient region of face images, subsequently restoring the original image with masked salient region. {The proposed} augmentation strategy enhances data diversity, thereby improving model performance and debiasing effects. We quantify dataset diversity using Image Similarity Score (ISS) across five datasets, including Flickr Faces HQ (FFHQ), WIKI, IMDB, Labelled Faces in the Wild (LFW), UTK Faces, and Diverse Dataset. The proposed approach demonstrates superior diversity metrics, as evaluated by ISS-intra and ISS-inter algorithms. Furthermore, we evaluate the effectiveness of our approach in mitigating gender bias on CEO, Engineer, Nurse, and School Teacher datasets. We use the Image-Image Association Score (IIAS) to measure gender bias in these occupations. Our experiments reveal a reduction in gender bias for both CNNs and ViTs, indicating the efficacy of our method in promoting fairness and inclusivity in computer vision models.
comment: Accepted at Image Signal and Video processing
☆ On Partial Prototype Collapse in the DINO Family of Self-Supervised Methods BMVC 2024
A prominent self-supervised learning paradigm is to model the representations as clusters, or more generally as a mixture model. Learning to map the data samples to compact representations and fitting the mixture model simultaneously leads to the representation collapse problem. Regularizing the distribution of data points over the clusters is the prevalent strategy to avoid this issue. While this is sufficient to prevent full representation collapse, we show that a partial prototype collapse problem still exists in the DINO family of methods, that leads to significant redundancies in the prototypes. Such prototype redundancies serve as shortcuts for the method to achieve a marginal latent class distribution that matches the prescribed prior. We show that by encouraging the model to use diverse prototypes, the partial prototype collapse can be mitigated. Effective utilization of the prototypes enables the methods to learn more fine-grained clusters, encouraging more informative representations. We demonstrate that this is especially beneficial when pre-training on a long-tailed fine-grained dataset.
comment: First version of the paper appeared in OpenReview on 22 Sep 2023. Accepted to BMVC 2024
☆ Learning Multimodal Cues of Children's Uncertainty SIGDIAL 2023
Understanding uncertainty plays a critical role in achieving common ground (Clark et al.,1983). This is especially important for multimodal AI systems that collaborate with users to solve a problem or guide the user through a challenging concept. In this work, for the first time, we present a dataset annotated in collaboration with developmental and cognitive psychologists for the purpose of studying nonverbal cues of uncertainty. We then present an analysis of the data, studying different roles of uncertainty and its relationship with task difficulty and performance. Lastly, we present a multimodal machine learning model that can predict uncertainty given a real-time video clip of a participant, which we find improves upon a baseline multimodal transformer model. This work informs research on cognitive coordination between human-human and human-AI and has broad implications for gesture understanding and generation. The anonymized version of our data and code will be publicly available upon the completion of the required consent forms and data sheets.
comment: SIGDIAL 2023
☆ Human Action Anticipation: A Survey
Predicting future human behavior is an increasingly popular topic in computer vision, driven by the interest in applications such as autonomous vehicles, digital assistants and human-robot interactions. The literature on behavior prediction spans various tasks, including action anticipation, activity forecasting, intent prediction, goal prediction, and so on. Our survey aims to tie together this fragmented literature, covering recent technical innovations as well as the development of new large-scale datasets for model training and evaluation. We also summarize the widely-used metrics for different tasks and provide a comprehensive performance comparison of existing approaches on eleven action anticipation datasets. This survey serves as not only a reference for contemporary methodologies in action anticipation, but also a guideline for future research direction of this evolving landscape.
comment: 30 pages, 9 figures, 12 tables
☆ Segmentation of Pediatric Brain Tumors using a Radiologically informed, Deep Learning Cascade
Monitoring of Diffuse Intrinsic Pontine Glioma (DIPG) and Diffuse Midline Glioma (DMG) brain tumors in pediatric patients is key for assessment of treatment response. Response Assessment in Pediatric Neuro-Oncology (RAPNO) guidelines recommend the volumetric measurement of these tumors using MRI. Segmentation challenges, such as the Brain Tumor Segmentation (BraTS) Challenge, promote development of automated approaches which are replicable, generalizable and accurate, to aid in these tasks. The current study presents a novel adaptation of existing nnU-Net approaches for pediatric brain tumor segmentation, submitted to the BraTS-PEDs 2024 challenge. We apply an adapted nnU-Net with hierarchical cascades to the segmentation task of the BraTS-PEDs 2024 challenge. The residual encoder variant of nnU-Net, used as our baseline model, already provides high quality segmentations. We incorporate multiple changes to the implementation of nnU-Net and devise a novel two-stage cascaded nnU-Net to segment the substructures of brain tumors from coarse to fine. Using outputs from the nnU-Net Residual Encoder (trained to segment CC, ED, ET and NET tumor labels from T1w, T1w-CE, T2w and T2-FLAIR MRI), these are passed to two additional models one classifying ET versus NET and a second classifying CC vs ED using cascade learning. We use radiological guidelines to steer which multi parametric MRI (mpMRI) to use in these cascading models. Compared to a default nnU-Net and an ensembled nnU-net as baseline approaches, our novel method provides robust segmentations for the BraTS-PEDs 2024 challenge, achieving mean Dice scores of 0.657, 0.904, 0.703, and 0.967, and HD95 of 76.2, 10.1, 111.0, and 12.3 for the ET, NET, CC and ED, respectively.
☆ Probabilistic U-Net with Kendall Shape Spaces for Geometry-Aware Segmentations of Images
One of the fundamental problems in computer vision is image segmentation, the task of detecting distinct regions or objects in given images. Deep Neural Networks (DNN) have been shown to be very effective in segmenting challenging images, producing convincing segmentations. There is further need for probabilistic DNNs that can reflect the uncertainties from the input images and the models into the computed segmentations, in other words, new DNNs that can generate multiple plausible segmentations and their distributions depending on the input or the model uncertainties. While there are existing probabilistic segmentation models, many of them do not take into account the geometry or shape underlying the segmented regions. In this paper, we propose a probabilistic image segmentation model that can incorporate the geometry of a segmentation. Our proposed model builds on the Probabilistic U-Net of \cite{kohl2018probabilistic} to generate probabilistic segmentations, i.e.\! multiple likely segmentations for an input image. Our model also adopts the Kendall Shape Variational Auto-Encoder of \cite{vadgama2023kendall} to encode a Kendall shape space in the latent variable layers of the prior and posterior networks of the Probabilistic U-Net. Incorporating the shape space in this manner leads to a more robust segmentation with spatially coherent regions, respecting the underlying geometry in the input images.
comment: 22 pages, 13 figures
☆ Reproducibility study of "LICO: Explainable Models with Language-Image Consistency"
The growing reproducibility crisis in machine learning has brought forward a need for careful examination of research findings. This paper investigates the claims made by Lei et al. (2023) regarding their proposed method, LICO, for enhancing post-hoc interpretability techniques and improving image classification performance. LICO leverages natural language supervision from a vision-language model to enrich feature representations and guide the learning process. We conduct a comprehensive reproducibility study, employing (Wide) ResNets and established interpretability methods like Grad-CAM and RISE. We were mostly unable to reproduce the authors' results. In particular, we did not find that LICO consistently led to improved classification performance or improvements in quantitative and qualitative measures of interpretability. Thus, our findings highlight the importance of rigorous evaluation and transparent reporting in interpretability research.
comment: 15 pages, 2 figures, Machine Learning Reproducibility Challenge 2024
☆ Debiasing Large Vision-Language Models by Ablating Protected Attribute Representations NeurIPS
Large Vision Language Models (LVLMs) such as LLaVA have demonstrated impressive capabilities as general-purpose chatbots that can engage in conversations about a provided input image. However, their responses are influenced by societal biases present in their training datasets, leading to undesirable differences in how the model responds when presented with images depicting people of different demographics. In this work, we propose a novel debiasing framework for LVLMs by directly ablating biased attributes during text generation to avoid generating text related to protected attributes, or even representing them internally. Our method requires no training and a relatively small amount of representative biased outputs (~1000 samples). Our experiments show that not only can we can minimize the propensity of LVLMs to generate text related to protected attributes, but we can even use synthetic data to inform the ablation while retaining captioning performance on real data such as COCO. Furthermore, we find the resulting generations from a debiased LVLM exhibit similar accuracy as a baseline biased model, showing that debiasing effects can be achieved without sacrificing model performance.
comment: NeurIPS workshop on SafeGenAI, 10 pages, 2 figures
☆ Satellite Streaming Video QoE Prediction: A Real-World Subjective Database and Network-Level Prediction Models
Demand for streaming services, including satellite, continues to exhibit unprecedented growth. Internet Service Providers find themselves at the crossroads of technological advancements and rising customer expectations. To stay relevant and competitive, these ISPs must ensure their networks deliver optimal video streaming quality, a key determinant of user satisfaction. Towards this end, it is important to have accurate Quality of Experience prediction models in place. However, achieving robust performance by these models requires extensive data sets labeled by subjective opinion scores on videos impaired by diverse playback disruptions. To bridge this data gap, we introduce the LIVE-Viasat Real-World Satellite QoE Database. This database consists of 179 videos recorded from real-world streaming services affected by various authentic distortion patterns. We also conducted a comprehensive subjective study involving 54 participants, who contributed both continuous-time opinion scores and endpoint (retrospective) QoE scores. Our analysis sheds light on various determinants influencing subjective QoE, such as stall events, spatial resolutions, bitrate, and certain network parameters. We demonstrate the usefulness of this unique new resource by evaluating the efficacy of prevalent QoE-prediction models on it. We also created a new model that maps the network parameters to predicted human perception scores, which can be used by ISPs to optimize the video streaming quality of their networks. Our proposed model, which we call SatQA, is able to accurately predict QoE using only network parameters, without any access to pixel data or video-specific metadata, estimated by Spearman's Rank Order Correlation Coefficient (SROCC), Pearson Linear Correlation Coefficient (PLCC), and Root Mean Squared Error (RMSE), indicating high accuracy and reliability.
☆ ARKit LabelMaker: A New Scale for Indoor 3D Scene Understanding
The performance of neural networks scales with both their size and the amount of data they have been trained on. This is shown in both language and image generation. However, this requires scaling-friendly network architectures as well as large-scale datasets. Even though scaling-friendly architectures like transformers have emerged for 3D vision tasks, the GPT-moment of 3D vision remains distant due to the lack of training data. In this paper, we introduce ARKit LabelMaker, the first large-scale, real-world 3D dataset with dense semantic annotations. Specifically, we complement ARKitScenes dataset with dense semantic annotations that are automatically generated at scale. To this end, we extend LabelMaker, a recent automatic annotation pipeline, to serve the needs of large-scale pre-training. This involves extending the pipeline with cutting-edge segmentation models as well as making it robust to the challenges of large-scale processing. Further, we push forward the state-of-the-art performance on ScanNet and ScanNet200 dataset with prevalent 3D semantic segmentation models, demonstrating the efficacy of our generated dataset.
♻ ☆ Corrective Machine Unlearning
Machine Learning models increasingly face data integrity challenges due to the use of large-scale training datasets drawn from the Internet. We study what model developers can do if they detect that some data was manipulated or incorrect. Such manipulated data can cause adverse effects including vulnerability to backdoored samples, systemic biases, and reduced accuracy on certain input domains. Realistically, all manipulated training samples cannot be identified, and only a small, representative subset of the affected data can be flagged. We formalize Corrective Machine Unlearning as the problem of mitigating the impact of data affected by unknown manipulations on a trained model, only having identified a subset of the corrupted data. We demonstrate that the problem of corrective unlearning has significantly different requirements from traditional privacy-oriented unlearning. We find most existing unlearning methods, including retraining-from-scratch without the deletion set, require most of the manipulated data to be identified for effective corrective unlearning. However, one approach, Selective Synaptic Dampening, achieves limited success, unlearning adverse effects with just a small portion of the manipulated samples in our setting, which shows encouraging signs for future progress. We hope our work spurs research towards developing better methods for corrective unlearning and offers practitioners a new strategy to handle data integrity challenges arising from web-scale training. Code is available at https://github.com/drimpossible/corrective-unlearning-bench.
comment: Published in Transactions of Machine Learning Research (TMLR), 17 pages, 7 figures
♻ ☆ Order-aware Interactive Segmentation
Interactive segmentation aims to accurately segment target objects with minimal user interactions. However, current methods often fail to accurately separate target objects from the background, due to a limited understanding of order, the relative depth between objects in a scene. To address this issue, we propose OIS: order-aware interactive segmentation, where we explicitly encode the relative depth between objects into order maps. We introduce a novel order-aware attention, where the order maps seamlessly guide the user interactions (in the form of clicks) to attend to the image features. We further present an object-aware attention module to incorporate a strong object-level understanding to better differentiate objects with similar order. Our approach allows both dense and sparse integration of user clicks, enhancing both accuracy and efficiency as compared to prior works. Experimental results demonstrate that OIS achieves state-of-the-art performance, improving mIoU after one click by 7.61 on the HQSeg44K dataset and 1.32 on the DAVIS dataset as compared to the previous state-of-the-art SegNext, while also doubling inference speed compared to current leading methods. The project page is https://ukaukaaaa.github.io/projects/OIS/index.html
comment: Interactive demo can be found in project page: https://ukaukaaaa.github.io/projects/OIS/index.html
♻ ☆ EchoApex: A General-Purpose Vision Foundation Model for Echocardiography
Quantitative evaluation of echocardiography is essential for precise assessment of cardiac condition, monitoring disease progression, and guiding treatment decisions. The diverse nature of echo images, including variations in probe types, manufacturers, and pathologies, poses challenges for developing artificial intelligent models that can generalize across different clinical practice. We introduce EchoApex, the first general-purpose vision foundation model echocardiography with applications on a variety of clinical practice. Leveraging self-supervised learning, EchoApex is pretrained on over 20 million echo images from 11 clinical centres. By incorporating task-specific decoders and adapter modules, we demonstrate the effectiveness of EchoApex on 4 different kind of clinical applications with 28 sub-tasks, including view classification, interactive structure segmentation, left ventricle hypertrophy detection and automated ejection fraction estimation from view sequences. Compared to state-of-the-art task-specific models, EchoApex attains improved performance with a unified image encoding architecture, demonstrating the benefits of model pretraining at scale with in-domain data. Furthermore, EchoApex illustrates the potential for developing a general-purpose vision foundation model tailored specifically for echocardiography, capable of addressing a diverse range of clinical applications with high efficiency and efficacy.
♻ ☆ Comprehensive Performance Evaluation of YOLO11, YOLOv10, YOLOv9 and YOLOv8 on Detecting and Counting Fruitlet in Complex Orchard Environments
This study extensively evaluated You Only Look Once (YOLO) object detection algorithms across all configurations (total 22) of YOLOv8, YOLOv9, YOLOv10, and YOLO11 for green fruit detection in commercial orchards. The research also validated in-field fruitlet counting using an iPhone and machine vision sensors across four apple varieties: Scifresh, Scilate, Honeycrisp and Cosmic Crisp. Among the 22 configurations evaluated, YOLO11s and YOLOv9 gelan-base outperformed others with mAP@50 scores of 0.933 and 0.935 respectively. In terms of recall, YOLOv9 gelan-base achieved the highest value among YOLOv9 configurations at 0.899, while YOLO11m led YOLO11 variants with 0.897. YOLO11n emerged as the fastest model, achieving fastest inference speed of only 2.4 ms, significantly outpacing the leading configurations of YOLOv10n, YOLOv9 gelan-s, and YOLOv8n, with speeds of 5.5, 11.5, and 4.1 ms, respectively. This comparative evaluation highlights the strengths of YOLO11, YOLOv9, and YOLOv10, offering researchers essential insights to choose the best-suited model for fruitlet detection and possible automation in commercial orchards. For real-time automation related work in relevant datasets, we recommend using YOLO11n due to its high detection and image processing speed. Keywords: YOLO11, YOLO11 Object Detection, YOLOv10, YOLOv9, YOLOv8, You Only Look Once, Fruitlet Detection, Greenfruit Detection, Green Apple Detection, Agricultural Automation, Artificial Intelligence, Deep Learning, Machine Learning, Zero-shot Detection
comment: 15 figures, 2 tables
♻ ☆ LieRE: Generalizing Rotary Position Encodings
While Rotary Position Embeddings (RoPE) for large language models have become widely adopted, their application for other modalities has been slower. Here, we introduce Lie group Relative position Encodings (LieRE) that goes beyond RoPE in supporting n-dimensional inputs. We evaluate the performance of LieRE on 2D and 3D image classification tasks and observe that LieRE leads to marked relative improvements in performance (up to 9.7% for 2D and up to 25.5% for 3D), training efficiency (3.5x reduction), data efficiency (30%) compared to the baselines of DeiT III, RoPE-Mixed and Vision-Llama. https://github.com/Stanford-AIMI/LieRE
♻ ☆ Beyond Coarse-Grained Matching in Video-Text Retrieval ACCV 2024
Video-text retrieval has seen significant advancements, yet the ability of models to discern subtle differences in captions still requires verification. In this paper, we introduce a new approach for fine-grained evaluation. Our approach can be applied to existing datasets by automatically generating hard negative test captions with subtle single-word variations across nouns, verbs, adjectives, adverbs, and prepositions. We perform comprehensive experiments using four state-of-the-art models across two standard benchmarks (MSR-VTT and VATEX) and two specially curated datasets enriched with detailed descriptions (VLN-UVO and VLN-OOPS), resulting in a number of novel insights: 1) our analyses show that the current evaluation benchmarks fall short in detecting a model's ability to perceive subtle single-word differences, 2) our fine-grained evaluation highlights the difficulty models face in distinguishing such subtle variations. To enhance fine-grained understanding, we propose a new baseline that can be easily combined with current methods. Experiments on our fine-grained evaluations demonstrate that this approach enhances a model's ability to understand fine-grained differences.
comment: Accepted to ACCV 2024
♻ ☆ FlashTex: Fast Relightable Mesh Texturing with LightControlNet
Manually creating textures for 3D meshes is time-consuming, even for expert visual content creators. We propose a fast approach for automatically texturing an input 3D mesh based on a user-provided text prompt. Importantly, our approach disentangles lighting from surface material/reflectance in the resulting texture so that the mesh can be properly relit and rendered in any lighting environment. We introduce LightControlNet, a new text-to-image model based on the ControlNet architecture, which allows the specification of the desired lighting as a conditioning image to the model. Our text-to-texture pipeline then constructs the texture in two stages. The first stage produces a sparse set of visually consistent reference views of the mesh using LightControlNet. The second stage applies a texture optimization based on Score Distillation Sampling (SDS) that works with LightControlNet to increase the texture quality while disentangling surface material from lighting. Our algorithm is significantly faster than previous text-to-texture methods, while producing high-quality and relightable textures.
comment: Project page: https://flashtex.github.io/
♻ ☆ PTQ4DiT: Post-training Quantization for Diffusion Transformers NeurIPS 2024
The recent introduction of Diffusion Transformers (DiTs) has demonstrated exceptional capabilities in image generation by using a different backbone architecture, departing from traditional U-Nets and embracing the scalable nature of transformers. Despite their advanced capabilities, the wide deployment of DiTs, particularly for real-time applications, is currently hampered by considerable computational demands at the inference stage. Post-training Quantization (PTQ) has emerged as a fast and data-efficient solution that can significantly reduce computation and memory footprint by using low-bit weights and activations. However, its applicability to DiTs has not yet been explored and faces non-trivial difficulties due to the unique design of DiTs. In this paper, we propose PTQ4DiT, a specifically designed PTQ method for DiTs. We discover two primary quantization challenges inherent in DiTs, notably the presence of salient channels with extreme magnitudes and the temporal variability in distributions of salient activation over multiple timesteps. To tackle these challenges, we propose Channel-wise Salience Balancing (CSB) and Spearmen's $\rho$-guided Salience Calibration (SSC). CSB leverages the complementarity property of channel magnitudes to redistribute the extremes, alleviating quantization errors for both activations and weights. SSC extends this approach by dynamically adjusting the balanced salience to capture the temporal variations in activation. Additionally, to eliminate extra computational costs caused by PTQ4DiT during inference, we design an offline re-parameterization strategy for DiTs. Experiments demonstrate that our PTQ4DiT successfully quantizes DiTs to 8-bit precision (W8A8) while preserving comparable generation ability and further enables effective quantization to 4-bit weight precision (W4A8) for the first time.
comment: NeurIPS 2024. Code is available at https://github.com/adreamwu/PTQ4DiT
♻ ☆ Learning Action and Reasoning-Centric Image Editing from Videos and Simulations NeurIPS 2024
An image editing model should be able to perform diverse edits, ranging from object replacement, changing attributes or style, to performing actions or movement, which require many forms of reasoning. Current general instruction-guided editing models have significant shortcomings with action and reasoning-centric edits. Object, attribute or stylistic changes can be learned from visually static datasets. On the other hand, high-quality data for action and reasoning-centric edits is scarce and has to come from entirely different sources that cover e.g. physical dynamics, temporality and spatial reasoning. To this end, we meticulously curate the AURORA Dataset (Action-Reasoning-Object-Attribute), a collection of high-quality training data, human-annotated and curated from videos and simulation engines. We focus on a key aspect of quality training data: triplets (source image, prompt, target image) contain a single meaningful visual change described by the prompt, i.e., truly minimal changes between source and target images. To demonstrate the value of our dataset, we evaluate an AURORA-finetuned model on a new expert-curated benchmark (AURORA-Bench) covering 8 diverse editing tasks. Our model significantly outperforms previous editing models as judged by human raters. For automatic evaluations, we find important flaws in previous metrics and caution their use for semantically hard editing tasks. Instead, we propose a new automatic metric that focuses on discriminative understanding. We hope that our efforts : (1) curating a quality training dataset and an evaluation benchmark, (2) developing critical evaluations, and (3) releasing a state-of-the-art model, will fuel further progress on general image editing.
comment: NeurIPS 2024 (Dataset & Benchmarks)
♻ ☆ Stratified Domain Adaptation: A Progressive Self-Training Approach for Scene Text Recognition
Unsupervised domain adaptation (UDA) has become increasingly prevalent in scene text recognition (STR), especially where training and testing data reside in different domains. The efficacy of existing UDA approaches tends to degrade when there is a large gap between the source and target domains. To deal with this problem, gradually shifting or progressively learning to shift from domain to domain is the key issue. In this paper, we introduce the Stratified Domain Adaptation (StrDA) approach, which examines the gradual escalation of the domain gap for the learning process. The objective is to partition the training data into subsets so that the progressively self-trained model can adapt to gradual changes. We stratify the training data by evaluating the proximity of each data sample to both the source and target domains. We propose a novel method for employing domain discriminators to estimate the out-of-distribution and domain discriminative levels of data samples. Extensive experiments on benchmark scene-text datasets show that our approach significantly improves the performance of baseline (source-trained) STR models.
comment: 15 pages, 12 figures, 5 tables, include supplementary materials
♻ ☆ Granular Privacy Control for Geolocation with Vision Language Models EMNLP 2024
Vision Language Models (VLMs) are rapidly advancing in their capability to answer information-seeking questions. As these models are widely deployed in consumer applications, they could lead to new privacy risks due to emergent abilities to identify people in photos, geolocate images, etc. As we demonstrate, somewhat surprisingly, current open-source and proprietary VLMs are very capable image geolocators, making widespread geolocation with VLMs an immediate privacy risk, rather than merely a theoretical future concern. As a first step to address this challenge, we develop a new benchmark, GPTGeoChat, to test the ability of VLMs to moderate geolocation dialogues with users. We collect a set of 1,000 image geolocation conversations between in-house annotators and GPT-4v, which are annotated with the granularity of location information revealed at each turn. Using this new dataset, we evaluate the ability of various VLMs to moderate GPT-4v geolocation conversations by determining when too much location information has been revealed. We find that custom fine-tuned models perform on par with prompted API-based models when identifying leaked location information at the country or city level; however, fine-tuning on supervised data appears to be needed to accurately moderate finer granularities, such as the name of a restaurant or building.
comment: Accepted to EMNLP 2024 main conference
♻ ☆ Learning Contrastive Feature Representations for Facial Action Unit Detection
Facial action unit (AU) detection has long encountered the challenge of detecting subtle feature differences when AUs activate. Existing methods often rely on encoding pixel-level information of AUs, which not only encodes additional redundant information but also leads to increased model complexity and limited generalizability. Additionally, the accuracy of AU detection is negatively impacted by the class imbalance issue of each AU type, and the presence of noisy and false AU labels. In this paper, we introduce a novel contrastive learning framework aimed for AU detection that incorporates both self-supervised and supervised signals, thereby enhancing the learning of discriminative features for accurate AU detection. To tackle the class imbalance issue, we employ a negative sample re-weighting strategy that adjusts the step size of updating parameters for minority and majority class samples. Moreover, to address the challenges posed by noisy and false AU labels, we employ a sampling technique that encompasses three distinct types of positive sample pairs. This enables us to inject self-supervised signals into the supervised signal, effectively mitigating the adverse effects of noisy labels. Our experimental assessments, conducted on four widely-utilized benchmark datasets (BP4D, DISFA, GFT and Aff-Wild2), underscore the superior performance of our approach compared to state-of-the-art methods of AU detection. Our code is available at \url{https://github.com/Ziqiao-Shang/AUNCE}.
comment: 35 pages, 18 figures, submitted to Pattern Recognition (PR)
♻ ☆ Context-Aware Full Body Anonymization using Text-to-Image Diffusion Models
Anonymization plays a key role in protecting sensible information of individuals in real world datasets. Self-driving cars for example need high resolution facial features to track people and their viewing direction to predict future behaviour and react accordingly. In order to protect people's privacy whilst keeping important features in the dataset, it is important to replace the full body of a person with a highly detailed anonymized one. In contrast to doing face anonymization, full body replacement decreases the ability of recognizing people by their hairstyle or clothes. In this paper, we propose a workflow for full body person anonymization utilizing Stable Diffusion as a generative backend. Text-to-image diffusion models, like Stable Diffusion, OpenAI's DALL-E or Midjourney, have become very popular in recent time, being able to create photorealistic images from a single text prompt. We show that our method outperforms state-of-the art anonymization pipelines with respect to image quality, resolution, Inception Score (IS) and Frechet Inception Distance (FID). Additionally, our method is invariant with respect to the image generator and thus able to be used with the latest models available.
♻ ☆ MuJo: Multimodal Joint Feature Space Learning for Human Activity Recognition
Human Activity Recognition (HAR) is a longstanding problem in AI with applications in a broad range of areas, including healthcare, sports and fitness, security, and more. The performance of HAR in real-world settings is strongly dependent on the type and quality of the input signal that can be acquired. Given an unobstructed, high-quality camera view of a scene, computer vision systems, in particular in conjunction with foundation models, can today fairly reliably distinguish complex activities. On the other hand, recognition using modalities such as wearable sensors (which are often more broadly available, e.g., in mobile phones and smartwatches) is a more difficult problem, as the signals often contain less information and labeled training data is more difficult to acquire. To alleviate the need for labeled data, we introduce our comprehensive Fitness Multimodal Activity Dataset (FiMAD) in this work, which can be used with the proposed pre-training method MuJo (Multimodal Joint Feature Space Learning) to enhance HAR performance across various modalities. FiMAD was created using YouTube fitness videos and contains parallel video, language, pose, and simulated IMU sensor data. MuJo utilizes this dataset to learn a joint feature space for these modalities. We show that classifiers pre-trained on FiMAD can increase the performance on real HAR datasets such as MM-Fit, MyoGym, MotionSense, and MHEALTH. For instance, on MM-Fit, we achieve an Macro F1-Score of up to 0.855 when fine-tuning on only 2% of the training data and 0.942 when utilizing the full training set for classification tasks. We have compared our approach to other self-supervised ones and showed that, unlike them, ours can consistently improve on the baseline network performance as well as provide a better data-efficiency.
♻ ☆ Automatic Mapping of Anatomical Landmarks from Free-Text Using Large Language Models: Insights from Llama-2
Anatomical landmarks are vital in medical imaging for navigation and anomaly detection. Modern large language models (LLMs), like Llama-2, offer promise for automating the mapping of these landmarks in free-text radiology reports to corresponding positions in image data. Recent studies propose LLMs may develop coherent representations of generative processes. Motivated by these insights, we investigated whether LLMs accurately represent the spatial positions of anatomical landmarks. Through experiments with Llama-2 models, we found that they can linearly represent anatomical landmarks in space with considerable robustness to different prompts. These results underscore the potential of LLMs to enhance the efficiency and accuracy of medical imaging workflows.
comment: 6 pages, 2 figures, 1 table
♻ ☆ Estimating Atmospheric Variables from Digital Typhoon Satellite Images via Conditional Denoising Diffusion Models NeurIPS 2024
This study explores the application of diffusion models in the field of typhoons, predicting multiple ERA5 meteorological variables simultaneously from Digital Typhoon satellite images. The focus of this study is taken to be Taiwan, an area very vulnerable to typhoons. By comparing the performance of Conditional Denoising Diffusion Probability Model (CDDPM) with Convolutional Neural Networks (CNN) and Squeeze-and-Excitation Networks (SENet), results suggest that the CDDPM performs best in generating accurate and realistic meteorological data. Specifically, CDDPM achieved a PSNR of 32.807, which is approximately 7.9% higher than CNN and 5.5% higher than SENet. Furthermore, CDDPM recorded an RMSE of 0.032, showing a 11.1% improvement over CNN and 8.6% improvement over SENet. A key application of this research can be for imputation purposes in missing meteorological datasets and generate additional high-quality meteorological data using satellite images. It is hoped that the results of this analysis will enable more robust and detailed forecasting, reducing the impact of severe weather events on vulnerable regions. Code accessible at https://github.com/TammyLing/Typhoon-forecasting.
comment: Accepted for spotlight presentation at the NeurIPS 2024 workshop on Tackling Climate Change with Machine Learning. 8 pages, 5 figures
♻ ☆ Rethinking Human Evaluation Protocol for Text-to-Video Models: Enhancing Reliability,Reproducibility, and Practicality
Recent text-to-video (T2V) technology advancements, as demonstrated by models such as Gen2, Pika, and Sora, have significantly broadened its applicability and popularity. Despite these strides, evaluating these models poses substantial challenges. Primarily, due to the limitations inherent in automatic metrics, manual evaluation is often considered a superior method for assessing T2V generation. However, existing manual evaluation protocols face reproducibility, reliability, and practicality issues. To address these challenges, this paper introduces the Text-to-Video Human Evaluation (T2VHE) protocol, a comprehensive and standardized protocol for T2V models. The T2VHE protocol includes well-defined metrics, thorough annotator training, and an effective dynamic evaluation module. Experimental results demonstrate that this protocol not only ensures high-quality annotations but can also reduce evaluation costs by nearly 50\%. We will open-source the entire setup of the T2VHE protocol, including the complete protocol workflow, the dynamic evaluation component details, and the annotation interface code. This will help communities establish more sophisticated human assessment protocols.
♻ ☆ MirrorCheck: Efficient Adversarial Defense for Vision-Language Models
Vision-Language Models (VLMs) are becoming increasingly vulnerable to adversarial attacks as various novel attack strategies are being proposed against these models. While existing defenses excel in unimodal contexts, they currently fall short in safeguarding VLMs against adversarial threats. To mitigate this vulnerability, we propose a novel, yet elegantly simple approach for detecting adversarial samples in VLMs. Our method leverages Text-to-Image (T2I) models to generate images based on captions produced by target VLMs. Subsequently, we calculate the similarities of the embeddings of both input and generated images in the feature space to identify adversarial samples. Empirical evaluations conducted on different datasets validate the efficacy of our approach, outperforming baseline methods adapted from image classification domains. Furthermore, we extend our methodology to classification tasks, showcasing its adaptability and model-agnostic nature. Theoretical analyses and empirical findings also show the resilience of our approach against adaptive attacks, positioning it as an excellent defense mechanism for real-world deployment against adversarial threats.
♻ ☆ Beyond Thumbs Up/Down: Untangling Challenges of Fine-Grained Feedback for Text-to-Image Generation
Human feedback plays a critical role in learning and refining reward models for text-to-image generation, but the optimal form the feedback should take for learning an accurate reward function has not been conclusively established. This paper investigates the effectiveness of fine-grained feedback which captures nuanced distinctions in image quality and prompt-alignment, compared to traditional coarse-grained feedback (for example, thumbs up/down or ranking between a set of options). While fine-grained feedback holds promise, particularly for systems catering to diverse societal preferences, we show that demonstrating its superiority to coarse-grained feedback is not automatic. Through experiments on real and synthetic preference data, we surface the complexities of building effective models due to the interplay of model choice, feedback type, and the alignment between human judgment and computational interpretation. We identify key challenges in eliciting and utilizing fine-grained feedback, prompting a reassessment of its assumed benefits and practicality. Our findings -- e.g., that fine-grained feedback can lead to worse models for a fixed budget, in some settings; however, in controlled settings with known attributes, fine grained rewards can indeed be more helpful -- call for careful consideration of feedback attributes and potentially beckon novel modeling approaches to appropriately unlock the potential value of fine-grained feedback in-the-wild.
♻ ☆ G2D: From Global to Dense Radiography Representation Learning via Vision-Language Pre-training NeurIPS2024
Recently, medical vision-language pre-training (VLP) has reached substantial progress to learn global visual representation from medical images and their paired radiology reports. However, medical imaging tasks in real world usually require finer granularity in visual features. These tasks include visual localization tasks (e.g., semantic segmentation, object detection) and visual grounding task. Yet, current medical VLP methods face challenges in learning these fine-grained features, as they primarily focus on brute-force alignment between image patches and individual text tokens for local visual feature learning, which is suboptimal for downstream dense prediction tasks. In this work, we propose a new VLP framework, named \textbf{G}lobal to \textbf{D}ense level representation learning (G2D) that achieves significantly improved granularity and more accurate grounding for the learned features, compared to existing medical VLP approaches. In particular, G2D learns dense and semantically-grounded image representations via a pseudo segmentation task parallel with the global vision-language alignment. Notably, generating pseudo segmentation targets does not incur extra trainable parameters: they are obtained on the fly during VLP with a parameter-free processor. G2D achieves superior performance across 6 medical imaging tasks and 25 diseases, particularly in semantic segmentation, which necessitates fine-grained, semantically-grounded image features. In this task, G2D surpasses peer models even when fine-tuned with just 1\% of the training data, compared to the 100\% used by these models. The code will be released upon acceptance.
comment: Accepted by NeurIPS2024
♻ ☆ t-READi: Transformer-Powered Robust and Efficient Multimodal Inference for Autonomous Driving
Given the wide adoption of multimodal sensors (e.g., camera, lidar, radar) by autonomous vehicles (AVs), deep analytics to fuse their outputs for a robust perception become imperative. However, existing fusion methods often make two assumptions rarely holding in practice: i) similar data distributions for all inputs and ii) constant availability for all sensors. Because, for example, lidars have various resolutions and failures of radars may occur, such variability often results in significant performance degradation in fusion. To this end, we present tREADi, an adaptive inference system that accommodates the variability of multimodal sensory data and thus enables robust and efficient perception. t-READi identifies variation-sensitive yet structure-specific model parameters; it then adapts only these parameters while keeping the rest intact. t-READi also leverages a cross-modality contrastive learning method to compensate for the loss from missing modalities. Both functions are implemented to maintain compatibility with existing multimodal deep fusion methods. The extensive experiments evidently demonstrate that compared with the status quo approaches, t-READi not only improves the average inference accuracy by more than 6% but also reduces the inference latency by almost 15x with the cost of only 5% extra memory overhead in the worst case under realistic data and modal variations.
comment: 14 pages, 16 figures
♻ ☆ Fine-grained Image-to-LiDAR Contrastive Distillation with Visual Foundation Models NeurIPS 2024
Contrastive image-to-LiDAR knowledge transfer, commonly used for learning 3D representations with synchronized images and point clouds, often faces a self-conflict dilemma. This issue arises as contrastive losses unintentionally dissociate features of unmatched points and pixels that share semantic labels, compromising the integrity of learned representations. To overcome this, we harness Visual Foundation Models (VFMs), which have revolutionized the acquisition of pixel-level semantics, to enhance 3D representation learning. Specifically, we utilize off-the-shelf VFMs to generate semantic labels for weakly-supervised pixel-to-point contrastive distillation. Additionally, we employ von Mises-Fisher distributions to structure the feature space, ensuring semantic embeddings within the same class remain consistent across varying inputs. Furthermore, we adapt sampling probabilities of points to address imbalances in spatial distribution and category frequency, promoting comprehensive and balanced learning. Extensive experiments demonstrate that our approach mitigates the challenges posed by traditional methods and consistently surpasses existing image-to-LiDAR contrastive distillation methods in downstream tasks. The source code is available at \href{https://github.com/Eaphan/OLIVINE.}{\color{black}https://github.com/Eaphan/OLIVINE}.
comment: Accepted to NeurIPS 2024
♻ ☆ Bias Behind the Wheel: Fairness Testing of Autonomous Driving Systems
This paper conducts fairness testing of automated pedestrian detection, a crucial but under-explored issue in autonomous driving systems. We evaluate eight state-of-the-art deep learning-based pedestrian detectors across demographic groups on large-scale real-world datasets. To enable thorough fairness testing, we provide extensive annotations for the datasets, resulting in 8,311 images with 16,070 gender labels, 20,115 age labels, and 3,513 skin tone labels. Our findings reveal significant fairness issues, particularly related to age. The proportion of undetected children is 20.14% higher compared to adults. Furthermore, we explore how various driving scenarios affect the fairness of pedestrian detectors. We find that pedestrian detectors demonstrate significant gender biases during night time, potentially exacerbating the prevalent societal issue of female safety concerns during nighttime out. Moreover, we observe that pedestrian detectors can demonstrate both enhanced fairness and superior performance under specific driving conditions, which challenges the fairness-performance trade-off theory widely acknowledged in the fairness literature. We publicly release the code, data, and results to support future research on fairness in autonomous driving.
comment: Accepted by ACM Transactions on Software Engineering and Methodology (TOSEM)
♻ ☆ AnyDesign: Versatile Area Fashion Editing via Mask-Free Diffusion
Fashion image editing aims to modify a person's appearance based on a given instruction. Existing methods require auxiliary tools like segmenters and keypoint extractors, lacking a flexible and unified framework. Moreover, these methods are limited in the variety of clothing types they can handle, as most datasets focus on people in clean backgrounds and only include generic garments such as tops, pants, and dresses. These limitations restrict their applicability in real-world scenarios. In this paper, we first extend an existing dataset for human generation to include a wider range of apparel and more complex backgrounds. This extended dataset features people wearing diverse items such as tops, pants, dresses, skirts, headwear, scarves, shoes, socks, and bags. Additionally, we propose AnyDesign, a diffusion-based method that enables mask-free editing on versatile areas. Users can simply input a human image along with a corresponding prompt in either text or image format. Our approach incorporates Fashion DiT, equipped with a Fashion-Guidance Attention (FGA) module designed to fuse explicit apparel types and CLIP-encoded apparel features. Both Qualitative and quantitative experiments demonstrate that our method delivers high-quality fashion editing and outperforms contemporary text-guided fashion editing methods.
♻ ☆ SCMM: Calibrating Cross-modal Representations for Text-Based Person Search
Text-Based Person Search (TBPS) is a crucial task that enables accurate retrieval of target individuals from large-scale galleries with only given textual caption. For cross-modal TBPS tasks, it is critical to obtain well-distributed representation in the common embedding space to reduce the inter-modal gap. Furthermore, learning detailed image-text correspondences is essential to discriminate similar targets and enable fine-grained search. To address these challenges, we present a simple yet effective method named Sew Calibration and Masked Modeling (SCMM) that calibrates cross-modal representations by learning compact and well-aligned embeddings. SCMM is distinguished by two novel losses to provide fine-grained cross-modal representations: 1) a Sew calibration loss that takes the quality of textual captions as guidance and aligns features between image and text modalities, and 2) a Masked Caption Modeling (MCM) loss that leverages a masked caption prediction task to establish detailed and generic relationships between textual and visual parts. The dual-pronged strategy refines feature alignment and enriches cross-modal correspondences, enabling the accurate distinction of similar individuals. Consequently, its streamlined dual-encoder architecture avoids complex branches and interactions and facilitates high-speed inference suitable for real-time requirements. Consequently, high-speed inference is achieved, which is essential for resource-limited applications often demanding real-time processing. Extensive experiments on three popular TBPS benchmarks demonstrate the superiority of SCMM, achieving top results with 73.81%, 74.25%, and 57.35% Rank-1 accuracy on CUHK-PEDES, ICFG-PEDES, and RSTPReID, respectively. We hope SCMM's scalable and cost-effective design will serve as a strong baseline and facilitate future research in this field.
comment: This version of manuscript is under IEEE TMM review
♻ ☆ Lost in Tracking: Uncertainty-guided Cardiac Cine MRI Segmentation at Right Ventricle Base
Accurate biventricular segmentation of cardiac magnetic resonance (CMR) cine images is essential for the clinical evaluation of heart function. However, compared to left ventricle (LV), right ventricle (RV) segmentation is still more challenging and less reproducible. Degenerate performance frequently occurs at the RV base, where the in-plane anatomical structures are complex (with atria, valve, and aorta) and vary due to the strong interplanar motion. In this work, we propose to address the currently unsolved issues in CMR segmentation, specifically at the RV base, with two strategies: first, we complemented the public resource by reannotating the RV base in the ACDC dataset, with refined delineation of the right ventricle outflow tract (RVOT), under the guidance of an expert cardiologist. Second, we proposed a novel dual encoder U-Net architecture that leverages temporal incoherence to inform the segmentation when interplanar motions occur. The inter-planar motion is characterized by loss-of-tracking, via Bayesian uncertainty of a motion-tracking model. Our experiments showed that our method significantly improved RV base segmentation taking into account temporal incoherence. Furthermore, we investigated the reproducibility of deep learning-based segmentation and showed that the combination of consistent annotation and loss of tracking could enhance the reproducibility of RV segmentation, potentially facilitating a large number of clinical studies focusing on RV.
♻ ☆ Sliding Gaussian ball adaptive growth (SlingBAG): point cloud-based iterative algorithm for large-scale 3D photoacoustic imaging
Large-scale photoacoustic (PA) 3D imaging has become increasingly important for both clinical and pre-clinical applications. Limited by resource and application constrains, only sparsely-distributed transducer arrays can be applied, which necessitates advanced image reconstruction algorithms to overcome artifacts caused by using back-projection algorithm. However, high computing memory consumption of traditional iterative algorithms for large-scale 3D cases is practically unacceptable. Here, we propose a point cloud-based iterative algorithm that reduces memory consumption by several orders, wherein a 3D photoacoustic scene is modeled as a series of Gaussian-distributed spherical sources. During the iterative reconstruction process, the properties of each Gaussian source, including peak intensities, standard deviations and means are stored in form of point cloud, then continuously optimized and adaptively undergoing destroying, splitting, and duplication along the gradient direction, thus manifesting the sliding ball adaptive growth effect. This method, named the sliding Gaussian ball adaptive growth (SlingBAG) algorithm, enables high-quality 3D large-scale PA reconstruction with fast iteration and extremely less memory usage. We validated SlingBAG algorithm in both simulation study and in vivo animal experiments.
comment: Added SlingBAG reconstruction of rat kidney and rat liver results; updated methods; added references
♻ ☆ Steerable Conditional Diffusion for Out-of-Distribution Adaptation in Medical Image Reconstruction
Denoising diffusion models have emerged as the go-to generative framework for solving inverse problems in imaging. A critical concern regarding these models is their performance on out-of-distribution tasks, which remains an under-explored challenge. Using a diffusion model on an out-of-distribution dataset, realistic reconstructions can be generated, but with hallucinating image features that are uniquely present in the training dataset. To address this discrepancy during train-test time and improve reconstruction accuracy, we introduce a novel sampling framework called Steerable Conditional Diffusion. Specifically, this framework adapts the diffusion model, concurrently with image reconstruction, based solely on the information provided by the available measurement. Utilising our proposed method, we achieve substantial enhancements in out-of-distribution performance across diverse imaging modalities, advancing the robust deployment of denoising diffusion models in real-world applications.
♻ ☆ ASTM :Autonomous Smart Traffic Management System Using Artificial Intelligence CNN and LSTM
In the modern world, the development of Artificial Intelligence (AI) has contributed to improvements in various areas, including automation, computer vision, fraud detection, and more. AI can be leveraged to enhance the efficiency of Autonomous Smart Traffic Management (ASTM) systems and reduce traffic congestion rates. This paper presents an Autonomous Smart Traffic Management (STM) system that uses AI to improve traffic flow rates. The system employs the YOLO V5 Convolutional Neural Network to detect vehicles in traffic management images. Additionally, it predicts the number of vehicles for the next 12 hours using a Recurrent Neural Network with Long Short-Term Memory (RNN-LSTM). The Smart Traffic Management Cycle Length Analysis manages the traffic cycle length based on these vehicle predictions, aided by AI. From the results of the RNN-LSTM model for predicting vehicle numbers over the next 12 hours, we observe that the model predicts traffic with a Mean Squared Error (MSE) of 4.521 vehicles and a Root Mean Squared Error (RMSE) of 2.232 vehicles. After simulating the STM system in the CARLA simulation environment, we found that the Traffic Management Congestion Flow Rate with ASTM (21 vehicles per minute) is 50\% higher than the rate without STM (around 15 vehicles per minute). Additionally, the Traffic Management Vehicle Pass Delay with STM (5 seconds per vehicle) is 70\% lower than without STM (around 12 seconds per vehicle). These results demonstrate that the STM system using AI can increase traffic flow by 50\% and reduce vehicle pass delays by 70\%.
comment: In process to IEEE Intelligent Vehicle Symposium 2025
♻ ☆ LinFusion: 1 GPU, 1 Minute, 16K Image
Modern diffusion models, particularly those utilizing a Transformer-based UNet for denoising, rely heavily on self-attention operations to manage complex spatial relationships, thus achieving impressive generation performance. However, this existing paradigm faces significant challenges in generating high-resolution visual content due to its quadratic time and memory complexity with respect to the number of spatial tokens. To address this limitation, we aim at a novel linear attention mechanism as an alternative in this paper. Specifically, we begin our exploration from recently introduced models with linear complexity, e.g., Mamba2, RWKV6, Gated Linear Attention, etc, and identify two key features--attention normalization and non-causal inference--that enhance high-resolution visual generation performance. Building on these insights, we introduce a generalized linear attention paradigm, which serves as a low-rank approximation of a wide spectrum of popular linear token mixers. To save the training cost and better leverage pre-trained models, we initialize our models and distill the knowledge from pre-trained StableDiffusion (SD). We find that the distilled model, termed LinFusion, achieves performance on par with or superior to the original SD after only modest training, while significantly reducing time and memory complexity. Extensive experiments on SD-v1.5, SD-v2.1, and SD-XL demonstrate that LinFusion enables satisfactory and efficient zero-shot cross-resolution generation, accommodating ultra-resolution images like 16K on a single GPU. Moreover, it is highly compatible with pre-trained SD components and pipelines, such as ControlNet, IP-Adapter, DemoFusion, DistriFusion, etc, requiring no adaptation efforts. Codes are available at https://github.com/Huage001/LinFusion.
comment: Work in Progress. Codes are available at https://github.com/Huage001/LinFusion
♻ ☆ Synthetic Augmentation for Anatomical Landmark Localization using DDPMs MICCAI 2024
Deep learning techniques for anatomical landmark localization (ALL) have shown great success, but their reliance on large annotated datasets remains a problem due to the tedious and costly nature of medical data acquisition and annotation. While traditional data augmentation, variational autoencoders (VAEs), and generative adversarial networks (GANs) have already been used to synthetically expand medical datasets, diffusion-based generative models have recently started to gain attention for their ability to generate high-quality synthetic images. In this study, we explore the use of denoising diffusion probabilistic models (DDPMs) for generating medical images and their corresponding heatmaps of landmarks to enhance the training of a supervised deep learning model for ALL. Our novel approach involves a DDPM with a 2-channel input, incorporating both the original medical image and its heatmap of annotated landmarks. We also propose a novel way to assess the quality of the generated images using a Markov Random Field (MRF) model for landmark matching and a Statistical Shape Model (SSM) to check landmark plausibility, before we evaluate the DDPM-augmented dataset in the context of an ALL task involving hand X-Rays.
comment: Accepted for the SASHIMI workshop of MICCAI 2024
♻ ☆ PixLore: A Dataset-driven Approach to Rich Image Captioning
In the domain of vision-language integration, generating detailed image captions poses a significant challenge due to the lack of curated and rich datasets. This study introduces PixLore, a novel method that leverages Querying Transformers through the fine-tuning of the BLIP-2 model using the LoRa method on a standard commercial GPU. The followed approach, which involves training on a carefully assembled dataset from state-of-the-art Computer Vision models combined and augmented by ChatGPT, addresses the question of whether intricate image understanding can be achieved with an ensemble of smaller-scale models, referred to as Knowledge Stitching. Comparative evaluations against major models such as GPT-4 and Google Bard demonstrate that PixLore-2.7B, despite having considerably fewer parameters, is rated higher than the existing State-of-the-Art models in over half of the assessments. Precisely, PixLore outperform Bard and BLIP-2, which score approximately 35.18% and 27.98% lower than PixLore in the task of image captioning. This research not only presents a groundbreaking approach but also highlights the importance of well-curated datasets in enhancing the performance of smaller models.
comment: Paper in preprint pending of publication
♻ ☆ SCA: Highly Efficient Semantic-Consistent Unrestricted Adversarial Attack
Deep neural network based systems deployed in sensitive environments are vulnerable to adversarial attacks. Unrestricted adversarial attacks typically manipulate the semantic content of an image (e.g., color or texture) to create adversarial examples that are both effective and photorealistic. Recent works have utilized the diffusion inversion process to map images into a latent space, where high-level semantics are manipulated by introducing perturbations. However, they often results in substantial semantic distortions in the denoised output and suffers from low efficiency. In this study, we propose a novel framework called Semantic-Consistent Unrestricted Adversarial Attacks (SCA), which employs an inversion method to extract edit-friendly noise maps and utilizes Multimodal Large Language Model (MLLM) to provide semantic guidance throughout the process. Under the condition of rich semantic information provided by MLLM, we perform the DDPM denoising process of each step using a series of edit-friendly noise maps, and leverage DPM Solver++ to accelerate this process, enabling efficient sampling with semantic consistency. Compared to existing methods, our framework enables the efficient generation of adversarial examples that exhibit minimal discernible semantic changes. Consequently, we for the first time introduce Semantic-Consistent Adversarial Examples (SCAE). Extensive experiments and visualizations have demonstrated the high efficiency of SCA, particularly in being on average 12 times faster than the state-of-the-art attacks. Our research can further draw attention to the security of multimedia information.
♻ ☆ SafeGen: Mitigating Sexually Explicit Content Generation in Text-to-Image Models CCS 2024
Text-to-image (T2I) models, such as Stable Diffusion, have exhibited remarkable performance in generating high-quality images from text descriptions in recent years. However, text-to-image models may be tricked into generating not-safe-for-work (NSFW) content, particularly in sexually explicit scenarios. Existing countermeasures mostly focus on filtering inappropriate inputs and outputs, or suppressing improper text embeddings, which can block sexually explicit content (e.g., naked) but may still be vulnerable to adversarial prompts -- inputs that appear innocent but are ill-intended. In this paper, we present SafeGen, a framework to mitigate sexual content generation by text-to-image models in a text-agnostic manner. The key idea is to eliminate explicit visual representations from the model regardless of the text input. In this way, the text-to-image model is resistant to adversarial prompts since such unsafe visual representations are obstructed from within. Extensive experiments conducted on four datasets and large-scale user studies demonstrate SafeGen's effectiveness in mitigating sexually explicit content generation while preserving the high-fidelity of benign images. SafeGen outperforms eight state-of-the-art baseline methods and achieves 99.4% sexual content removal performance. Furthermore, our constructed benchmark of adversarial prompts provides a basis for future development and evaluation of anti-NSFW-generation methods.
comment: Accepted by ACM CCS 2024. Please cite this paper as "Xinfeng Li, Yuchen Yang, Jiangyi Deng, Chen Yan, Yanjiao Chen, Xiaoyu Ji, Wenyuan Xu. SafeGen: Mitigating Sexually Explicit Content Generation in Text-to-Image Models. In Proceedings of ACM Conference on Computer and Communications Security (CCS), 2024."
♻ ☆ SAM-Guided Masked Token Prediction for 3D Scene Understanding NeurIPS 2024
Foundation models have significantly enhanced 2D task performance, and recent works like Bridge3D have successfully applied these models to improve 3D scene understanding through knowledge distillation, marking considerable advancements. Nonetheless, challenges such as the misalignment between 2D and 3D representations and the persistent long-tail distribution in 3D datasets still restrict the effectiveness of knowledge distillation from 2D to 3D using foundation models. To tackle these issues, we introduce a novel SAM-guided tokenization method that seamlessly aligns 3D transformer structures with region-level knowledge distillation, replacing the traditional KNN-based tokenization techniques. Additionally, we implement a group-balanced re-weighting strategy to effectively address the long-tail problem in knowledge distillation. Furthermore, inspired by the recent success of masked feature prediction, our framework incorporates a two-stage masked token prediction process in which the student model predicts both the global embeddings and the token-wise local embeddings derived from the teacher models trained in the first stage. Our methodology has been validated across multiple datasets, including SUN RGB-D, ScanNet, and S3DIS, for tasks like 3D object detection and semantic segmentation. The results demonstrate significant improvements over current State-of-the-art self-supervised methods, establishing new benchmarks in this field.
comment: Accepted by NeurIPS 2024
♻ ☆ GeoCalib: Learning Single-image Calibration with Geometric Optimization ECCV 2024
From a single image, visual cues can help deduce intrinsic and extrinsic camera parameters like the focal length and the gravity direction. This single-image calibration can benefit various downstream applications like image editing and 3D mapping. Current approaches to this problem are based on either classical geometry with lines and vanishing points or on deep neural networks trained end-to-end. The learned approaches are more robust but struggle to generalize to new environments and are less accurate than their classical counterparts. We hypothesize that they lack the constraints that 3D geometry provides. In this work, we introduce GeoCalib, a deep neural network that leverages universal rules of 3D geometry through an optimization process. GeoCalib is trained end-to-end to estimate camera parameters and learns to find useful visual cues from the data. Experiments on various benchmarks show that GeoCalib is more robust and more accurate than existing classical and learned approaches. Its internal optimization estimates uncertainties, which help flag failure cases and benefit downstream applications like visual localization. The code and trained models are publicly available at https://github.com/cvg/GeoCalib.
comment: Presented at ECCV 2024
♻ ☆ Octree-GS: Towards Consistent Real-time Rendering with LOD-Structured 3D Gaussians
The recent 3D Gaussian splatting (3D-GS) has shown remarkable rendering fidelity and efficiency compared to NeRF-based neural scene representations. While demonstrating the potential for real-time rendering, 3D-GS encounters rendering bottlenecks in large scenes with complex details due to an excessive number of Gaussian primitives located within the viewing frustum. This limitation is particularly noticeable in zoom-out views and can lead to inconsistent rendering speeds in scenes with varying details. Moreover, it often struggles to capture the corresponding level of details at different scales with its heuristic density control operation. Inspired by the Level-of-Detail (LOD) techniques, we introduce Octree-GS, featuring an LOD-structured 3D Gaussian approach supporting level-of-detail decomposition for scene representation that contributes to the final rendering results. Our model dynamically selects the appropriate level from the set of multi-resolution anchor points, ensuring consistent rendering performance with adaptive LOD adjustments while maintaining high-fidelity rendering results.
comment: Project page: https://city-super.github.io/octree-gs/
♻ ☆ OpenDAS: Open-Vocabulary Domain Adaptation for Segmentation
Recently, Vision-Language Models (VLMs) have advanced segmentation techniques by shifting from the traditional segmentation of a closed-set of predefined object classes to open-vocabulary segmentation (OVS), allowing users to segment novel classes and concepts unseen during training of the segmentation model. However, this flexibility comes with a trade-off: fully-supervised closed-set methods still outperform OVS methods on base classes, that is on classes on which they have been explicitly trained. This is due to the lack of pixel-aligned training masks for VLMs (which are trained on image-caption pairs), and the absence of domain-specific knowledge, such as autonomous driving. Therefore, we propose the task of open-vocabulary domain adaptation to infuse domain-specific knowledge into VLMs while preserving their open-vocabulary nature. By doing so, we achieve improved performance in base and novel classes. Existing VLM adaptation methods improve performance on base (training) queries, but fail to fully preserve the open-set capabilities of VLMs on novel queries. To address this shortcoming, we combine parameter-efficient prompt tuning with a triplet-loss-based training strategy that uses auxiliary negative queries. Notably, our approach is the only parameter-efficient method that consistently surpasses the original VLM on novel classes. Our adapted VLMs can seamlessly be integrated into existing OVS pipelines, e.g., improving OVSeg by +6.0% mIoU on ADE20K for open-vocabulary 2D segmentation, and OpenMask3D by +4.1% AP on ScanNet++ Offices for open-vocabulary 3D instance segmentation without other changes.
♻ ☆ Tri-Cam: Practical Eye Gaze Tracking via Camera Network
As human eyes serve as conduits of rich information, unveiling emotions, intentions, and even aspects of an individual's health and overall well-being, gaze tracking also enables various human-computer interaction applications, as well as insights in psychological and medical research. However, existing gaze tracking solutions fall short at handling free user movement, and also require laborious user effort in system calibration. We introduce Tri-Cam, a practical deep learning-based gaze tracking system using three affordable RGB webcams. It features a split network structure for efficient training, as well as designated network designs to handle the separated gaze tracking tasks. Tri-Cam is also equipped with an implicit calibration module, which makes use of mouse click opportunities to reduce calibration overhead on the user's end. We evaluate Tri-Cam against Tobii, the state-of-the-art commercial eye tracker, achieving comparable accuracy, while supporting a wider free movement area. In conclusion, Tri-Cam provides a user-friendly, affordable, and robust gaze tracking solution that could practically enable various applications.
comment: 12 pages
♻ ☆ See Where You Read with Eye Gaze Tracking and Large Language Model
Losing track of reading progress during line switching can be frustrating. Eye gaze tracking technology offers a potential solution by highlighting read paragraphs, aiding users in avoiding wrong line switches. However, the gap between gaze tracking accuracy (2-3 cm) and text line spacing (3-5 mm) makes direct application impractical. Existing methods leverage the linear reading pattern but fail during jump reading. This paper presents a reading tracking and highlighting system that supports both linear and jump reading. Based on experimental insights from the gaze nature study of 16 users, two gaze error models are designed to enable both jump reading detection and relocation. The system further leverages the large language model's contextual perception capability in aiding reading tracking. A reading tracking domain-specific line-gaze alignment opportunity is also exploited to enable dynamic and frequent calibration of the gaze results. Controlled experiments demonstrate reliable linear reading tracking, as well as 84% accuracy in tracking jump reading. Furthermore, real field tests with 18 volunteers demonstrated the system's effectiveness in tracking and highlighting read paragraphs, improving reading efficiency, and enhancing user experience.
comment: 9 pages
♻ ☆ Tables as Texts or Images: Evaluating the Table Reasoning Ability of LLMs and MLLMs ACL 2024
In this paper, we investigate the effectiveness of various LLMs in interpreting tabular data through different prompting strategies and data formats. Our analyses extend across six benchmarks for table-related tasks such as question-answering and fact-checking. We introduce for the first time the assessment of LLMs' performance on image-based table representations. Specifically, we compare five text-based and three image-based table representations, demonstrating the role of representation and prompting on LLM performance. Our study provides insights into the effective use of LLMs on table-related tasks.
comment: Accepted to ACL 2024 Findings; Naihao and Zhenjie contributed equally to the project; Data available at: https://github.com/dnaihao/Tables-as-Texts-or-Images
♻ ☆ SurrogatePrompt: Bypassing the Safety Filter of Text-to-Image Models via Substitution CCS
Advanced text-to-image models such as DALL$\cdot$E 2 and Midjourney possess the capacity to generate highly realistic images, raising significant concerns regarding the potential proliferation of unsafe content. This includes adult, violent, or deceptive imagery of political figures. Despite claims of rigorous safety mechanisms implemented in these models to restrict the generation of not-safe-for-work (NSFW) content, we successfully devise and exhibit the first prompt attacks on Midjourney, resulting in the production of abundant photorealistic NSFW images. We reveal the fundamental principles of such prompt attacks and suggest strategically substituting high-risk sections within a suspect prompt to evade closed-source safety measures. Our novel framework, SurrogatePrompt, systematically generates attack prompts, utilizing large language models, image-to-text, and image-to-image modules to automate attack prompt creation at scale. Evaluation results disclose an 88% success rate in bypassing Midjourney's proprietary safety filter with our attack prompts, leading to the generation of counterfeit images depicting political figures in violent scenarios. Both subjective and objective assessments validate that the images generated from our attack prompts present considerable safety hazards.
comment: To appear in the the 31st ACM Conference on Computer and Communications Security (CCS)
♻ ☆ GeoReasoner: Geo-localization with Reasoning in Street Views using a Large Vision-Language Model ICML 2024
This work tackles the problem of geo-localization with a new paradigm using a large vision-language model (LVLM) augmented with human inference knowledge. A primary challenge here is the scarcity of data for training the LVLM - existing street-view datasets often contain numerous low-quality images lacking visual clues, and lack any reasoning inference. To address the data-quality issue, we devise a CLIP-based network to quantify the degree of street-view images being locatable, leading to the creation of a new dataset comprising highly locatable street views. To enhance reasoning inference, we integrate external knowledge obtained from real geo-localization games, tapping into valuable human inference capabilities. The data are utilized to train GeoReasoner, which undergoes fine-tuning through dedicated reasoning and location-tuning stages. Qualitative and quantitative evaluations illustrate that GeoReasoner outperforms counterpart LVLMs by more than 25% at country-level and 38% at city-level geo-localization tasks, and surpasses StreetCLIP performance while requiring fewer training resources. The data and code are available at https://github.com/lingli1996/GeoReasoner.
comment: ICML 2024
♻ ☆ Spiking GS: Towards High-Accuracy and Low-Cost Surface Reconstruction via Spiking Neuron-based Gaussian Splatting
3D Gaussian Splatting is capable of reconstructing 3D scenes in minutes. Despite recent advances in improving surface reconstruction accuracy, the reconstructed results still exhibit bias and suffer from inefficiency in storage and training. This paper provides a different observation on the cause of the inefficiency and the reconstruction bias, which is attributed to the integration of the low-opacity parts (LOPs) of the generated Gaussians. We show that LOPs consist of Gaussians with overall low-opacity (LOGs) and the low-opacity tails (LOTs) of Gaussians. We propose Spiking GS to reduce such two types of LOPs by integrating spiking neurons into the Gaussian Splatting pipeline. Specifically, we introduce global and local full-precision integrate-and-fire spiking neurons to the opacity and representation function of flattened 3D Gaussians, respectively. Furthermore, we enhance the density control strategy with spiking neurons' thresholds and a new criterion on the scale of Gaussians. Our method can represent more accurate reconstructed surfaces at a lower cost. The supplementary material and code are available at https://github.com/zju-bmi-lab/SpikingGS.
♻ ☆ Cefdet: Cognitive Effectiveness Network Based on Fuzzy Inference for Action Detection ACM MM
Action detection and understanding provide the foundation for the generation and interaction of multimedia content. However, existing methods mainly focus on constructing complex relational inference networks, overlooking the judgment of detection effectiveness. Moreover, these methods frequently generate detection results with cognitive abnormalities. To solve the above problems, this study proposes a cognitive effectiveness network based on fuzzy inference (Cefdet), which introduces the concept of "cognition-based detection" to simulate human cognition. First, a fuzzy-driven cognitive effectiveness evaluation module (FCM) is established to introduce fuzzy inference into action detection. FCM is combined with human action features to simulate the cognition-based detection process, which clearly locates the position of frames with cognitive abnormalities. Then, a fuzzy cognitive update strategy (FCS) is proposed based on the FCM, which utilizes fuzzy logic to re-detect the cognition-based detection results and effectively update the results with cognitive abnormalities. Experimental results demonstrate that Cefdet exhibits superior performance against several mainstream algorithms on the public datasets, validating its effectiveness and superiority. Code is available at https://github.com/12sakura/Cefdet.
comment: The paper has been accepted by ACM MM. If you find this work helpful, please consider citing our paper. Zhe Luo, Weina Fu, Shuai Liu, Saeed Anwar, Muhammad Saqib, Sambit Bakshi, Khan Muhammad (2024) Cefdet: Cognitive Effectiveness Network Based on Fuzzy Inference for Action Detection, 32nd ACM International Conference on Multimedia, online first, 10.1145/3664647.3681226
♻ ☆ Model Supply Chain Poisoning: Backdooring Pre-trained Models via Embedding Indistinguishability
Pre-trained models (PTMs) are widely adopted across various downstream tasks in the machine learning supply chain. Adopting untrustworthy PTMs introduces significant security risks, where adversaries can poison the model supply chain by embedding hidden malicious behaviors (backdoors) into PTMs. However, existing backdoor attacks to PTMs can only achieve partially task-agnostic and the embedded backdoors are easily erased during the fine-tuning process. This makes it challenging for the backdoors to persist and propagate through the supply chain. In this paper, we propose a novel and severer backdoor attack, TransTroj, which enables the backdoors embedded in PTMs to efficiently transfer in the model supply chain. In particular, we first formalize this attack as an indistinguishability problem between poisoned and clean samples in the embedding space. We decompose embedding indistinguishability into pre- and post-indistinguishability, representing the similarity of the poisoned and reference embeddings before and after the attack. Then, we propose a two-stage optimization that separately optimizes triggers and victim PTMs to achieve embedding indistinguishability. We evaluate TransTroj on four PTMs and six downstream tasks. Experimental results show that our method significantly outperforms SOTA task-agnostic backdoor attacks -- achieving nearly 100\% attack success rate on most downstream tasks -- and demonstrates robustness under various system settings. Our findings underscore the urgent need to secure the model supply chain against such transferable backdoor attacks. The code is available at https://github.com/haowang-cqu/TransTroj .
♻ ☆ SeeClear: Semantic Distillation Enhances Pixel Condensation for Video Super-Resolution NeurIPS 2024
Diffusion-based Video Super-Resolution (VSR) is renowned for generating perceptually realistic videos, yet it grapples with maintaining detail consistency across frames due to stochastic fluctuations. The traditional approach of pixel-level alignment is ineffective for diffusion-processed frames because of iterative disruptions. To overcome this, we introduce SeeClear--a novel VSR framework leveraging conditional video generation, orchestrated by instance-centric and channel-wise semantic controls. This framework integrates a Semantic Distiller and a Pixel Condenser, which synergize to extract and upscale semantic details from low-resolution frames. The Instance-Centric Alignment Module (InCAM) utilizes video-clip-wise tokens to dynamically relate pixels within and across frames, enhancing coherency. Additionally, the Channel-wise Texture Aggregation Memory (CaTeGory) infuses extrinsic knowledge, capitalizing on long-standing semantic textures. Our method also innovates the blurring diffusion process with the ResShift mechanism, finely balancing between sharpness and diffusion effects. Comprehensive experiments confirm our framework's advantage over state-of-the-art diffusion-based VSR techniques. The code is available: https://github.com/Tang1705/SeeClear-NeurIPS24.
comment: Accepted to NeurIPS 2024
♻ ☆ A Diffusion-based Xray2MRI Model: Generating Pseudo-MRI Volumes From one Single X-ray
Knee osteoarthritis (KOA) is a prevalent musculoskeletal disorder, and X-rays are commonly used for its diagnosis due to their cost-effectiveness. Magnetic Resonance Imaging (MRI), on the other hand, offers detailed soft tissue visualization and has become a valuable supplementary diagnostic tool for KOA. Unfortunately, the high cost and limited accessibility of MRI hinders its widespread use, leaving many patients with KOA to rely solely on X-ray imaging. In this study, we introduce a novel diffusion-based Xray2MRI model capable of generating pseudo-MRI volumes from a single X-ray image. In addition to using X-rays as conditional input, our model integrates target depth, KOA probability distribution, and image intensity distribution modules to guide the synthesis process, ensuring that the generated corresponding slices accurately correspond to the anatomical structures. Experimental results demonstrate that by integrating information from X-rays with additional input data, our proposed approach is capable of generating pseudo-MRI sequences that approximate real MRI scans. In addition, by increasing the number of inference steps, the model achieves effective interpolation, which further improves the continuity and smoothness of the generated MRI sequences, representing a promising first attempt at cost-effective medical imaging solutions. This study is available on https://zwang78.github.io/.
♻ ☆ UAV3D: A Large-scale 3D Perception Benchmark for Unmanned Aerial Vehicles NeurIPS 2024
Unmanned Aerial Vehicles (UAVs), equipped with cameras, are employed in numerous applications, including aerial photography, surveillance, and agriculture. In these applications, robust object detection and tracking are essential for the effective deployment of UAVs. However, existing benchmarks for UAV applications are mainly designed for traditional 2D perception tasks, restricting the development of real-world applications that require a 3D understanding of the environment. Furthermore, despite recent advancements in single-UAV perception, limited views of a single UAV platform significantly constrain its perception capabilities over long distances or in occluded areas. To address these challenges, we introduce UAV3D, a benchmark designed to advance research in both 3D and collaborative 3D perception tasks with UAVs. UAV3D comprises 1,000 scenes, each of which has 20 frames with fully annotated 3D bounding boxes on vehicles. We provide the benchmark for four 3D perception tasks: single-UAV 3D object detection, single-UAV object tracking, collaborative-UAV 3D object detection, and collaborative-UAV object tracking. Our dataset and code are available at https://huiyegit.github.io/UAV3D_Benchmark/.
comment: Accepted at NeurIPS 2024
♻ ☆ RULE: Reliable Multimodal RAG for Factuality in Medical Vision Language Models EMNLP 2024
The recent emergence of Medical Large Vision Language Models (Med-LVLMs) has enhanced medical diagnosis. However, current Med-LVLMs frequently encounter factual issues, often generating responses that do not align with established medical facts. Retrieval-Augmented Generation (RAG), which utilizes external knowledge, can improve the factual accuracy of these models but introduces two major challenges. First, limited retrieved contexts might not cover all necessary information, while excessive retrieval can introduce irrelevant and inaccurate references, interfering with the model's generation. Second, in cases where the model originally responds correctly, applying RAG can lead to an over-reliance on retrieved contexts, resulting in incorrect answers. To address these issues, we propose RULE, which consists of two components. First, we introduce a provably effective strategy for controlling factuality risk through the calibrated selection of the number of retrieved contexts. Second, based on samples where over-reliance on retrieved contexts led to errors, we curate a preference dataset to fine-tune the model, balancing its dependence on inherent knowledge and retrieved contexts for generation. We demonstrate the effectiveness of RULE on medical VQA and report generation tasks across three datasets, achieving an average improvement of 47.4% in factual accuracy. We publicly release our benchmark and code in https://github.com/richard-peng-xia/RULE.
comment: EMNLP 2024 main
♻ ☆ InstructAny2Pix: Flexible Visual Editing via Multimodal Instruction Following
The ability to provide fine-grained control for generating and editing visual imagery has profound implications for computer vision and its applications. Previous works have explored extending controllability in two directions: instruction tuning with text-based prompts and multi-modal conditioning. However, these works make one or more unnatural assumptions on the number and/or type of modality inputs used to express controllability. We propose InstructAny2Pix, a flexible multi-modal instruction-following system that enables users to edit an input image using instructions involving audio, images, and text. InstructAny2Pix consists of three building blocks that facilitate this capability: a multi-modal encoder that encodes different modalities such as images and audio into a unified latent space, a diffusion model that learns to decode representations in this latent space into images, and a multi-modal LLM that can understand instructions involving multiple images and audio pieces and generate a conditional embedding of the desired output, which can be used by the diffusion decoder. Additionally, to facilitate training efficiency and improve generation quality, we include an additional refinement prior module that enhances the visual quality of LLM outputs. These designs are critical to the performance of our system. We demonstrate that our system can perform a series of novel instruction-guided editing tasks. The code is available at https://github.com/jacklishufan/InstructAny2Pix.git
comment: 25 pages, 19 figures
♻ ☆ From Redundancy to Relevance: Information Flow in LVLMs Across Reasoning Tasks
Large Vision Language Models (LVLMs) achieve great performance on visual-language reasoning tasks, however, the black-box nature of LVLMs hinders in-depth research on the reasoning mechanism. As all images need to be converted into image tokens to fit the input format of large language models (LLMs) along with natural language prompts, sequential visual representation is essential to the performance of LVLMs, and the information flow analysis approach can be an effective tool for determining interactions between these representations. In this paper, we propose integrating attention analysis with LLaVA-CAM, concretely, attention scores highlight relevant regions during forward propagation, while LLaVA-CAM captures gradient changes through backward propagation, revealing key image features. By exploring the information flow from the perspective of visual representation contribution, we observe that it tends to converge in shallow layers but diversify in deeper layers. To validate our analysis, we conduct comprehensive experiments with truncation strategies across various LVLMs for visual question answering and image captioning tasks, and experimental results not only verify our hypothesis but also reveal a consistent pattern of information flow convergence in the corresponding layers, and the information flow cliff layer will be different due to different contexts. The paper's source code can be accessed from \url{https://github.com/zhangbaijin/From-Redundancy-to-Relevance}
♻ ☆ D-Net: Dynamic Large Kernel with Dynamic Feature Fusion for Volumetric Medical Image Segmentation
Hierarchical transformers have achieved significant success in medical image segmentation due to their large receptive field and capabilities of effectively leveraging global long-range contextual information. Convolutional neural networks (CNNs) can also deliver a large receptive field by using large kernels, enabling them to achieve competitive performance with fewer model parameters. However, CNNs incorporated with large convolutional kernels remain constrained in adaptively capturing multi-scale features from organs with large variations in shape and size due to the employment of fixed-sized kernels. Additionally, they are unable to utilize global contextual information efficiently. To address these limitations, we propose Dynamic Large Kernel (DLK) and Dynamic Feature Fusion (DFF) modules. The DLK module employs multiple large kernels with varying kernel sizes and dilation rates to capture multi-scale features. Subsequently, a dynamic selection mechanism is utilized to adaptively highlight the most important spatial features based on global information. Additionally, the DFF module is proposed to adaptively fuse multi-scale local feature maps based on their global information. We integrate DLK and DFF in a hierarchical transformer architecture to develop a novel architecture, termed D-Net. D-Net is able to effectively utilize a multi-scale large receptive field and adaptively harness global contextual information. Extensive experimental results demonstrate that D-Net outperforms other state-of-the-art models in the two volumetric segmentation tasks, including abdominal multi-organ segmentation and multi-modality brain tumor segmentation. Our code is available at https://github.com/sotiraslab/DLK.
comment: 18 pages, 8 figures, 9 tables
♻ ☆ Degraded Polygons Raise Fundamental Questions of Neural Network Perception NeurIPS 2023
It is well-known that modern computer vision systems often exhibit behaviors misaligned with those of humans: from adversarial attacks to image corruptions, deep learning vision models suffer in a variety of settings that humans capably handle. In light of these phenomena, here we introduce another, orthogonal perspective studying the human-machine vision gap. We revisit the task of recovering images under degradation, first introduced over 30 years ago in the Recognition-by-Components theory of human vision. Specifically, we study the performance and behavior of neural networks on the seemingly simple task of classifying regular polygons at varying orders of degradation along their perimeters. To this end, we implement the Automated Shape Recoverability Test for rapidly generating large-scale datasets of perimeter-degraded regular polygons, modernizing the historically manual creation of image recoverability experiments. We then investigate the capacity of neural networks to recognize and recover such degraded shapes when initialized with different priors. Ultimately, we find that neural networks' behavior on this simple task conflicts with human behavior, raising a fundamental question of the robustness and learning capabilities of modern computer vision models.
comment: Accepted as a conference paper to NeurIPS 2023 (Datasets & Benchmarks Track)
♻ ☆ Depth-supervised NeRF: Fewer Views and Faster Training for Free DSN
A commonly observed failure mode of Neural Radiance Field (NeRF) is fitting incorrect geometries when given an insufficient number of input views. One potential reason is that standard volumetric rendering does not enforce the constraint that most of a scene's geometry consist of empty space and opaque surfaces. We formalize the above assumption through DS-NeRF (Depth-supervised Neural Radiance Fields), a loss for learning radiance fields that takes advantage of readily-available depth supervision. We leverage the fact that current NeRF pipelines require images with known camera poses that are typically estimated by running structure-from-motion (SFM). Crucially, SFM also produces sparse 3D points that can be used as "free" depth supervision during training: we add a loss to encourage the distribution of a ray's terminating depth matches a given 3D keypoint, incorporating depth uncertainty. DS-NeRF can render better images given fewer training views while training 2-3x faster. Further, we show that our loss is compatible with other recently proposed NeRF methods, demonstrating that depth is a cheap and easily digestible supervisory signal. And finally, we find that DS-NeRF can support other types of depth supervision such as scanned depth sensors and RGB-D reconstruction outputs.
comment: Project page: http://www.cs.cmu.edu/~dsnerf/ GitHub: https://github.com/dunbar12138/DSNeRF
♻ ☆ Adversarial Exposure Attack on Diabetic Retinopathy Imagery Grading
Diabetic Retinopathy (DR) is a leading cause of vision loss around the world. To help diagnose it, numerous cutting-edge works have built powerful deep neural networks (DNNs) to automatically grade DR via retinal fundus images (RFIs). However, RFIs are commonly affected by camera exposure issues that may lead to incorrect grades. The mis-graded results can potentially pose high risks to an aggravation of the condition. In this paper, we study this problem from the viewpoint of adversarial attacks. We identify and introduce a novel solution to an entirely new task, termed as adversarial exposure attack, which is able to produce natural exposure images and mislead the state-of-the-art DNNs. We validate our proposed method on a real-world public DR dataset with three DNNs, e.g., ResNet50, MobileNet, and EfficientNet, demonstrating that our method achieves high image quality and success rate in transferring the attacks. Our method reveals the potential threats to DNN-based automatic DR grading and would benefit the development of exposure-robust DR grading methods in the future.
comment: 13 pages, 7 figures
♻ ☆ Preserving Cardiac Integrity: A Topology-Infused Approach to Whole Heart Segmentation
Whole heart segmentation (WHS) supports cardiovascular disease (CVD) diagnosis, disease monitoring, treatment planning, and prognosis. Deep learning has become the most widely used method for WHS applications in recent years. However, segmentation of whole-heart structures faces numerous challenges including heart shape variability during the cardiac cycle, clinical artifacts like motion and poor contrast-to-noise ratio, domain shifts in multi-center data, and the distinct modalities of CT and MRI. To address these limitations and improve segmentation quality, this paper introduces a new topology-preserving module that is integrated into deep neural networks. The implementation achieves anatomically plausible segmentation by using learned topology-preserving fields, which are based entirely on 3D convolution and are therefore very effective for 3D voxel data. We incorporate natural constraints between structures into the end-to-end training and enrich the feature representation of the neural network. The effectiveness of the proposed method is validated on an open-source medical heart dataset, specifically using the WHS++ data. The results demonstrate that the architecture performs exceptionally well, achieving a Dice coefficient of 0.939 during testing. This indicates full topology preservation for individual structures and significantly outperforms other baselines in preserving the overall scene topology.
♻ ☆ Suitability of KANs for Computer Vision: A preliminary investigation
Kolmogorov-Arnold Networks (KANs) introduce a paradigm of neural modeling that implements learnable functions on the edges of the networks, diverging from the traditional node-centric activations in neural networks. This work assesses the applicability and efficacy of KANs in visual modeling, focusing on fundamental recognition and segmentation tasks. We mainly analyze the performance and efficiency of different network architectures built using KAN concepts along with conventional building blocks of convolutional and linear layers, enabling a comparative analysis with the conventional models. Our findings are aimed at contributing to understanding the potential of KANs in computer vision, highlighting both their strengths and areas for further research. Our evaluation point toward the fact that while KAN-based architectures perform in line with the original claims, it may often be important to employ more complex functions on the network edges to retain the performance advantage of KANs on more complex visual data.
♻ ☆ Performance of a GPU- and Time-Efficient Pseudo 3D Network for Magnetic Resonance Image Super-Resolution and Motion Artifact Reduction
Shortening acquisition time and reducing motion artifacts are the most critical challenges in magnetic resonance imaging (MRI). Deep learning-based image restoration has emerged as a promising solution capable of generating high-resolution and motion-artifact-free MRI images from low-resolution images acquired with shortened acquisition times or from motion-artifact-corrupted images. To facilitate clinical integration, a time- and GPU-efficient network with reliable accuracy is essential. In this study, we adopted a unified 2D deep learning framework for pseudo-3D MRI image super-resolution reconstruction (SRR) and motion artifact reduction (MAR). The optimal down-sampling factors to optimize the acquisition time in SRR were identified. Training for MAR was performed using publicly available in vivo data, employing a novel standardized method to induce motion artifacts of varying severity in a controlled way. The accuracy of the network was evaluated through a pixel-wise uncertainty map, and performance was benchmarked against state-of-the-art methods. The results demonstrated that the down-sampling factor of 1x1x2 for x2 acceleration and 2x2x2 for x4 acceleration was optimal. For SRR, the proposed TS-RCAN outperformed the 3D networks of mDCSRN and ReCNN, with an improvement of more than 0.01 in SSIM and 1.5 dB in PSNR while reducing GPU load by up to and inference time by up to 90%. For MAR, TS-RCAN exceeded UNet's performance by up to 0.014 in SSIM and 1.48 dB in PSNR. Additionally, TS-RCAN provided uncertainty information, which can be used to estimate the quality of the reconstructed images. TS-RCAN has potential use for SRR and MAR in the clinical setting.
comment: 16 pages, 9 figures
♻ ☆ Duoduo CLIP: Efficient 3D Understanding with Multi-View Images
We introduce Duoduo CLIP, a model for 3D representation learning that learns shape encodings from multi-view images instead of point-clouds. The choice of multi-view images allows us to leverage 2D priors from off-the-shelf CLIP models to facilitate fine-tuning with 3D data. Our approach not only shows better generalization compared to existing point cloud methods, but also reduces GPU requirements and training time. In addition, the model is modified with cross-view attention to leverage information across multiple frames of the object which further boosts performance. Notably, our model is permutation invariant to the order of multi-view images while being pose-free. Compared to the current SOTA point cloud method that requires 480 A100 hours to train 1 billion model parameters we only require 57 A5000 hours and 87 million parameters. Multi-view images also provide more flexibility including being able to encode objects with a variable number of images, and performance scales when more views are used. In contrast, point cloud based methods require an entire scan or model of the object. We showcase this flexibility with benchmarks from images of real-world objects. Our model also achieves better performance in more fine-grained text to shape retrieval, demonstrating better text-and-shape alignment than point cloud based models.
♻ ☆ Estimating Body and Hand Motion in an Ego-sensed World
We present EgoAllo, a system for human motion estimation from a head-mounted device. Using only egocentric SLAM poses and images, EgoAllo guides sampling from a conditional diffusion model to estimate 3D body pose, height, and hand parameters that capture the wearer's actions in the allocentric coordinate frame of the scene. To achieve this, our key insight is in representation: we propose spatial and temporal invariance criteria for improving model performance, from which we derive a head motion conditioning parameterization that improves estimation by up to 18%. We also show how the bodies estimated by our system can improve the hands: the resulting kinematic and temporal constraints result in over 40% lower hand estimation errors compared to noisy monocular estimates. Project page: https://egoallo.github.io/
comment: v2: fixed figures for Safari, typos
♻ ☆ Efficient Anatomical Labeling of Pulmonary Tree Structures via Deep Point-Graph Representation-based Implicit Fields
Pulmonary diseases rank prominently among the principal causes of death worldwide. Curing them will require, among other things, a better understanding of the complex 3D tree-shaped structures within the pulmonary system, such as airways, arteries, and veins. Traditional approaches using high-resolution image stacks and standard CNNs on dense voxel grids face challenges in computational efficiency, limited resolution, local context, and inadequate preservation of shape topology. Our method addresses these issues by shifting from dense voxel to sparse point representation, offering better memory efficiency and global context utilization. However, the inherent sparsity in point representation can lead to a loss of crucial connectivity in tree-shaped structures. To mitigate this, we introduce graph learning on skeletonized structures, incorporating differentiable feature fusion for improved topology and long-distance context capture. Furthermore, we employ an implicit function for efficient conversion of sparse representations into dense reconstructions end-to-end. The proposed method not only delivers state-of-the-art performance in labeling accuracy, both overall and at key locations, but also enables efficient inference and the generation of closed surface shapes. Addressing data scarcity in this field, we have also curated a comprehensive dataset to validate our approach. Data and code are available at \url{https://github.com/M3DV/pulmonary-tree-labeling}.
comment: Accepted by Medical Image Analysis
♻ ☆ CYCLO: Cyclic Graph Transformer Approach to Multi-Object Relationship Modeling in Aerial Videos NeurIPS 2024
Video scene graph generation (VidSGG) has emerged as a transformative approach to capturing and interpreting the intricate relationships among objects and their temporal dynamics in video sequences. In this paper, we introduce the new AeroEye dataset that focuses on multi-object relationship modeling in aerial videos. Our AeroEye dataset features various drone scenes and includes a visually comprehensive and precise collection of predicates that capture the intricate relationships and spatial arrangements among objects. To this end, we propose the novel Cyclic Graph Transformer (CYCLO) approach that allows the model to capture both direct and long-range temporal dependencies by continuously updating the history of interactions in a circular manner. The proposed approach also allows one to handle sequences with inherent cyclical patterns and process object relationships in the correct sequential order. Therefore, it can effectively capture periodic and overlapping relationships while minimizing information loss. The extensive experiments on the AeroEye dataset demonstrate the effectiveness of the proposed CYCLO model, demonstrating its potential to perform scene understanding on drone videos. Finally, the CYCLO method consistently achieves State-of-the-Art (SOTA) results on two in-the-wild scene graph generation benchmarks, i.e., PVSG and ASPIRe.
comment: Accepted to NeurIPS 2024
Information Retrieval 37
☆ Knowledge-Aware Query Expansion with Large Language Models for Textual and Relational Retrieval
Large language models (LLMs) have been used to generate query expansions augmenting original queries for improving information search. Recent studies also explore providing LLMs with initial retrieval results to generate query expansions more grounded to document corpus. However, these methods mostly focus on enhancing textual similarities between search queries and target documents, overlooking document relations. For queries like "Find me a highly rated camera for wildlife photography compatible with my Nikon F-Mount lenses", existing methods may generate expansions that are semantically similar but structurally unrelated to user intents. To handle such semi-structured queries with both textual and relational requirements, in this paper we propose a knowledge-aware query expansion framework, augmenting LLMs with structured document relations from knowledge graph (KG). To further address the limitation of entity-based scoring in existing KG-based methods, we leverage document texts as rich KG node representations and use document-based relation filtering for our Knowledge-Aware Retrieval (KAR). Extensive experiments on three datasets of diverse domains show the advantages of our method compared against state-of-the-art baselines on textual and relational semi-structured retrieval.
☆ Disjointness Violations in Wikidata
Disjointness checks are among the most important constraint checks in a knowledge base and can be used to help detect and correct incorrect statements and internal contradictions. Wikidata is a very large, community-managed knowledge base. Because of both its size and construction, Wikidata contains many incorrect statements and internal contradictions. We analyze the current modeling of disjointness on Wikidata, identify patterns that cause these disjointness violations and categorize them. We use SPARQL queries to identify each ``culprit'' causing a disjointness violation and lay out formulas to identify and fix conflicting information. We finally discuss how disjointness information could be better modeled and expanded in Wikidata in the future.
comment: Sixth International Knowledge Graph and Semantic Web Conference
☆ Pessimistic Evaluation
Traditional evaluation of information access systems has focused primarily on average utility across a set of information needs (information retrieval) or users (recommender systems). In this work, we argue that evaluating only with average metric measurements assumes utilitarian values not aligned with traditions of information access based on equal access. We advocate for pessimistic evaluation of information access systems focusing on worst case utility. These methods are (a) grounded in ethical and pragmatic concepts, (b) theoretically complementary to existing robustness and fairness methods, and (c) empirically validated across a set of retrieval and recommendation tasks. These results suggest that pessimistic evaluation should be included in existing experimentation processes to better understand the behavior of systems, especially when concerned with principles of social good.
☆ Large Language Models as Narrative-Driven Recommenders
Narrative-driven recommenders aim to provide personalized suggestions for user requests expressed in free-form text such as "I want to watch a thriller with a mind-bending story, like Shutter Island." Although large language models (LLMs) have been shown to excel in processing general natural language queries, their effectiveness for handling such recommendation requests remains relatively unexplored. To close this gap, we compare the performance of 38 open- and closed-source LLMs of various sizes, such as LLama 3.2 and GPT-4o, in a movie recommendation setting. For this, we utilize a gold-standard, crowdworker-annotated dataset of posts from reddit's movie suggestion community and employ various prompting strategies, including zero-shot, identity, and few-shot prompting. Our findings demonstrate the ability of LLMs to generate contextually relevant movie recommendations, significantly outperforming other state-of-the-art approaches, such as doc2vec. While we find that closed-source and large-parameterized models generally perform best, medium-sized open-source models remain competitive, being only slightly outperformed by their more computationally expensive counterparts. Furthermore, we observe no significant differences across prompting strategies for most models, underscoring the effectiveness of simple approaches such as zero-shot prompting for narrative-driven recommendations. Overall, this work offers valuable insights for recommender system researchers as well as practitioners aiming to integrate LLMs into real-world recommendation tools.
comment: Under review; 19 pages
☆ Cross-Domain Sequential Recommendation via Neural Process
Cross-Domain Sequential Recommendation (CDSR) is a hot topic in sequence-based user interest modeling, which aims at utilizing a single model to predict the next items for different domains. To tackle the CDSR, many methods are focused on domain overlapped users' behaviors fitting, which heavily relies on the same user's different-domain item sequences collaborating signals to capture the synergy of cross-domain item-item correlation. Indeed, these overlapped users occupy a small fraction of the entire user set only, which introduces a strong assumption that the small group of domain overlapped users is enough to represent all domain user behavior characteristics. However, intuitively, such a suggestion is biased, and the insufficient learning paradigm in non-overlapped users will inevitably limit model performance. Further, it is not trivial to model non-overlapped user behaviors in CDSR because there are no other domain behaviors to collaborate with, which causes the observed single-domain users' behavior sequences to be hard to contribute to cross-domain knowledge mining. Considering such a phenomenon, we raise a challenging and unexplored question: How to unleash the potential of non-overlapped users' behaviors to empower CDSR?
comment: Work in progress
☆ Generate and Instantiate What You Prefer: Text-Guided Diffusion for Sequential Recommendation
Recent advancements in generative recommendation systems, particularly in the realm of sequential recommendation tasks, have shown promise in enhancing generalization to new items. Among these approaches, diffusion-based generative recommendation has emerged as an effective tool, leveraging its ability to capture data distributions and generate high-quality samples. Despite effectiveness, two primary challenges have been identified: 1) the lack of consistent modeling of data distribution for oracle items; and 2) the difficulty in scaling to more informative control signals beyond historical interactions. These issues stem from the uninformative nature of ID embeddings, which necessitate random initialization and limit the incorporation of additional control signals. To address these limitations, we propose iDreamRe } to involve more concrete prior knowledge to establish item embeddings, particularly through detailed item text descriptions and advanced Text Embedding Models (TEM). More importantly, by converting item descriptions into embeddings aligned with TEM, we enable the integration of intention instructions as control signals to guide the generation of oracle items. Experimental results on four datasets demonstrate that iDreamRec not only outperforms existing diffusion-based generative recommenders but also facilitates the incorporation of intention instructions for more precise and effective recommendation generation.
☆ Context-aware adaptive personalised recommendation: a meta-hybrid
Recommenders take place on a wide scale of e-commerce systems, reducing the problem of information overload. The most common approach is to choose a recommender used by the system to make predictions. However, users vary from each other; thus, a one-fits-all approach seems to be sub-optimal. In this paper, we propose a meta-hybrid recommender that uses machine learning to predict an optimal algorithm. In this way, the best-performing recommender is used for each specific session and user. This selection depends on contextual and preferential information collected about the user. We use standard MovieLens and The Movie DB datasets for offline evaluation. We show that based on the proposed model, it is possible to predict which recommender will provide the most precise recommendations to a user. The theoretical performance of our meta-hybrid outperforms separate approaches by 20-50% in normalized Discounted Gain and Root Mean Square Error metrics. However, it is hard to obtain the optimal performance based on widely-used standard information stored about users.
☆ Comparing the Utility, Preference, and Performance of Course Material Search Functionality and Retrieval-Augmented Generation Large Language Model (RAG-LLM) AI Chatbots in Information-Seeking Tasks
Providing sufficient support for students requires substantial resources, especially considering the growing enrollment numbers. Students need help in a variety of tasks, ranging from information-seeking to requiring support with course assignments. To explore the utility of recent large language models (LLMs) as a support mechanism, we developed an LLM-powered AI chatbot that augments the answers that are produced with information from the course materials. To study the effect of the LLM-powered AI chatbot, we conducted a lab-based user study (N=14), in which the participants worked on tasks from a web software development course. The participants were divided into two groups, where one of the groups first had access to the chatbot and then to a more traditional search functionality, while another group started with the search functionality and was then given the chatbot. We assessed the participants' performance and perceptions towards the chatbot and the search functionality and explored their preferences towards the support functionalities. Our findings highlight that both support mechanisms are seen as useful and that support mechanisms work well for specific tasks, while less so for other tasks. We also observe that students tended to prefer the second support mechanism more, where students who were first given the chatbot tended to prefer the search functionality and vice versa.
comment: 12 pages, 4 figures
☆ SBI-RAG: Enhancing Math Word Problem Solving for Students through Schema-Based Instruction and Retrieval-Augmented Generation NeurIPS'24
Many students struggle with math word problems (MWPs), often finding it difficult to identify key information and select the appropriate mathematical operations.Schema-based instruction (SBI) is an evidence-based strategy that helps students categorize problems based on their structure, improving problem-solving accuracy. Building on this, we propose a Schema-Based Instruction Retrieval-Augmented Generation (SBI-RAG) framework that incorporates a large language model (LLM).Our approach emphasizes step-by-step reasoning by leveraging schemas to guide solution generation. We evaluate its performance on the GSM8K dataset, comparing it with GPT-4 and GPT-3.5 Turbo, and introduce a "reasoning score" metric to assess solution quality. Our findings suggest that SBI-RAG enhances reasoning clarity and problem-solving accuracy, potentially providing educational benefits for students
comment: Accepted to the 4th MATH-AI Workshop at NeurIPS'24
☆ Disentangling Likes and Dislikes in Personalized Generative Explainable Recommendation
Recent research on explainable recommendation generally frames the task as a standard text generation problem, and evaluates models simply based on the textual similarity between the predicted and ground-truth explanations. However, this approach fails to consider one crucial aspect of the systems: whether their outputs accurately reflect the users' (post-purchase) sentiments, i.e., whether and why they would like and/or dislike the recommended items. To shed light on this issue, we introduce new datasets and evaluation methods that focus on the users' sentiments. Specifically, we construct the datasets by explicitly extracting users' positive and negative opinions from their post-purchase reviews using an LLM, and propose to evaluate systems based on whether the generated explanations 1) align well with the users' sentiments, and 2) accurately identify both positive and negative opinions of users on the target items. We benchmark several recent models on our datasets and demonstrate that achieving strong performance on existing metrics does not ensure that the generated explanations align well with the users' sentiments. Lastly, we find that existing models can provide more sentiment-aware explanations when the users' (predicted) ratings for the target items are directly fed into the models as input. We will release our code and datasets upon acceptance.
☆ Starbucks: Improved Training for 2D Matryoshka Embeddings
Effective approaches that can scale embedding model depth (i.e. layers) and embedding size allow for the creation of models that are highly scalable across different computational resources and task requirements. While the recently proposed 2D Matryoshka training approach can efficiently produce a single embedding model such that its sub-layers and sub-dimensions can measure text similarity, its effectiveness is significantly worse than if smaller models were trained separately. To address this issue, we propose Starbucks, a new training strategy for Matryoshka-like embedding models, which encompasses both the fine-tuning and pre-training phases. For the fine-tuning phase, we discover that, rather than sampling a random sub-layer and sub-dimensions for each training steps, providing a fixed list of layer-dimension pairs, from small size to large sizes, and computing the loss across all pairs significantly improves the effectiveness of 2D Matryoshka embedding models, bringing them on par with their separately trained counterparts. To further enhance performance, we introduce a new pre-training strategy, which applies masked autoencoder language modelling to sub-layers and sub-dimensions during pre-training, resulting in a stronger backbone for subsequent fine-tuning of the embedding model. Experimental results on both semantic text similarity and retrieval benchmarks demonstrate that the proposed pre-training and fine-tuning strategies significantly improved the effectiveness over 2D Matryoshka models, enabling Starbucks models to perform more efficiently and effectively than separately trained models.
☆ Research on Travel Route Planing Problems Based on Greedy Algorithm
The greedy algorithm based route planning problem is a method of finding the optimal or near optimal route between a given starting and ending point. This article first uses PCA method to reduce the dimensionality of urban evaluation indicators, extracts key principal components, and KMO and TOPSIS algorithms to reduce the dimensionality of the data. Secondly, for datasets that have not passed the KMO test, a comprehensive evaluation will be conducted using the entropy weight method and TOPSIS method. Finally, based on the greedy algorithm, a route planning algorithm was proposed and optimized to provide personalized route customization according to the different needs of tourists. We also took into account the local travel efficiency, the time required to visit tourist attractions, and necessary daily rest time to reduce costs and avoid falling into the local optimal solution.
☆ MixEHR-Nest: Identifying Subphenotypes within Electronic Health Records through Hierarchical Guided-Topic Modeling
Automatic subphenotyping from electronic health records (EHRs)provides numerous opportunities to understand diseases with unique subgroups and enhance personalized medicine for patients. However, existing machine learning algorithms either focus on specific diseases for better interpretability or produce coarse-grained phenotype topics without considering nuanced disease patterns. In this study, we propose a guided topic model, MixEHR-Nest, to infer sub-phenotype topics from thousands of disease using multi-modal EHR data. Specifically, MixEHR-Nest detects multiple subtopics from each phenotype topic, whose prior is guided by the expert-curated phenotype concepts such as Phenotype Codes (PheCodes) or Clinical Classification Software (CCS) codes. We evaluated MixEHR-Nest on two EHR datasets: (1) the MIMIC-III dataset consisting of over 38 thousand patients from intensive care unit (ICU) from Beth Israel Deaconess Medical Center (BIDMC) in Boston, USA; (2) the healthcare administrative database PopHR, comprising 1.3 million patients from Montreal, Canada. Experimental results demonstrate that MixEHR-Nest can identify subphenotypes with distinct patterns within each phenotype, which are predictive for disease progression and severity. Consequently, MixEHR-Nest distinguishes between type 1 and type 2 diabetes by inferring subphenotypes using CCS codes, which do not differentiate these two subtype concepts. Additionally, MixEHR-Nest not only improved the prediction accuracy of short-term mortality of ICU patients and initial insulin treatment in diabetic patients but also revealed the contributions of subphenotypes. For longitudinal analysis, MixEHR-Nest identified subphenotypes of distinct age prevalence under the same phenotypes, such as asthma, leukemia, epilepsy, and depression. The MixEHR-Nest software is available at GitHub: https://github.com/li-lab-mcgill/MixEHR-Nest.
Transformers4NewsRec: A Transformer-based News Recommendation Framework
Pre-trained transformer models have shown great promise in various natural language processing tasks, including personalized news recommendations. To harness the power of these models, we introduce Transformers4NewsRec, a new Python framework built on the \textbf{Transformers} library. This framework is designed to unify and compare the performance of various news recommendation models, including deep neural networks and graph-based models. Transformers4NewsRec offers flexibility in terms of model selection, data preprocessing, and evaluation, allowing both quantitative and qualitative analysis.
☆ Retrieval-Enhanced Named Entity Recognition
When combined with In-Context Learning, a technique that enables models to adapt to new tasks by incorporating task-specific examples or demonstrations directly within the input prompt, autoregressive language models have achieved good performance in a wide range of tasks and applications. However, this combination has not been properly explored in the context of named entity recognition, where the structure of this task poses unique challenges. We propose RENER (Retrieval-Enhanced Named Entity Recognition), a technique for named entity recognition using autoregressive language models based on In-Context Learning and information retrieval techniques. When presented with an input text, RENER fetches similar examples from a dataset of training examples that are used to enhance a language model to recognize named entities from this input text. RENER is modular and independent of the underlying language model and information retrieval algorithms. Experimental results show that in the CrossNER collection we achieve state-of-the-art performance with the proposed technique and that information retrieval can increase the F-score by up to 11 percentage points.
comment: 13 pages, 6 figures, 3 tables
☆ Preference Diffusion for Recommendation
Recommender systems predict personalized item rankings based on user preference distributions derived from historical behavior data. Recently, diffusion models (DMs) have gained attention in recommendation for their ability to model complex distributions, yet current DM-based recommenders often rely on traditional objectives like mean squared error (MSE) or recommendation objectives, which are not optimized for personalized ranking tasks or fail to fully leverage DM's generative potential. To address this, we propose PreferDiff, a tailored optimization objective for DM-based recommenders. PreferDiff transforms BPR into a log-likelihood ranking objective and integrates multiple negative samples to better capture user preferences. Specifically, we employ variational inference to handle the intractability through minimizing the variational upper bound and replaces MSE with cosine error to improve alignment with recommendation tasks. Finally, we balance learning generation and preference to enhance the training stability of DMs. PreferDiff offers three key benefits: it is the first personalized ranking loss designed specifically for DM-based recommenders and it improves ranking and faster convergence by addressing hard negatives. We also prove that it is theoretically connected to Direct Preference Optimization which indicates that it has the potential to align user preferences in DM-based recommenders via generative modeling. Extensive experiments across three benchmarks validate its superior recommendation performance and commendable general sequential recommendation capabilities. Our codes are available at \url{https://github.com/lswhim/PreferDiff}.
☆ Lightweight Correlation-Aware Table Compression
The growing adoption of data lakes for managing relational data necessitates efficient, open storage formats that provide high scan performance and competitive compression ratios. While existing formats achieve fast scans through lightweight encoding techniques, they have reached a plateau in terms of minimizing storage footprint. Recently, correlation-aware compression schemes have been shown to reduce file sizes further. Yet, current approaches either incur significant scan overheads or require manual specification of correlations, limiting their practicability. We present $\texttt{Virtual}$, a framework that integrates seamlessly with existing open formats to automatically leverage data correlations, achieving substantial compression gains while having minimal scan performance overhead. Experiments on $\texttt{data.gov}$ datasets show that $\texttt{Virtual}$ reduces file sizes by up to 40% compared to Apache Parquet.
comment: Third Table Representation Learning Workshop (TRL 2024)
☆ Best in Tau@LLMJudge: Criteria-Based Relevance Evaluation with Llama3
Traditional evaluation of information retrieval (IR) systems relies on human-annotated relevance labels, which can be both biased and costly at scale. In this context, large language models (LLMs) offer an alternative by allowing us to directly prompt them to assign relevance labels for passages associated with each query. In this study, we explore alternative methods to directly prompt LLMs for assigned relevance labels, by exploring two hypotheses: Hypothesis 1 assumes that it is helpful to break down "relevance" into specific criteria - exactness, coverage, topicality, and contextual fit. We explore different approaches that prompt large language models (LLMs) to obtain criteria-level grades for all passages, and we consider various ways to aggregate criteria-level grades into a relevance label. Hypothesis 2 assumes that differences in linguistic style between queries and passages may negatively impact the automatic relevance label prediction. We explore whether improvements can be achieved by first synthesizing a summary of the passage in the linguistic style of a query, and then using this summary in place of the passage to assess its relevance. We include an empirical evaluation of our approaches based on data from the LLMJudge challenge run in Summer 2024, where our "Four Prompts" approach obtained the highest scores in Kendall's tau.
☆ Efficient Retrieval of Temporal Event Sequences from Textual Descriptions
Retrieving temporal event sequences from textual descriptions is essential for applications such as analyzing e-commerce behavior, monitoring social media activities, and tracking criminal incidents. In this paper, we introduce TPP-LLM-Embedding, a unified model for efficiently embedding and retrieving event sequences based on natural language descriptions. Built on the TPP-LLM framework, which integrates large language models with temporal point processes, our model encodes both event types and times, generating a sequence-level representation through pooling. Textual descriptions are embedded using the same architecture, ensuring a shared embedding space for both sequences and descriptions. We optimize a contrastive loss based on similarity between these embeddings, bringing matching pairs closer and separating non-matching ones. TPP-LLM-Embedding enables efficient retrieval and demonstrates superior performance compared to baseline models across diverse datasets.
☆ FinQAPT: Empowering Financial Decisions with End-to-End LLM-driven Question Answering Pipeline
Financial decision-making hinges on the analysis of relevant information embedded in the enormous volume of documents in the financial domain. To address this challenge, we developed FinQAPT, an end-to-end pipeline that streamlines the identification of relevant financial reports based on a query, extracts pertinent context, and leverages Large Language Models (LLMs) to perform downstream tasks. To evaluate the pipeline, we experimented with various techniques to optimize the performance of each module using the FinQA dataset. We introduced a novel clustering-based negative sampling technique to enhance context extraction and a novel prompting method called Dynamic N-shot Prompting to boost the numerical question-answering capabilities of LLMs. At the module level, we achieved state-of-the-art accuracy on FinQA, attaining an accuracy of 80.6\%. However, at the pipeline level, we observed decreased performance due to challenges in extracting relevant context from financial reports. We conducted a detailed error analysis of each module and the end-to-end pipeline, pinpointing specific challenges that must be addressed to develop a robust solution for handling complex financial tasks.
comment: Accepted in ICAIF 2024, 8 pages, 5 figures, 4 tables
☆ Identifying High Consideration E-Commerce Search Queries EMNLP 2024
In e-commerce, high consideration search missions typically require careful and elaborate decision making, and involve a substantial research investment from customers. We consider the task of identifying High Consideration (HC) queries. Identifying such queries enables e-commerce sites to better serve user needs using targeted experiences such as curated QA widgets that help users reach purchase decisions. We explore the task by proposing an Engagement-based Query Ranking (EQR) approach, focusing on query ranking to indicate potential engagement levels with query-related shopping knowledge content during product search. Unlike previous studies on predicting trends, EQR prioritizes query-level features related to customer behavior, finance, and catalog information rather than popularity signals. We introduce an accurate and scalable method for EQR and present experimental results demonstrating its effectiveness. Offline experiments show strong ranking performance. Human evaluation shows a precision of 96% for HC queries identified by our model. The model was commercially deployed, and shown to outperform human-selected queries in terms of downstream customer impact, as measured through engagement.
comment: Accepted by EMNLP 2024 (Industry Track)
♻ ☆ Mixed-Precision Embeddings for Large-Scale Recommendation Models
Embedding techniques have become essential components of large databases in the deep learning era. By encoding discrete entities, such as words, items, or graph nodes, into continuous vector spaces, embeddings facilitate more efficient storage, retrieval, and processing in large databases. Especially in the domain of recommender systems, millions of categorical features are encoded as unique embedding vectors, which facilitates the modeling of similarities and interactions among features. However, numerous embedding vectors can result in significant storage overhead. In this paper, we aim to compress the embedding table through quantization techniques. Given that features vary in importance levels, we seek to identify an appropriate precision for each feature to balance model accuracy and memory usage. To this end, we propose a novel embedding compression method, termed Mixed-Precision Embeddings (MPE). Specifically, to reduce the size of the search space, we first group features by frequency and then search precision for each feature group. MPE further learns the probability distribution over precision levels for each feature group, which can be used to identify the most suitable precision with a specially designed sampling strategy. Extensive experiments on three public datasets demonstrate that MPE significantly outperforms existing embedding compression methods. Remarkably, MPE achieves about 200x compression on the Criteo dataset without comprising the prediction accuracy.
comment: under submision
♻ ☆ A Survey on Intent-aware Recommender Systems
Many modern online services feature personalized recommendations. A central challenge when providing such recommendations is that the reason why an individual user accesses the service may change from visit to visit or even during an ongoing usage session. To be effective, a recommender system should therefore aim to take the users' probable intent of using the service at a certain point in time into account. In recent years, researchers have thus started to address this challenge by incorporating intent-awareness into recommender systems. Correspondingly, a number of technical approaches were put forward, including diversification techniques, intent prediction models or latent intent modeling approaches. In this paper, we survey and categorize existing approaches to building the next generation of Intent-Aware Recommender Systems (IARS). Based on an analysis of current evaluation practices, we outline open gaps and possible future directions in this area, which in particular include the consideration of additional interaction signals and contextual information to further improve the effectiveness of such systems.
♻ ☆ Chatbot-Based Ontology Interaction Using Large Language Models and Domain-Specific Standards
The following contribution introduces a concept that employs Large Language Models (LLMs) and a chatbot interface to enhance SPARQL query generation for ontologies, thereby facilitating intuitive access to formalized knowledge. Utilizing natural language inputs, the system converts user inquiries into accurate SPARQL queries that strictly query the factual content of the ontology, effectively preventing misinformation or fabrication by the LLM. To enhance the quality and precision of outcomes, additional textual information from established domain-specific standards is integrated into the ontology for precise descriptions of its concepts and relationships. An experimental study assesses the accuracy of generated SPARQL queries, revealing significant benefits of using LLMs for querying ontologies and highlighting areas for future research.
comment: \c{opyright} 2024 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works
♻ ☆ DiffATR: Diffusion-based Generative Modeling for Audio-Text Retrieval
Existing audio-text retrieval (ATR) methods are essentially discriminative models that aim to maximize the conditional likelihood, represented as p(candidates|query). Nevertheless, this methodology fails to consider the intrinsic data distribution p(query), leading to difficulties in discerning out-of-distribution data. In this work, we attempt to tackle this constraint through a generative perspective and model the relationship between audio and text as their joint probability p(candidates,query). To this end, we present a diffusion-based ATR framework (DiffATR), which models ATR as an iterative procedure that progressively generates joint distribution from noise. Throughout its training phase, DiffATR is optimized from both generative and discriminative viewpoints: the generator is refined through a generation loss, while the feature extractor benefits from a contrastive loss, thus combining the merits of both methodologies. Experiments on the AudioCaps and Clotho datasets with superior performances, verify the effectiveness of our approach. Notably, without any alterations, our DiffATR consistently exhibits strong performance in out-of-domain retrieval settings.
comment: Accepted by Interspeech2024
♻ ☆ Probability Distribution Learning: A theoretical framework for Deep Learning
This paper introduces probability distribution learning (PD learning), a novel theoretical learning framework. Departing from the traditional statistical learning framework, PD learning focuses on learning the underlying probability distribution, which is modeled as a random variable within the probability simplex. Within this framework, the learning error is decomposed into uncertainty, estimation error, and the model's fitting error. Subsequently, we present the methodology for calculating uncertainty, along with optimization strategies for both estimation error and fitting error. Given that minimizing the fitting error typically constitutes a non-convex optimization problem, we introduce a standard loss function and the gradient structural control (GSC) algorithm, and demonstrate that by employing this function, the optima of fitting error minimization can be approached by reducing the gradient norm and structural error. Furthermore, we apply the PD learning framework to deep learning, elucidating the mechanisms by which techniques such as random parameter initialization, over-parameterization, bias-variance trade-off, and dropout influence deep model training. Finally, experimental results on various models validate the effectiveness of the proposed framework.
comment: arXiv admin note: text overlap with arXiv:2105.04026 by other authors. arXiv admin note: text overlap with arXiv:2105.04026 by other authors
♻ ☆ NFT1000: A Cross-Modal Dataset for Non-Fungible Token Retrieval
With the rise of "Metaverse" and "Web 3.0", Non-Fungible Token (NFT) has emerged as a kind of pivotal digital asset, garnering significant attention. By the end of March 2024, more than 1.7 billion NFTs have been minted across various blockchain platforms. To effectively locate a desired NFT, conducting searches within a vast array of NFTs is essential. The challenge in NFT retrieval is heightened due to the high degree of similarity among different NFTs, regarding regional and semantic aspects. In this paper, we will introduce a benchmark dataset named "NFT Top1000 Visual-Text Dataset" (NFT1000), containing 7.56 million image-text pairs, and being collected from 1000 most famous PFP1 NFT collections2 by sales volume on the Ethereum blockchain. Based on this dataset and leveraging the CLIP series of pre-trained models as our foundation, we propose the dynamic masking fine-tuning scheme. This innovative approach results in a 7.4\% improvement in the top1 accuracy rate, while utilizing merely 13\% of the total training data (0.79 million vs. 6.1 million). We also propose a robust metric Comprehensive Variance Index (CVI) to assess the similarity and retrieval difficulty of visual-text pairs data. The dataset will be released as an open-source resource. For more details, please refer to: https://github.com/ShuxunoO/NFT-Net.git.
comment: 11 pages,12figures to be published in ACM Multimedia 2024, see https://openreview.net/forum?id=xUtNrKH8iB¬eId=xUtNrKH8iB
♻ ☆ RAGEval: Scenario Specific RAG Evaluation Dataset Generation Framework
Retrieval-Augmented Generation (RAG) is a powerful approach that enables large language models (LLMs) to incorporate external knowledge. However, evaluating the effectiveness of RAG systems in specialized scenarios remains challenging due to the high costs of data construction and the lack of suitable evaluation metrics. This paper introduces RAGEval, a framework designed to assess RAG systems across diverse scenarios by generating high-quality documents, questions, answers, and references through a schema-based pipeline. With a focus on factual accuracy, we propose three novel metrics Completeness, Hallucination, and Irrelevance to rigorously evaluate LLM-generated responses. Experimental results show that RAGEval outperforms zero-shot and one-shot methods in terms of clarity, safety, conformity, and richness of generated samples. Furthermore, the use of LLMs for scoring the proposed metrics demonstrates a high level of consistency with human evaluations. RAGEval establishes a new paradigm for evaluating RAG systems in real-world applications.
comment: https://github.com/OpenBMB/RAGEval
♻ ☆ A Theory for Token-Level Harmonization in Retrieval-Augmented Generation
Retrieval-augmented generation (RAG) utilizes retrieved texts to enhance large language models (LLMs). Studies show that while RAG provides valuable external information (benefit), it may also mislead LLMs (detriment) with noisy or incorrect retrieved texts. Although many existing methods attempt to preserve benefit and avoid detriment, they lack a theoretical explanation for RAG. The benefit and detriment in the next token prediction of RAG remain a black box that cannot be quantified or compared in an explainable manner, so existing methods are data-driven, need additional utility evaluators or post-hoc. This paper takes the first step towards providing a theory to explain and trade off the benefit and detriment in RAG. First, we model RAG as the fusion between distribution of LLMs knowledge and distribution of retrieved texts. Then, we formalize the trade-off between the value of external knowledge (benefit) and its potential risk of misleading LLMs (detriment) in next token prediction of RAG by distribution difference in this fusion. Finally, we prove that the actual effect of RAG on the token, which is the comparison between benefit and detriment, can be predicted without any training or accessing the utility of retrieval. Based on our theory, we propose a practical novel method, Tok-RAG, which achieves collaborative generation between the pure LLM and RAG at token level to preserve benefit and avoid detriment. Experiments in real-world tasks using LLMs such as OPT, LLaMA-2, and Mistral show the effectiveness of our method and support our theoretical findings.
comment: 25 pages
♻ ☆ Improving Bilingual Lexicon Induction with Cross-Encoder Reranking EMNLP 2022
Bilingual lexicon induction (BLI) with limited bilingual supervision is a crucial yet challenging task in multilingual NLP. Current state-of-the-art BLI methods rely on the induction of cross-lingual word embeddings (CLWEs) to capture cross-lingual word similarities; such CLWEs are obtained 1) via traditional static models (e.g., VecMap), or 2) by extracting type-level CLWEs from multilingual pretrained language models (mPLMs), or 3) through combining the former two options. In this work, we propose a novel semi-supervised post-hoc reranking method termed BLICEr (BLI with Cross-Encoder Reranking), applicable to any precalculated CLWE space, which improves their BLI capability. The key idea is to 'extract' cross-lingual lexical knowledge from mPLMs, and then combine it with the original CLWEs. This crucial step is done via 1) creating a word similarity dataset, comprising positive word pairs (i.e., true translations) and hard negative pairs induced from the original CLWE space, and then 2) fine-tuning an mPLM (e.g., mBERT or XLM-R) in a cross-encoder manner to predict the similarity scores. At inference, we 3) combine the similarity score from the original CLWE space with the score from the BLI-tuned cross-encoder. BLICEr establishes new state-of-the-art results on two standard BLI benchmarks spanning a wide spectrum of diverse languages: it substantially outperforms a series of strong baselines across the board. We also validate the robustness of BLICEr with different CLWEs.
comment: Findings of EMNLP 2022
♻ ☆ The Moral Case for Using Language Model Agents for Recommendation
Our information and communication environment has fallen short of the ideals that networked global communication might have served. Identifying all the causes of its pathologies is difficult, but existing recommender systems very likely play a contributing role. In this paper, which draws on the normative tools of philosophy of computing, informed by empirical and technical insights from natural language processing and recommender systems, we make the moral case for an alternative approach. We argue that existing recommenders incentivise mass surveillance, concentrate power, fall prey to narrow behaviourism, and compromise user agency. Rather than just trying to avoid algorithms entirely, or to make incremental improvements to the current paradigm, researchers and engineers should explore an alternative paradigm: the use of language model (LM) agents to source and curate content that matches users' preferences and values, expressed in natural language. The use of LM agents for recommendation poses its own challenges, including those related to candidate generation, computational efficiency, preference modelling, and prompt injection. Nonetheless, if implemented successfully LM agents could: guide us through the digital public sphere without relying on mass surveillance; shift power away from platforms towards users; optimise for what matters instead of just for behavioural proxies; and scaffold our agency instead of undermining it.
♻ ☆ Improving Word Translation via Two-Stage Contrastive Learning ACL 2022
Word translation or bilingual lexicon induction (BLI) is a key cross-lingual task, aiming to bridge the lexical gap between different languages. In this work, we propose a robust and effective two-stage contrastive learning framework for the BLI task. At Stage C1, we propose to refine standard cross-lingual linear maps between static word embeddings (WEs) via a contrastive learning objective; we also show how to integrate it into the self-learning procedure for even more refined cross-lingual maps. In Stage C2, we conduct BLI-oriented contrastive fine-tuning of mBERT, unlocking its word translation capability. We also show that static WEs induced from the `C2-tuned' mBERT complement static WEs from Stage C1. Comprehensive experiments on standard BLI datasets for diverse languages and different experimental setups demonstrate substantial gains achieved by our framework. While the BLI method from Stage C1 already yields substantial gains over all state-of-the-art BLI methods in our comparison, even stronger improvements are met with the full two-stage framework: e.g., we report gains for 112/112 BLI setups, spanning 28 language pairs.
comment: ACL 2022 Main
♻ ☆ Into the Unknown Unknowns: Engaged Human Learning through Participation in Language Model Agent Conversations EMNLP 2024
While language model (LM)-powered chatbots and generative search engines excel at answering concrete queries, discovering information in the terrain of unknown unknowns remains challenging for users. To emulate the common educational scenario where children/students learn by listening to and participating in conversations of their parents/teachers, we create Collaborative STORM (Co-STORM). Unlike QA systems that require users to ask all the questions, Co-STORM lets users observe and occasionally steer the discourse among several LM agents. The agents ask questions on the user's behalf, allowing the user to discover unknown unknowns serendipitously. To facilitate user interaction, Co-STORM assists users in tracking the discourse by organizing the uncovered information into a dynamic mind map, ultimately generating a comprehensive report as takeaways. For automatic evaluation, we construct the WildSeek dataset by collecting real information-seeking records with user goals. Co-STORM outperforms baseline methods on both discourse trace and report quality. In a further human evaluation, 70% of participants prefer Co-STORM over a search engine, and 78% favor it over a RAG chatbot.
comment: EMNLP 2024 Main
♻ ☆ Mixed-initiative Query Rewriting in Conversational Passage Retrieval
In this paper, we report our methods and experiments for the TREC Conversational Assistance Track (CAsT) 2022. In this work, we aim to reproduce multi-stage retrieval pipelines and explore one of the potential benefits of involving mixed-initiative interaction in conversational passage retrieval scenarios: reformulating raw queries. Before the first ranking stage of a multi-stage retrieval pipeline, we propose a mixed-initiative query rewriting module, which achieves query rewriting based on the mixed-initiative interaction between the users and the system, as the replacement for the neural rewriting method. Specifically, we design an algorithm to generate appropriate questions related to the ambiguities in raw queries, and another algorithm to reformulate raw queries by parsing users' feedback and incorporating it into the raw query. For the first ranking stage of our multi-stage pipelines, we adopt a sparse ranking function: BM25, and a dense retrieval method: TCT-ColBERT. For the second-ranking step, we adopt a pointwise reranker: MonoT5, and a pairwise reranker: DuoT5. Experiments on both TREC CAsT 2021 and TREC CAsT 2022 datasets show the effectiveness of our mixed-initiative-based query rewriting (or query reformulation) method on improving retrieval performance compared with two popular reformulators: a neural reformulator: CANARD-T5 and a rule-based reformulator: historical query reformulator(HQE).
comment: https://trec.nist.gov/pubs/trec31/papers/udel_fang.C.pdf
♻ ☆ Behavior Alignment: A New Perspective of Evaluating LLM-based Conversational Recommender Systems SIGIR
Large Language Models (LLMs) have demonstrated great potential in Conversational Recommender Systems (CRS). However, the application of LLMs to CRS has exposed a notable discrepancy in behavior between LLM-based CRS and human recommenders: LLMs often appear inflexible and passive, frequently rushing to complete the recommendation task without sufficient inquiry.This behavior discrepancy can lead to decreased accuracy in recommendations and lower user satisfaction. Despite its importance, existing studies in CRS lack a study about how to measure such behavior discrepancy. To fill this gap, we propose Behavior Alignment, a new evaluation metric to measure how well the recommendation strategies made by a LLM-based CRS are consistent with human recommenders'. Our experiment results show that the new metric is better aligned with human preferences and can better differentiate how systems perform than existing evaluation metrics. As Behavior Alignment requires explicit and costly human annotations on the recommendation strategies, we also propose a classification-based method to implicitly measure the Behavior Alignment based on the responses. The evaluation results confirm the robustness of the method.
comment: Accepted by the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval
♻ ☆ ZeQR: Zero-shot Query Reformulation for Conversational Search SIGIR
As the popularity of voice assistants continues to surge, conversational search has gained increased attention in Information Retrieval. However, data sparsity issues in conversational search significantly hinder the progress of supervised conversational search methods. Consequently, researchers are focusing more on zero-shot conversational search approaches. Nevertheless, existing zero-shot methods face three primary limitations: they are not universally applicable to all retrievers, their effectiveness lacks sufficient explainability, and they struggle to resolve common conversational ambiguities caused by omission. To address these limitations, we introduce a novel Zero-shot Query Reformulation (or Query Rewriting) (ZeQR) framework that reformulates queries based on previous dialogue contexts without requiring supervision from conversational search data. Specifically, our framework utilizes language models designed for machine reading comprehension tasks to explicitly resolve two common ambiguities: coreference and omission, in raw queries. In comparison to existing zero-shot methods, our approach is universally applicable to any retriever without additional adaptation or indexing. It also provides greater explainability and effectively enhances query intent understanding because ambiguities are explicitly and proactively resolved. Through extensive experiments on four TREC conversational datasets, we demonstrate the effectiveness of our method, which consistently outperforms state-of-the-art baselines.
comment: Accepted by the 9th ACM SIGIR International Conference on the Theory of Information Retrieval
♻ ☆ Dataset Condensation for Recommendation
Training recommendation models on large datasets requires significant time and resources. It is desired to construct concise yet informative datasets for efficient training. Recent advances in dataset condensation show promise in addressing this problem by synthesizing small datasets. However, applying existing methods of dataset condensation to recommendation has limitations: (1) they fail to generate discrete user-item interactions, and (2) they could not preserve users' potential preferences. To address the limitations, we propose a lightweight condensation framework tailored for recommendation (DConRec), focusing on condensing user-item historical interaction sets. Specifically, we model the discrete user-item interactions via a probabilistic approach and design a pre-augmentation module to incorporate the potential preferences of users into the condensed datasets. While the substantial size of datasets leads to costly optimization, we propose a lightweight policy gradient estimation to accelerate the data synthesis. Experimental results on multiple real-world datasets have demonstrated the effectiveness and efficiency of our framework. Besides, we provide a theoretical analysis of the provable convergence of DConRec. Our implementation is available at: https://github.com/JiahaoWuGit/DConRec.
comment: Accepted by IEEE TKDE. Also titled as "Condensing Pre-augmented Recommendation Data via Lightweight Policy Gradient Estimation"
Machine Learning 201
Fluid: Scaling Autoregressive Text-to-image Generative Models with Continuous Tokens
Scaling up autoregressive models in vision has not proven as beneficial as in large language models. In this work, we investigate this scaling problem in the context of text-to-image generation, focusing on two critical factors: whether models use discrete or continuous tokens, and whether tokens are generated in a random or fixed raster order using BERT- or GPT-like transformer architectures. Our empirical results show that, while all models scale effectively in terms of validation loss, their evaluation performance -- measured by FID, GenEval score, and visual quality -- follows different trends. Models based on continuous tokens achieve significantly better visual quality than those using discrete tokens. Furthermore, the generation order and attention mechanisms significantly affect the GenEval score: random-order models achieve notably better GenEval scores compared to raster-order models. Inspired by these findings, we train Fluid, a random-order autoregressive model on continuous tokens. Fluid 10.5B model achieves a new state-of-the-art zero-shot FID of 6.16 on MS-COCO 30K, and 0.69 overall score on the GenEval benchmark. We hope our findings and results will encourage future efforts to further bridge the scaling gap between vision and language models.
comment: Tech report
☆ How Numerical Precision Affects Mathematical Reasoning Capabilities of LLMs
Despite the remarkable success of Transformer-based Large Language Models (LLMs) across various domains, understanding and enhancing their mathematical capabilities remains a significant challenge. In this paper, we conduct a rigorous theoretical analysis of LLMs' mathematical abilities, with a specific focus on their arithmetic performances. We identify numerical precision as a key factor that influences their effectiveness in mathematical tasks. Our results show that Transformers operating with low numerical precision fail to address arithmetic tasks, such as iterated addition and integer multiplication, unless the model size grows super-polynomially with respect to the input length. In contrast, Transformers with standard numerical precision can efficiently handle these tasks with significantly smaller model sizes. We further support our theoretical findings through empirical experiments that explore the impact of varying numerical precision on arithmetic tasks, providing valuable insights for improving the mathematical reasoning capabilities of LLMs.
☆ Diffusing States and Matching Scores: A New Framework for Imitation Learning
Adversarial Imitation Learning is traditionally framed as a two-player zero-sum game between a learner and an adversarially chosen cost function, and can therefore be thought of as the sequential generalization of a Generative Adversarial Network (GAN). A prominent example of this framework is Generative Adversarial Imitation Learning (GAIL). However, in recent years, diffusion models have emerged as a non-adversarial alternative to GANs that merely require training a score function via regression, yet produce generations of a higher quality. In response, we investigate how to lift insights from diffusion modeling to the sequential setting. We propose diffusing states and performing score-matching along diffused states to measure the discrepancy between the expert's and learner's states. Thus, our approach only requires training score functions to predict noises via standard regression, making it significantly easier and more stable to train than adversarial methods. Theoretically, we prove first- and second-order instance-dependent bounds with linear scaling in the horizon, proving that our approach avoids the compounding errors that stymie offline approaches to imitation learning. Empirically, we show our approach outperforms GAN-style imitation learning baselines across various continuous control problems, including complex tasks like controlling humanoids to walk, sit, and crawl.
☆ AutoAL: Automated Active Learning with Differentiable Query Strategy Search
As deep learning continues to evolve, the need for data efficiency becomes increasingly important. Considering labeling large datasets is both time-consuming and expensive, active learning (AL) provides a promising solution to this challenge by iteratively selecting the most informative subsets of examples to train deep neural networks, thereby reducing the labeling cost. However, the effectiveness of different AL algorithms can vary significantly across data scenarios, and determining which AL algorithm best fits a given task remains a challenging problem. This work presents the first differentiable AL strategy search method, named AutoAL, which is designed on top of existing AL sampling strategies. AutoAL consists of two neural nets, named SearchNet and FitNet, which are optimized concurrently under a differentiable bi-level optimization framework. For any given task, SearchNet and FitNet are iteratively co-optimized using the labeled data, learning how well a set of candidate AL algorithms perform on that task. With the optimal AL strategies identified, SearchNet selects a small subset from the unlabeled pool for querying their annotations, enabling efficient training of the task model. Experimental results demonstrate that AutoAL consistently achieves superior accuracy compared to all candidate AL algorithms and other selective AL approaches, showcasing its potential for adapting and integrating multiple existing AL methods across diverse tasks and domains. Code will be available at: https://github.com/haizailache999/AutoAL.
☆ Retrospective Learning from Interactions
Multi-turn interactions between large language models (LLMs) and users naturally include implicit feedback signals. If an LLM responds in an unexpected way to an instruction, the user is likely to signal it by rephrasing the request, expressing frustration, or pivoting to an alternative task. Such signals are task-independent and occupy a relatively constrained subspace of language, allowing the LLM to identify them even if it fails on the actual task. This creates an avenue for continually learning from interactions without additional annotations. We introduce ReSpect, a method to learn from such signals in past interactions via retrospection. We deploy ReSpect in a new multimodal interaction scenario, where humans instruct an LLM to solve an abstract reasoning task with a combinatorial solution space. Through thousands of interactions with humans, we show how ReSpect gradually improves task completion rate from 31% to 82%, all without any external annotation.
☆ Influence Functions for Scalable Data Attribution in Diffusion Models
Diffusion models have led to significant advancements in generative modelling. Yet their widespread adoption poses challenges regarding data attribution and interpretability. In this paper, we aim to help address such challenges in diffusion models by developing an \textit{influence functions} framework. Influence function-based data attribution methods approximate how a model's output would have changed if some training data were removed. In supervised learning, this is usually used for predicting how the loss on a particular example would change. For diffusion models, we focus on predicting the change in the probability of generating a particular example via several proxy measurements. We show how to formulate influence functions for such quantities and how previously proposed methods can be interpreted as particular design choices in our framework. To ensure scalability of the Hessian computations in influence functions, we systematically develop K-FAC approximations based on generalised Gauss-Newton matrices specifically tailored to diffusion models. We recast previously proposed methods as specific design choices in our framework and show that our recommended method outperforms previous data attribution approaches on common evaluations, such as the Linear Data-modelling Score (LDS) or retraining without top influences, without the need for method-specific hyperparameter tuning.
☆ From Gradient Clipping to Normalization for Heavy Tailed SGD
Recent empirical evidence indicates that many machine learning applications involve heavy-tailed gradient noise, which challenges the standard assumptions of bounded variance in stochastic optimization. Gradient clipping has emerged as a popular tool to handle this heavy-tailed noise, as it achieves good performance in this setting both theoretically and practically. However, our current theoretical understanding of non-convex gradient clipping has three main shortcomings. First, the theory hinges on large, increasing clipping thresholds, which are in stark contrast to the small constant clipping thresholds employed in practice. Second, clipping thresholds require knowledge of problem-dependent parameters to guarantee convergence. Lastly, even with this knowledge, current sampling complexity upper bounds for the method are sub-optimal in nearly all parameters. To address these issues, we study convergence of Normalized SGD (NSGD). First, we establish a parameter-free sample complexity for NSGD of $\mathcal{O}\left(\varepsilon^{-\frac{2p}{p-1}}\right)$ to find an $\varepsilon$-stationary point. Furthermore, we prove tightness of this result, by providing a matching algorithm-specific lower bound. In the setting where all problem parameters are known, we show this complexity is improved to $\mathcal{O}\left(\varepsilon^{-\frac{3p-2}{p-1}}\right)$, matching the previously known lower bound for all first-order methods in all problem dependent parameters. Finally, we establish high-probability convergence of NSGD with a mild logarithmic dependence on the failure probability. Our work complements the studies of gradient clipping under heavy tailed noise improving the sample complexities of existing algorithms and offering an alternative mechanism to achieve high probability convergence.
☆ SimLayerKV: A Simple Framework for Layer-Level KV Cache Reduction
Recent advancements in large language models (LLMs) have extended their capabilities to handle long contexts. However, increasing the number of model layers and the length of input sequences significantly escalates the memory required to store key-value (KV) cache, posing challenges for efficient inference. To mitigate this issue, we present SimLayerKV, a simple yet effective method that reduces inter-layer KV cache redundancies by selectively dropping cache in identified lazy layers. Our approach is based on the observation that certain layers in long-context LLMs exhibit "lazy" behavior, contributing less to modeling long-range dependencies compared to non-lazy layers. By analyzing attention weight patterns, we find that the behavior of these lazy layers is consistent across tokens during generation for a given input. This insight motivates our SimLayerKV, which identifies lazy layers and reduces their KV cache accordingly. SimLayerKV is training-free, generalizable, and can be implemented with only seven lines of code. We conduct extensive experiments on three representative LLMs, e.g., LLaMA2-7B, LLaMA3-8B, and Mistral-7B across 16 tasks from the LongBench benchmark. The results demonstrate that SimLayerKV achieves a KV cache compression ratio of 5$\times$ with only a 1.2% performance drop when combined with 4-bit quantization. Our code is available at https://github.com/sail-sg/SimLayerKV.
☆ A Unified View of Delta Parameter Editing in Post-Trained Large-Scale Models
Post-training has emerged as a crucial paradigm for adapting large-scale pre-trained models to various tasks, whose effects are fully reflected by delta parameters (i.e., the disparity between post-trained and pre-trained parameters). While numerous studies have explored delta parameter properties via operations like pruning, quantization, low-rank approximation, and extrapolation, a unified framework for systematically examining these characteristics has been lacking. In this paper, we propose a novel perspective based on Riemann sum approximation of the loss function to elucidate delta parameter editing operations. Our analysis categorizes existing methods into three classes based on their post-editing performance: competitive, decreased, and improved, explaining how they are expressed by the Riemann sum approximation term and how they alter the model performance. Extensive experiments on both visual and language models, including ViT, LLaMA 3, Qwen 2, and Mistral, corroborate our theoretical findings. Furthermore, we introduce extensions to existing techniques like DARE and BitDelta, highlighting their limitations in leveraging the properties of delta parameters and reorganizing them into general expressions to enhance the applicability and effectiveness of delta parameter editing in post-trained models.
☆ ORSO: Accelerating Reward Design via Online Reward Selection and Policy Optimization
Reward shaping is a critical component in reinforcement learning (RL), particularly for complex tasks where sparse rewards can hinder learning. While shaping rewards have been introduced to provide additional guidance, selecting effective shaping functions remains challenging and computationally expensive. This paper introduces Online Reward Selection and Policy Optimization (ORSO), a novel approach that frames shaping reward selection as an online model selection problem. ORSO employs principled exploration strategies to automatically identify promising shaping reward functions without human intervention, balancing exploration and exploitation with provable regret guarantees. We demonstrate ORSO's effectiveness across various continuous control tasks using the Isaac Gym simulator. Compared to traditional methods that fully evaluate each shaping reward function, ORSO significantly improves sample efficiency, reduces computational time, and consistently identifies high-quality reward functions that produce policies comparable to those generated by domain experts through hand-engineered rewards.
comment: preprint, 35 pages, 23 figures
☆ Active-Dormant Attention Heads: Mechanistically Demystifying Extreme-Token Phenomena in LLMs
Practitioners have consistently observed three puzzling phenomena in transformer-based large language models (LLMs): attention sinks, value-state drains, and residual-state peaks, collectively referred to as extreme-token phenomena. These phenomena are characterized by certain so-called "sink tokens" receiving disproportionately high attention weights, exhibiting significantly smaller value states, and having much larger residual-state norms than those of other tokens. These extreme tokens give rise to various challenges in LLM inference, quantization, and interpretability. We elucidate the mechanisms behind extreme-token phenomena. First, we show that these phenomena arise in very simple architectures -- transformers with one to three layers -- trained on a toy model, the Bigram-Backcopy (BB) task. In this setting, we identify an active-dormant mechanism, where attention heads become sinks for specific input domains while remaining non-sinks for others. Our theoretical analysis of the training dynamics reveals that these phenomena are driven by a mutual reinforcement mechanism. Building on these insights, we propose strategies to mitigate extreme-token phenomena during pretraining, including replacing softmax with ReLU and Adam with SGD. Next, we extend our analysis to pretrained LLMs, including Llama and OLMo, showing that many attention heads exhibit a similar active-dormant mechanism as in the BB task, and that the mutual reinforcement mechanism also governs the emergence of extreme-token phenomena during LLM pretraining. Our results reveal that many of the static and dynamic properties of extreme-token phenomena predicted by the BB task align with observations in pretrained LLMs.
☆ The Disparate Benefits of Deep Ensembles
Ensembles of Deep Neural Networks, Deep Ensembles, are widely used as a simple way to boost predictive performance. However, their impact on algorithmic fairness is not well understood yet. Algorithmic fairness investigates how a model's performance varies across different groups, typically defined by protected attributes such as age, gender, or race. In this work, we investigate the interplay between the performance gains from Deep Ensembles and fairness. Our analysis reveals that they unevenly favor different groups in what we refer to as a disparate benefits effect. We empirically investigate this effect with Deep Ensembles applied to popular facial analysis and medical imaging datasets, where protected group attributes are given and find that it occurs for multiple established group fairness metrics, including statistical parity and equal opportunity. Furthermore, we identify the per-group difference in predictive diversity of ensemble members as the potential cause of the disparate benefits effect. Finally, we evaluate different approaches to reduce unfairness due to the disparate benefits effect. Our findings show that post-processing is an effective method to mitigate this unfairness while preserving the improved performance of Deep Ensembles.
☆ A Common Pitfall of Margin-based Language Model Alignment: Gradient Entanglement
Reinforcement Learning from Human Feedback (RLHF) has become the predominant approach for language model (LM) alignment. At its core, RLHF uses a margin-based loss for preference optimization, specifying ideal LM behavior only by the difference between preferred and dispreferred responses. In this paper, we identify a common pitfall of margin-based methods -- the under-specification of ideal LM behavior on preferred and dispreferred responses individually, which leads to two unintended consequences as the margin increases: (1) The probability of dispreferred (e.g., unsafe) responses may increase, resulting in potential safety alignment failures. (2) The probability of preferred responses may decrease, even when those responses are ideal. We demystify the reasons behind these problematic behaviors: margin-based losses couple the change in the preferred probability to the gradient of the dispreferred one, and vice versa, often preventing the preferred probability from increasing while the dispreferred one decreases, and thus causing a synchronized increase or decrease in both probabilities. We term this effect, inherent in margin-based objectives, gradient entanglement. Formally, we derive conditions for general margin-based alignment objectives under which gradient entanglement becomes concerning: the inner product of the gradients of preferred and dispreferred log-probabilities is large relative to the individual gradient norms. We theoretically investigate why such inner products can be large when aligning language models and empirically validate our findings. Empirical implications of our framework extend to explaining important differences in the training dynamics of various preference optimization algorithms, and suggesting potential algorithm designs to mitigate the under-specification issue of margin-based methods and thereby improving language model alignment.
☆ Unearthing Skill-Level Insights for Understanding Trade-Offs of Foundation Models
With models getting stronger, evaluations have grown more complex, testing multiple skills in one benchmark and even in the same instance at once. However, skill-wise performance is obscured when inspecting aggregate accuracy, under-utilizing the rich signal modern benchmarks contain. We propose an automatic approach to recover the underlying skills relevant for any evaluation instance, by way of inspecting model-generated rationales. After validating the relevance of rationale-parsed skills and inferring skills for $46$k instances over $12$ benchmarks, we observe many skills to be common across benchmarks, resulting in the curation of hundreds of skill-slices (i.e. sets of instances testing a common skill). Inspecting accuracy over these slices yields novel insights on model trade-offs: e.g., compared to GPT-4o and Claude 3.5 Sonnet, on average, Gemini 1.5 Pro is $18\%$ more accurate in "computing molar mass", but $19\%$ less accurate in "applying constitutional law", despite the overall accuracies of the three models differing by a mere $0.4\%$. Furthermore, we demonstrate the practical utility of our approach by showing that insights derived from skill slice analysis can generalize to held-out instances: when routing each instance to the model strongest on the relevant skills, we see a $3\%$ accuracy improvement over our $12$ dataset corpus. Our skill-slices and framework open a new avenue in model evaluation, leveraging skill-specific analyses to unlock a more granular and actionable understanding of model capabilities.
comment: Code at: github.com/microsoft/skill-slice-insights
☆ Artificial Kuramoto Oscillatory Neurons
It has long been known in both neuroscience and AI that ``binding'' between neurons leads to a form of competitive learning where representations are compressed in order to represent more abstract concepts in deeper layers of the network. More recently, it was also hypothesized that dynamic (spatiotemporal) representations play an important role in both neuroscience and AI. Building on these ideas, we introduce Artificial Kuramoto Oscillatory Neurons (AKOrN) as a dynamical alternative to threshold units, which can be combined with arbitrary connectivity designs such as fully connected, convolutional, or attentive mechanisms. Our generalized Kuramoto updates bind neurons together through their synchronization dynamics. We show that this idea provides performance improvements across a wide spectrum of tasks such as unsupervised object discovery, adversarial robustness, calibrated uncertainty quantification, and reasoning. We believe that these empirical results show the importance of rethinking our assumptions at the most basic neuronal level of neural representation, and in particular show the importance of dynamical representations.
comment: Code: https://github.com/autonomousvision/akorn
☆ Steering Your Generalists: Improving Robotic Foundation Models via Value Guidance
Large, general-purpose robotic policies trained on diverse demonstration datasets have been shown to be remarkably effective both for controlling a variety of robots in a range of different scenes, and for acquiring broad repertoires of manipulation skills. However, the data that such policies are trained on is generally of mixed quality -- not only are human-collected demonstrations unlikely to perform the task perfectly, but the larger the dataset is, the harder it is to curate only the highest quality examples. It also remains unclear how optimal data from one embodiment is for training on another embodiment. In this paper, we present a general and broadly applicable approach that enhances the performance of such generalist robot policies at deployment time by re-ranking their actions according to a value function learned via offline RL. This approach, which we call Value-Guided Policy Steering (V-GPS), is compatible with a wide range of different generalist policies, without needing to fine-tune or even access the weights of the policy. We show that the same value function can improve the performance of five different state-of-the-art policies with different architectures, even though they were trained on distinct datasets, attaining consistent performance improvement on multiple robotic platforms across a total of 12 tasks. Code and videos can be found at: https://nakamotoo.github.io/V-GPS
comment: Conference on Robot Learning (CoRL) 2024. Project Page: https://nakamotoo.github.io/V-GPS
☆ Private Counterfactual Retrieval
Transparency and explainability are two extremely important aspects to be considered when employing black-box machine learning models in high-stake applications. Providing counterfactual explanations is one way of catering this requirement. However, this also poses a threat to the privacy of both the institution that is providing the explanation as well as the user who is requesting it. In this work, we propose multiple schemes inspired by private information retrieval (PIR) techniques which ensure the \emph{user's privacy} when retrieving counterfactual explanations. We present a scheme which retrieves the \emph{exact} nearest neighbor counterfactual explanation from a database of accepted points while achieving perfect (information-theoretic) privacy for the user. While the scheme achieves perfect privacy for the user, some leakage on the database is inevitable which we quantify using a mutual information based metric. Furthermore, we propose strategies to reduce this leakage to achieve an advanced degree of database privacy. We extend these schemes to incorporate user's preference on transforming their attributes, so that a more actionable explanation can be received. Since our schemes rely on finite field arithmetic, we empirically validate our schemes on real datasets to understand the trade-off between the accuracy and the finite field sizes.
☆ Adversarial Testing as a Tool for Interpretability: Length-based Overfitting of Elementary Functions in Transformers
The Transformer model has a tendency to overfit various aspects of the training data, such as the overall sequence length. We study elementary string edit functions using a defined set of error indicators to interpret the behaviour of the sequence-to-sequence Transformer. We show that generalization to shorter sequences is often possible, but confirm that longer sequences are highly problematic, although partially correct answers are often obtained. Additionally, we find that other structural characteristics of the sequences, such as subsegment length, may be equally important. We hypothesize that the models learn algorithmic aspects of the tasks simultaneously with structural aspects but adhering to the structural aspects is unfortunately often preferred by Transformer when they come into conflict.
comment: 9 pages, 8 figures, 2 tables; to be published
☆ Machine-Learning Analysis of Radiative Decays to Dark Matter at the LHC
The search for weakly interacting matter particles (WIMPs) is one of the main objectives of the High Luminosity Large Hadron Collider (HL-LHC). In this work we use Machine Learning (ML) techniques to explore WIMP radiative decays into a Dark Matter (DM) candidate in a supersymmetric framework. The minimal supersymmetric WIMP sector includes the lightest neutralino that can provide the observed DM relic density through its co-annihilation with the second lightest neutralino and lightest chargino. Moreover, the direct DM detection cross section rates fulfill current experimental bounds and provide discovery targets for the same region of model parameters in which the radiative decay of the second lightest neutralino into a photon and the lightest neutralino is enhanced. This strongly motivates the search for radiatively decaying neutralinos which, however, suffers from strong backgrounds. We investigate the LHC reach in the search for these radiatively decaying particles by means of cut-based and ML methods and estimate its discovery potential in this well-motivated, new physics scenario.
comment: 32 pages, 9 figures, 3 tables, 4 appendices
☆ Discrete distributions are learnable from metastable samples
Markov chain samplers designed to sample from multi-variable distributions often undesirably get stuck in specific regions of their state space. This causes such samplers to approximately sample from a metastable distribution which is usually quite different from the desired, stationary distribution of the chain. We show that single-variable conditionals of metastable distributions of reversible Markov chain samplers that satisfy a strong metastability condition are on average very close to those of the true distribution. This holds even when the metastable distribution is far away from the true model in terms of global metrics like Kullback-Leibler divergence or total variation distance. This property allows us to learn the true model using a conditional likelihood based estimator, even when the samples come from a metastable distribution concentrated in a small region of the state space. Explicit examples of such metastable states can be constructed from regions that effectively bottleneck the probability flow and cause poor mixing of the Markov chain. For specific cases of binary pairwise undirected graphical models, we extend our results to further rigorously show that data coming from metastable states can be used to learn the parameters of the energy function and recover the structure of the model.
comment: Preliminary version, 26 pages
☆ Learning Graph Quantized Tokenizers for Transformers
Transformers serve as the backbone architectures of Foundational Models, where a domain-specific tokenizer helps them adapt to various domains. Graph Transformers (GTs) have recently emerged as a leading model in geometric deep learning, outperforming Graph Neural Networks (GNNs) in various graph learning tasks. However, the development of tokenizers for graphs has lagged behind other modalities, with existing approaches relying on heuristics or GNNs co-trained with Transformers. To address this, we introduce GQT (\textbf{G}raph \textbf{Q}uantized \textbf{T}okenizer), which decouples tokenizer training from Transformer training by leveraging multi-task graph self-supervised learning, yielding robust and generalizable graph tokens. Furthermore, the GQT utilizes Residual Vector Quantization (RVQ) to learn hierarchical discrete tokens, resulting in significantly reduced memory requirements and improved generalization capabilities. By combining the GQT with token modulation, a Transformer encoder achieves state-of-the-art performance on 16 out of 18 benchmarks, including large-scale homophilic and heterophilic datasets. The code is available at: https://github.com/limei0307/graph-tokenizer
☆ Arbitrarily-Conditioned Multi-Functional Diffusion for Multi-Physics Emulation
Modern physics simulation often involves multiple functions of interests, and traditional numerical approaches are known to be complex and computationally costly. While machine learning-based surrogate models can offer significant cost reductions, most focus on a single task, such as forward prediction, and typically lack uncertainty quantification -- an essential component in many applications. To overcome these limitations, we propose Arbitrarily-Conditioned Multi-Functional Diffusion (ACMFD), a versatile probabilistic surrogate model for multi-physics emulation. ACMFD can perform a wide range of tasks within a single framework, including forward prediction, various inverse problems, and simulating data for entire systems or subsets of quantities conditioned on others. Specifically, we extend the standard Denoising Diffusion Probabilistic Model (DDPM) for multi-functional generation by modeling noise as Gaussian processes (GP). We then introduce an innovative denoising loss. The training involves randomly sampling the conditioned part and fitting the corresponding predicted noise to zero, enabling ACMFD to flexibly generate function values conditioned on any other functions or quantities. To enable efficient training and sampling, and to flexibly handle irregularly sampled data, we use GPs to interpolate function samples onto a grid, inducing a Kronecker product structure for efficient computation. We demonstrate the advantages of ACMFD across several fundamental multi-physics systems.
☆ Analyzing Deep Transformer Models for Time Series Forecasting via Manifold Learning
Transformer models have consistently achieved remarkable results in various domains such as natural language processing and computer vision. However, despite ongoing research efforts to better understand these models, the field still lacks a comprehensive understanding. This is particularly true for deep time series forecasting methods, where analysis and understanding work is relatively limited. Time series data, unlike image and text information, can be more challenging to interpret and analyze. To address this, we approach the problem from a manifold learning perspective, assuming that the latent representations of time series forecasting models lie next to a low-dimensional manifold. In our study, we focus on analyzing the geometric features of these latent data manifolds, including intrinsic dimension and principal curvatures. Our findings reveal that deep transformer models exhibit similar geometric behavior across layers, and these geometric features are correlated with model performance. Additionally, we observe that untrained models initially have different structures, but they rapidly converge during training. By leveraging our geometric analysis and differentiable tools, we can potentially design new and improved deep forecasting neural networks. This approach complements existing analysis studies and contributes to a better understanding of transformer models in the context of time series forecasting. Code is released at https://github.com/azencot-group/GATLM.
comment: Accepted to TMLR 2024
☆ DPLM-2: A Multimodal Diffusion Protein Language Model
Proteins are essential macromolecules defined by their amino acid sequences, which determine their three-dimensional structures and, consequently, their functions in all living organisms. Therefore, generative protein modeling necessitates a multimodal approach to simultaneously model, understand, and generate both sequences and structures. However, existing methods typically use separate models for each modality, limiting their ability to capture the intricate relationships between sequence and structure. This results in suboptimal performance in tasks that requires joint understanding and generation of both modalities. In this paper, we introduce DPLM-2, a multimodal protein foundation model that extends discrete diffusion protein language model (DPLM) to accommodate both sequences and structures. To enable structural learning with the language model, 3D coordinates are converted to discrete tokens using a lookup-free quantization-based tokenizer. By training on both experimental and high-quality synthetic structures, DPLM-2 learns the joint distribution of sequence and structure, as well as their marginals and conditionals. We also implement an efficient warm-up strategy to exploit the connection between large-scale evolutionary data and structural inductive biases from pre-trained sequence-based protein language models. Empirical evaluation shows that DPLM-2 can simultaneously generate highly compatible amino acid sequences and their corresponding 3D structures eliminating the need for a two-stage generation approach. Moreover, DPLM-2 demonstrates competitive performance in various conditional generation tasks, including folding, inverse folding, and scaffolding with multimodal motif inputs, as well as providing structure-aware representations for predictive tasks.
☆ Optimal Quantization for Matrix Multiplication
Recent work in machine learning community proposed multiple methods for performing lossy compression (quantization) of large matrices. This quantization is important for accelerating matrix multiplication (main component of large language models), which is often bottlenecked by the speed of loading these matrices from memory. Unlike classical vector quantization and rate-distortion theory, the goal of these new compression algorithms is to be able to approximate not the matrices themselves, but their matrix product. Specifically, given a pair of real matrices $A,B$ an encoder (compressor) is applied to each of them independently producing descriptions with $R$ bits per entry. These representations subsequently are used by the decoder to estimate matrix product $A^\top B$. In this work, we provide a non-asymptotic lower bound on the mean squared error of this approximation (as a function of rate $R$) for the case of matrices $A,B$ with iid Gaussian entries. Algorithmically, we construct a universal quantizer based on nested lattices with an explicit guarantee of approximation error for any (non-random) pair of matrices $A$, $B$ in terms of only Frobenius norms $\|A\|_F, \|B\|_F$ and $\|A^\top B\|_F$. For iid Gaussian matrices our quantizer achieves the lower bound and is, thus, asymptotically optimal. A practical low-complexity version of our quantizer achieves performance quite close to optimal. In information-theoretic terms we derive rate-distortion function for matrix multiplication of iid Gaussian matrices.
☆ The Mystery of the Pathological Path-star Task for Language Models EMNLP 2024
The recently introduced path-star task is a minimal task designed to exemplify limitations to the abilities of language models (Bachmann and Nagarajan, 2024). It involves a path-star graph where multiple arms radiate from a single starting node and each node is unique. Given the start node and a specified target node that ends an arm, the task is to generate the arm containing that target node. This is straightforward for a human but surprisingly difficult for language models, which did not outperform the random baseline. The authors hypothesized this is due to a deficiency in teacher-forcing and the next-token prediction paradigm. We demonstrate the task is learnable using teacher-forcing in alternative settings and that the issue is partially due to representation. We introduce a regularization method using structured samples of the same graph but with differing target nodes, improving results across a variety of model types. We provide RASP proofs showing the task is theoretically solvable. Finally, we find settings where an encoder-only model can consistently solve the task.
comment: EMNLP 2024 Main
☆ Change Detection in Multivariate data streams: Online Analysis with Kernel-QuantTree ALT
We present Kernel-QuantTree Exponentially Weighted Moving Average (KQT-EWMA), a non-parametric change-detection algorithm that combines the Kernel-QuantTree (KQT) histogram and the EWMA statistic to monitor multivariate data streams online. The resulting monitoring scheme is very flexible, since histograms can be used to model any stationary distribution, and practical, since the distribution of test statistics does not depend on the distribution of datastream in stationary conditions (non-parametric monitoring). KQT-EWMA enables controlling false alarms by operating at a pre-determined Average Run Length ($ARL_0$), which measures the expected number of stationary samples to be monitored before triggering a false alarm. The latter peculiarity is in contrast with most non-parametric change-detection tests, which rarely can control the $ARL_0$ a priori. Our experiments on synthetic and real-world datasets demonstrate that KQT-EWMA can control $ARL_0$ while achieving detection delays comparable to or lower than state-of-the-art methods designed to work in the same conditions.
comment: AALTD workshop at ECML 2024 (https://ecml-aaltd.github.io/aaltd2024/)
☆ Representing Model Weights with Language using Tree Experts
The increasing availability of public models begs the question: can we train neural networks that use other networks as input? This paper learns to represent models within a joint space that embeds both model weights and language. However, machine learning on model weights is challenging as model weights often exhibit significant variation unrelated to the models' semantic properties (nuisance variation). We identify a key property of real-world models: most public models belong to a small set of Model Trees, where all models within a tree are fine-tuned from a common ancestor (e.g., a foundation model). Importantly, we find that within each tree there is less nuisance variation between models. For example, while classifying models according to their training dataset generally requires complex architectures, in our case, even a linear classifier trained on a single layer is often effective. While effective, linear layers are computationally expensive as model weights are very high dimensional. To address this, we introduce Probing Experts (ProbeX), a theoretically motivated, lightweight probing method. Notably, ProbeX is the first probing method designed to learn from the weights of just a single model layer. We also construct and release a dataset that simulates the structure of public model repositories. Our results show that ProbeX can effectively map the weights of large models into a shared weight-language embedding space. Furthermore, we demonstrate the impressive generalization of our method, achieving zero-shot model classification and retrieval.
☆ Enhancing Retail Sales Forecasting with Optimized Machine Learning Models ICSE
In retail sales forecasting, accurately predicting future sales is crucial for inventory management and strategic planning. Traditional methods like LR often fall short due to the complexity of sales data, which includes seasonality and numerous product families. Recent advancements in machine learning (ML) provide more robust alternatives. This research benefits from the power of ML, particularly Random Forest (RF), Gradient Boosting (GB), Support Vector Regression (SVR), and XGBoost, to improve prediction accuracy. Despite advancements, a significant gap exists in handling complex datasets with high seasonality and multiple product families. The proposed solution involves implementing and optimizing a RF model, leveraging hyperparameter tuning through randomized search cross-validation. This approach addresses the complexities of the dataset, capturing intricate patterns that traditional methods miss. The optimized RF model achieved an R-squared value of 0.945, substantially higher than the initial RF model and traditional LR, which had an R-squared of 0.531. The model reduced the root mean squared logarithmic error (RMSLE) to 1.172, demonstrating its superior predictive capability. The optimized RF model did better than cutting-edge models like Gradient Boosting (R-squared: 0.942), SVR (R-squared: 0.940), and XGBoost (R-squared: 0.939), with more minor mean squared error (MSE) and mean absolute error (MAE) numbers. The results demonstrate that the optimized RF model excels in forecasting retail sales, handling the datasets complexity with higher accuracy and reliability. This research highlights the importance of advanced ML techniques in predictive analytics, offering a significant improvement over traditional methods and other contemporary models.
comment: IEEE 4th ICSES 2024
☆ Is Prior-Free Black-Box Non-Stationary Reinforcement Learning Feasible?
We study the problem of Non-Stationary Reinforcement Learning (NS-RL) without prior knowledge about the system's non-stationarity. A state-of-the-art, black-box algorithm, known as MASTER, is considered, with a focus on identifying the conditions under which it can achieve its stated goals. Specifically, we prove that MASTER's non-stationarity detection mechanism is not triggered for practical choices of horizon, leading to performance akin to a random restarting algorithm. Moreover, we show that the regret bound for MASTER, while being order optimal, stays above the worst-case linear regret until unreasonably large values of the horizon. To validate these observations, MASTER is tested for the special case of piecewise stationary multi-armed bandits, along with methods that employ random restarting, and others that use quickest change detection to restart. A simple, order optimal random restarting algorithm, that has prior knowledge of the non-stationarity is proposed as a baseline. The behavior of the MASTER algorithm is validated in simulations, and it is shown that methods employing quickest change detection are more robust and consistently outperform MASTER and other random restarting approaches.
☆ Probing the Latent Hierarchical Structure of Data via Diffusion Models
High-dimensional data must be highly structured to be learnable. Although the compositional and hierarchical nature of data is often put forward to explain learnability, quantitative measurements establishing these properties are scarce. Likewise, accessing the latent variables underlying such a data structure remains a challenge. In this work, we show that forward-backward experiments in diffusion-based models, where data is noised and then denoised to generate new samples, are a promising tool to probe the latent structure of data. We predict in simple hierarchical models that, in this process, changes in data occur by correlated chunks, with a length scale that diverges at a noise level where a phase transition is known to take place. Remarkably, we confirm this prediction in both text and image datasets using state-of-the-art diffusion models. Our results show how latent variable changes manifest in the data and establish how to measure these effects in real data using diffusion models.
comment: 11 pages, 6 figures
☆ Virtual Sensing for Real-Time Degradation Monitoring of Nuclear Systems: Leveraging DeepONet for Enhanced Sensing Coverage for Digital Twin-Enabling Technology
Effective real-time monitoring technique is crucial for detecting material degradation and maintaining the structural integrity of nuclear systems to ensure both safety and operational efficiency. Traditional physical sensor systems face limitations such as installation challenges, high costs, and difficulties in measuring critical parameters in hard-to-reach or harsh environments, often resulting in incomplete data coverage. Machine learning-driven virtual sensors offer a promising solution by enhancing physical sensor capabilities to monitor critical degradation indicators like pressure, velocity, and turbulence. However, conventional machine learning models struggle with real-time monitoring due to the high-dimensional nature of reactor data and the need for frequent retraining. This paper explores the use of Deep Operator Networks (DeepONet) within a digital twin (DT) framework to predict key thermal-hydraulic parameters in the hot leg of an AP-1000 Pressurized Water Reactor (PWR). In this study, DeepONet is trained with different operational conditions, which relaxes the requirement of continuous retraining, making it suitable for online and real-time prediction components for DT. Our results show that DeepONet achieves accurate predictions with low mean squared error and relative L2 error and can make predictions on unknown data 160,000 times faster than traditional finite element (FE) simulations. This speed and accuracy make DeepONet a powerful tool for tracking conditions that contribute to material degradation in real-time, enhancing reactor safety and longevity.
☆ GDeR: Safeguarding Efficiency, Balancing, and Robustness via Prototypical Graph Pruning NeurIPS 2024
Training high-quality deep models necessitates vast amounts of data, resulting in overwhelming computational and memory demands. Recently, data pruning, distillation, and coreset selection have been developed to streamline data volume by retaining, synthesizing, or selecting a small yet informative subset from the full set. Among these methods, data pruning incurs the least additional training cost and offers the most practical acceleration benefits. However, it is the most vulnerable, often suffering significant performance degradation with imbalanced or biased data schema, thus raising concerns about its accuracy and reliability in on-device deployment. Therefore, there is a looming need for a new data pruning paradigm that maintains the efficiency of previous practices while ensuring balance and robustness. Unlike the fields of computer vision and natural language processing, where mature solutions have been developed to address these issues, graph neural networks (GNNs) continue to struggle with increasingly large-scale, imbalanced, and noisy datasets, lacking a unified dataset pruning solution. To achieve this, we introduce a novel dynamic soft-pruning method, GDeR, designed to update the training ``basket'' during the process using trainable prototypes. GDeR first constructs a well-modeled graph embedding hypersphere and then samples \textit{representative, balanced, and unbiased subsets} from this embedding space, which achieves the goal we called Graph Training Debugging. Extensive experiments on five datasets across three GNN backbones, demonstrate that GDeR (I) achieves or surpasses the performance of the full dataset with 30%~50% fewer training samples, (II) attains up to a 2.81x lossless training speedup, and (III) outperforms state-of-the-art pruning methods in imbalanced training and noisy training scenarios by 0.3%~4.3% and 3.6%~7.8%, respectively.
comment: NeurIPS 2024
☆ CLIMB: Language-Guided Continual Learning for Task Planning with Iterative Model Building
Intelligent and reliable task planning is a core capability for generalized robotics, requiring a descriptive domain representation that sufficiently models all object and state information for the scene. We present CLIMB, a continual learning framework for robot task planning that leverages foundation models and execution feedback to guide domain model construction. CLIMB can build a model from a natural language description, learn non-obvious predicates while solving tasks, and store that information for future problems. We demonstrate the ability of CLIMB to improve performance in common planning environments compared to baseline methods. We also develop the BlocksWorld++ domain, a simulated environment with an easily usable real counterpart, together with a curriculum of tasks with progressing difficulty for evaluating continual learning. Additional details and demonstrations for this system can be found at https://plan-with-climb.github.io/ .
comment: 6 pages, 6 figures
☆ MixEval-X: Any-to-Any Evaluations from Real-World Data Mixtures
Perceiving and generating diverse modalities are crucial for AI models to effectively learn from and engage with real-world signals, necessitating reliable evaluations for their development. We identify two major issues in current evaluations: (1) inconsistent standards, shaped by different communities with varying protocols and maturity levels; and (2) significant query, grading, and generalization biases. To address these, we introduce MixEval-X, the first any-to-any real-world benchmark designed to optimize and standardize evaluations across input and output modalities. We propose multi-modal benchmark mixture and adaptation-rectification pipelines to reconstruct real-world task distributions, ensuring evaluations generalize effectively to real-world use cases. Extensive meta-evaluations show our approach effectively aligns benchmark samples with real-world task distributions and the model rankings correlate strongly with that of crowd-sourced real-world evaluations (up to 0.98). We provide comprehensive leaderboards to rerank existing models and organizations and offer insights to enhance understanding of multi-modal evaluations and inform future research.
☆ Supervised Kernel Thinning
The kernel thinning algorithm of Dwivedi & Mackey (2024) provides a better-than-i.i.d. compression of a generic set of points. By generating high-fidelity coresets of size significantly smaller than the input points, KT is known to speed up unsupervised tasks like Monte Carlo integration, uncertainty quantification, and non-parametric hypothesis testing, with minimal loss in statistical accuracy. In this work, we generalize the KT algorithm to speed up supervised learning problems involving kernel methods. Specifically, we combine two classical algorithms--Nadaraya-Watson (NW) regression or kernel smoothing, and kernel ridge regression (KRR)--with KT to provide a quadratic speed-up in both training and inference times. We show how distribution compression with KT in each setting reduces to constructing an appropriate kernel, and introduce the Kernel-Thinned NW and Kernel-Thinned KRR estimators. We prove that KT-based regression estimators enjoy significantly superior computational efficiency over the full-data estimators and improved statistical efficiency over i.i.d. subsampling of the training data. En route, we also provide a novel multiplicative error guarantee for compressing with KT. We validate our design choices with both simulations and real data experiments.
☆ Theory on Score-Mismatched Diffusion Models and Zero-Shot Conditional Samplers
The denoising diffusion model has recently emerged as a powerful generative technique, capable of transforming noise into meaningful data. While theoretical convergence guarantees for diffusion models are well established when the target distribution aligns with the training distribution, practical scenarios often present mismatches. One common case is in zero-shot conditional diffusion sampling, where the target conditional distribution is different from the (unconditional) training distribution. These score-mismatched diffusion models remain largely unexplored from a theoretical perspective. In this paper, we present the first performance guarantee with explicit dimensional dependencies for general score-mismatched diffusion samplers, focusing on target distributions with finite second moments. We show that score mismatches result in an asymptotic distributional bias between the target and sampling distributions, proportional to the accumulated mismatch between the target and training distributions. This result can be directly applied to zero-shot conditional samplers for any conditional model, irrespective of measurement noise. Interestingly, the derived convergence upper bound offers useful guidance for designing a novel bias-optimal zero-shot sampler in linear conditional models that minimizes the asymptotic bias. For such bias-optimal samplers, we further establish convergence guarantees with explicit dependencies on dimension and conditioning, applied to several interesting target distributions, including those with bounded support and Gaussian mixtures. Our findings are supported by numerical studies.
☆ Single-Timescale Multi-Sequence Stochastic Approximation Without Fixed Point Smoothness: Theories and Applications
Stochastic approximation (SA) that involves multiple coupled sequences, known as multiple-sequence SA (MSSA), finds diverse applications in the fields of signal processing and machine learning. However, existing theoretical understandings {of} MSSA are limited: the multi-timescale analysis implies a slow convergence rate, whereas the single-timescale analysis relies on a stringent fixed point smoothness assumption. This paper establishes tighter single-timescale analysis for MSSA, without assuming smoothness of the fixed points. Our theoretical findings reveal that, when all involved operators are strongly monotone, MSSA converges at a rate of $\tilde{\mathcal{O}}(K^{-1})$, where $K$ denotes the total number of iterations. In addition, when all involved operators are strongly monotone except for the main one, MSSA converges at a rate of $\mathcal{O}(K^{-\frac{1}{2}})$. These theoretical findings align with those established for single-sequence SA. Applying these theoretical findings to bilevel optimization and communication-efficient distributed learning offers relaxed assumptions and/or simpler algorithms with performance guarantees, as validated by numerical experiments.
☆ Improved Convergence Rate for Diffusion Probabilistic Models
Score-based diffusion models have achieved remarkable empirical performance in the field of machine learning and artificial intelligence for their ability to generate high-quality new data instances from complex distributions. Improving our understanding of diffusion models, including mainly convergence analysis for such models, has attracted a lot of interests. Despite a lot of theoretical attempts, there still exists significant gap between theory and practice. Towards to close this gap, we establish an iteration complexity at the order of $d^{1/3}\varepsilon^{-2/3}$, which is better than $d^{5/12}\varepsilon^{-1}$, the best known complexity achieved before our work. This convergence analysis is based on a randomized midpoint method, which is first proposed for log-concave sampling (Shen and Lee, 2019), and then extended to diffusion models by Gupta et al. (2024). Our theory accommodates $\varepsilon$-accurate score estimates, and does not require log-concavity on the target distribution. Moreover, the algorithm can also be parallelized to run in only $O(\log^2(d/\varepsilon))$ parallel rounds in a similar way to prior works.
comment: 20 pages
☆ Optimizing Probabilistic Conformal Prediction with Vectorized Non-Conformity Scores
Generative models have shown significant promise in critical domains such as medical diagnosis, autonomous driving, and climate science, where reliable decision-making hinges on accurate uncertainty quantification. While probabilistic conformal prediction (PCP) offers a powerful framework for this purpose, its coverage efficiency -- the size of the uncertainty set -- is limited when dealing with complex underlying distributions and a finite number of generated samples. In this paper, we propose a novel PCP framework that enhances efficiency by first vectorizing the non-conformity scores with ranked samples and then optimizing the shape of the prediction set by varying the quantiles for samples at the same rank. Our method delivers valid coverage while producing discontinuous and more efficient prediction sets, making it particularly suited for high-stakes applications. We demonstrate the effectiveness of our approach through experiments on both synthetic and real-world datasets.
☆ Reducing the Transformer Architecture to a Minimum
Transformers are a widespread and successful model architecture, particularly in Natural Language Processing (NLP) and Computer Vision (CV). The essential innovation of this architecture is the Attention Mechanism, which solves the problem of extracting relevant context information from long sequences in NLP and realistic scenes in CV. A classical neural network component, a Multi-Layer Perceptron (MLP), complements the attention mechanism. Its necessity is frequently justified by its capability of modeling nonlinear relationships. However, the attention mechanism itself is nonlinear through its internal use of similarity measures. A possible hypothesis is that this nonlinearity is sufficient for modeling typical application problems. As the MLPs usually contain the most trainable parameters of the whole model, their omission would substantially reduce the parameter set size. Further components can also be reorganized to reduce the number of parameters. Under some conditions, query and key matrices can be collapsed into a single matrix of the same size. The same is true about value and projection matrices, which can also be omitted without eliminating the substance of the attention mechanism. Initially, the similarity measure was defined asymmetrically, with peculiar properties such as that a token is possibly dissimilar to itself. A possible symmetric definition requires only half of the parameters. We have laid the groundwork by testing widespread CV benchmarks: MNIST and CIFAR-10. The tests have shown that simplified transformer architectures (a) without MLP, (b) with collapsed matrices, and (c) symmetric similarity matrices exhibit similar performance as the original architecture, saving up to 90% of parameters without hurting the classification performance.
comment: 8 pages, to appear in KDIR2024
☆ Movie Gen: A Cast of Media Foundation Models
We present Movie Gen, a cast of foundation models that generates high-quality, 1080p HD videos with different aspect ratios and synchronized audio. We also show additional capabilities such as precise instruction-based video editing and generation of personalized videos based on a user's image. Our models set a new state-of-the-art on multiple tasks: text-to-video synthesis, video personalization, video editing, video-to-audio generation, and text-to-audio generation. Our largest video generation model is a 30B parameter transformer trained with a maximum context length of 73K video tokens, corresponding to a generated video of 16 seconds at 16 frames-per-second. We show multiple technical innovations and simplifications on the architecture, latent spaces, training objectives and recipes, data curation, evaluation protocols, parallelization techniques, and inference optimizations that allow us to reap the benefits of scaling pre-training data, model size, and training compute for training large scale media generation models. We hope this paper helps the research community to accelerate progress and innovation in media generation models. All videos from this paper are available at https://go.fb.me/MovieGenResearchVideos.
☆ Generation through the lens of learning theory
We study generation through the lens of statistical learning theory. First, we abstract and formalize the results of Gold [1967], Angluin [1979, 1980], and Kleinberg and Mullainathan [2024] for language identification/generation in the limit in terms of a binary hypothesis class defined over an abstract instance space. Then, we formalize a different paradigm of generation studied by Kleinberg and Mullainathan [2024], which we call ``uniform generation," and provide a characterization of which hypothesis classes are uniformly generatable. As is standard in statistical learning theory, our characterization is in terms of the finiteness of a new combinatorial dimension we call the Closure dimension. By doing so, we are able to compare generatability with predictability (captured via PAC and online learnability) and show that these two properties of hypothesis classes are \emph{incompatible} - there are classes that are generatable but not predictable and vice versa.
comment: 16 pages
☆ CrystalX: Ultra-Precision Crystal Structure Resolution and Error Correction Using Deep Learning
Atomic structure analysis of crystalline materials is a paramount endeavor in both chemical and material sciences. This sophisticated technique necessitates not only a solid foundation in crystallography but also a profound comprehension of the intricacies of the accompanying software, posing a significant challenge in meeting the rigorous daily demands. For the first time, we confront this challenge head-on by harnessing the power of deep learning for ultra-precise structural analysis at the full-atom level. To validate the performance of the model, named CrystalX, we employed a vast dataset comprising over 50,000 X-ray diffraction measurements derived from authentic experiments, demonstrating performance that is commensurate with human experts and adept at deciphering intricate geometric patterns. Remarkably, CrystalX revealed that even peer-reviewed publications can harbor errors that are stealthy to human scrutiny, yet CrystalX adeptly rectifies them. This deep learning model revolutionizes the time frame for crystal structure analysis, slashing it down to seconds. It has already been successfully applied in the structure analysis of newly discovered compounds in the latest research without human intervention. Overall, CrystalX marks the beginning of a new era in automating routine structural analysis within self-driving laboratories.
☆ On-device Federated Learning in Smartphones for Detecting Depression from Reddit Posts
Depression detection using deep learning models has been widely explored in previous studies, especially due to the large amounts of data available from social media posts. These posts provide valuable information about individuals' mental health conditions and can be leveraged to train models and identify patterns in the data. However, distributed learning approaches have not been extensively explored in this domain. In this study, we adopt Federated Learning (FL) to facilitate decentralized training on smartphones while protecting user data privacy. We train three neural network architectures--GRU, RNN, and LSTM on Reddit posts to detect signs of depression and evaluate their performance under heterogeneous FL settings. To optimize the training process, we leverage a common tokenizer across all client devices, which reduces the computational load. Additionally, we analyze resource consumption and communication costs on smartphones to assess their impact in a real-world FL environment. Our experimental results demonstrate that the federated models achieve comparable performance to the centralized models. This study highlights the potential of FL for decentralized mental health prediction by providing a secure and efficient model training process on edge devices.
comment: 11 pages, 7 figures, Submitted to IEEE
☆ On the Role of Attention Heads in Large Language Model Safety
Large language models (LLMs) achieve state-of-the-art performance on multiple language tasks, yet their safety guardrails can be circumvented, leading to harmful generations. In light of this, recent research on safety mechanisms has emerged, revealing that when safety representations or component are suppressed, the safety capability of LLMs are compromised. However, existing research tends to overlook the safety impact of multi-head attention mechanisms, despite their crucial role in various model functionalities. Hence, in this paper, we aim to explore the connection between standard attention mechanisms and safety capability to fill this gap in the safety-related mechanistic interpretability. We propose a novel metric which tailored for multi-head attention, the Safety Head ImPortant Score (Ships), to assess the individual heads' contributions to model safety. Based on this, we generalize Ships to the dataset level and further introduce the Safety Attention Head AttRibution Algorithm (Sahara) to attribute the critical safety attention heads inside the model. Our findings show that the special attention head has a significant impact on safety. Ablating a single safety head allows aligned model (e.g., Llama-2-7b-chat) to respond to 16 times more harmful queries, while only modifying 0.006% of the parameters, in contrast to the ~ 5% modification required in previous studies. More importantly, we demonstrate that attention heads primarily function as feature extractors for safety and models fine-tuned from the same base model exhibit overlapping safety heads through comprehensive experiments. Together, our attribution approach and findings provide a novel perspective for unpacking the black box of safety mechanisms within large models.
comment: 28 pages, 18 figures, 7 tables
☆ Efficient Function Placement in Virtual Networks: An Online Learning Approach
We propose a model for the virtual function placement problem and several novel algorithms using ideas based on multi-armed bandits. We prove that these algorithms learn the optimal placement policy rapidly, and their regret grows at a rate at most $O( N M \sqrt{T\ln T} )$ while respecting the feasibility constraints with high probability. We show through numerical experiments that those algorithms both have good practical performance and modest computational complexity. Using the proposed acceleration technique, they can be used to learn in large networks where computational power is limited. Our experiments are fully reproducible, and the code is publicly available.
comment: 19 pages
☆ Ab initio nonparametric variable selection for scalable Symbolic Regression with large $p$
Symbolic regression (SR) is a powerful technique for discovering symbolic expressions that characterize nonlinear relationships in data, gaining increasing attention for its interpretability, compactness, and robustness. However, existing SR methods do not scale to datasets with a large number of input variables (referred to as extreme-scale SR), which are common in modern scientific applications. This ``large $p$'' setting, often accompanied by measurement error, leads to slow performance of SR methods and overly complex expressions that are difficult to interpret. To address this scalability challenge, we propose a method called PAN+SR, which combines a key idea of ab initio nonparametric variable selection with SR to efficiently pre-screen large input spaces and reduce search complexity while maintaining accuracy. The use of nonparametric methods eliminates model misspecification, supporting a strategy called parametric-assisted nonparametric (PAN). We also extend SRBench, an open-source benchmarking platform, by incorporating high-dimensional regression problems with various signal-to-noise ratios. Our results demonstrate that PAN+SR consistently enhances the performance of 17 contemporary SR methods, enabling several to achieve state-of-the-art performance on these challenging datasets.
☆ Automated Model Discovery for Tensional Homeostasis: Constitutive Machine Learning in Growth and Remodeling
Soft biological tissues exhibit a tendency to maintain a preferred state of tensile stress, known as tensional homeostasis, which is restored even after external mechanical stimuli. This macroscopic behavior can be described using the theory of kinematic growth, where the deformation gradient is multiplicatively decomposed into an elastic part and a part related to growth and remodeling. Recently, the concept of homeostatic surfaces was introduced to define the state of homeostasis and the evolution equations for inelastic deformations. However, identifying the optimal model and material parameters to accurately capture the macroscopic behavior of inelastic materials can only be accomplished with significant expertise, is often time-consuming, and prone to error, regardless of the specific inelastic phenomenon. To address this challenge, built-in physics machine learning algorithms offer significant potential. In this work, we extend our inelastic Constitutive Artificial Neural Networks (iCANNs) by incorporating kinematic growth and homeostatic surfaces to discover the scalar model equations, namely the Helmholtz free energy and the pseudo potential. The latter describes the state of homeostasis in a smeared sense. We evaluate the ability of the proposed network to learn from experimentally obtained tissue equivalent data at the material point level, assess its predictive accuracy beyond the training regime, and discuss its current limitations when applied at the structural level. Our source code, data, examples, and an implementation of the corresponding material subroutine are made accessible to the public at https://doi.org/10.5281/zenodo.13946282.
comment: 46 pages, 12 figures, 5 tables
☆ Fine-Tuning Discrete Diffusion Models via Reward Optimization with Applications to DNA and Protein Design
Recent studies have demonstrated the strong empirical performance of diffusion models on discrete sequences across domains from natural language to biological sequence generation. For example, in the protein inverse folding task, conditional diffusion models have achieved impressive results in generating natural-like sequences that fold back into the original structure. However, practical design tasks often require not only modeling a conditional distribution but also optimizing specific task objectives. For instance, we may prefer protein sequences with high stability. To address this, we consider the scenario where we have pre-trained discrete diffusion models that can generate natural-like sequences, as well as reward models that map sequences to task objectives. We then formulate the reward maximization problem within discrete diffusion models, analogous to reinforcement learning (RL), while minimizing the KL divergence against pretrained diffusion models to preserve naturalness. To solve this RL problem, we propose a novel algorithm, DRAKES, that enables direct backpropagation of rewards through entire trajectories generated by diffusion models, by making the originally non-differentiable trajectories differentiable using the Gumbel-Softmax trick. Our theoretical analysis indicates that our approach can generate sequences that are both natural-like and yield high rewards. While similar tasks have been recently explored in diffusion models for continuous domains, our work addresses unique algorithmic and theoretical challenges specific to discrete diffusion models, which arise from their foundation in continuous-time Markov chains rather than Brownian motion. Finally, we demonstrate the effectiveness of DRAKES in generating DNA and protein sequences that optimize enhancer activity and protein stability, respectively, important tasks for gene therapies and protein-based therapeutics.
☆ Latent Space Chain-of-Embedding Enables Output-free LLM Self-Evaluation
LLM self-evaluation relies on the LLM's own ability to estimate response correctness, which can greatly improve its deployment reliability. In this research track, we propose the Chain-of-Embedding (CoE) in the latent space to enable LLMs to perform output-free self-evaluation. CoE consists of all progressive hidden states produced during the inference time, which can be treated as the latent thinking path of LLMs. We find that when LLMs respond correctly and incorrectly, their CoE features differ, these discrepancies assist us in estimating LLM response correctness. Experiments in four diverse domains and seven LLMs fully demonstrate the effectiveness of our method. Meanwhile, its label-free design intent without any training and millisecond-level computational cost ensure real-time feedback in large-scale scenarios. More importantly, we provide interesting insights into LLM response correctness from the perspective of hidden state changes inside LLMs.
comment: 33 pages, 18 figures, 12 tables
☆ Scaling Wearable Foundation Models
Wearable sensors have become ubiquitous thanks to a variety of health tracking features. The resulting continuous and longitudinal measurements from everyday life generate large volumes of data; however, making sense of these observations for scientific and actionable insights is non-trivial. Inspired by the empirical success of generative modeling, where large neural networks learn powerful representations from vast amounts of text, image, video, or audio data, we investigate the scaling properties of sensor foundation models across compute, data, and model size. Using a dataset of up to 40 million hours of in-situ heart rate, heart rate variability, electrodermal activity, accelerometer, skin temperature, and altimeter per-minute data from over 165,000 people, we create LSM, a multimodal foundation model built on the largest wearable-signals dataset with the most extensive range of sensor modalities to date. Our results establish the scaling laws of LSM for tasks such as imputation, interpolation and extrapolation, both across time and sensor modalities. Moreover, we highlight how LSM enables sample-efficient downstream learning for tasks like exercise and activity recognition.
☆ Normalizing self-supervised learning for provably reliable Change Point Detection
Change point detection (CPD) methods aim to identify abrupt shifts in the distribution of input data streams. Accurate estimators for this task are crucial across various real-world scenarios. Yet, traditional unsupervised CPD techniques face significant limitations, often relying on strong assumptions or suffering from low expressive power due to inherent model simplicity. In contrast, representation learning methods overcome these drawbacks by offering flexibility and the ability to capture the full complexity of the data without imposing restrictive assumptions. However, these approaches are still emerging in the CPD field and lack robust theoretical foundations to ensure their reliability. Our work addresses this gap by integrating the expressive power of representation learning with the groundedness of traditional CPD techniques. We adopt spectral normalization (SN) for deep representation learning in CPD tasks and prove that the embeddings after SN are highly informative for CPD. Our method significantly outperforms current state-of-the-art methods during the comprehensive evaluation via three standard CPD datasets.
☆ H2OVL-Mississippi Vision Language Models Technical Report
Smaller vision-language models (VLMs) are becoming increasingly important for privacy-focused, on-device applications due to their ability to run efficiently on consumer hardware for processing enterprise commercial documents and images. These models require strong language understanding and visual capabilities to enhance human-machine interaction. To address this need, we present H2OVL-Mississippi, a pair of small VLMs trained on 37 million image-text pairs using 240 hours of compute on 8 x H100 GPUs. H2OVL-Mississippi-0.8B is a tiny model with 0.8 billion parameters that specializes in text recognition, achieving state of the art performance on the Text Recognition portion of OCRBench and surpassing much larger models in this area. Additionally, we are releasing H2OVL-Mississippi-2B, a 2 billion parameter model for general use cases, exhibiting highly competitive metrics across various academic benchmarks. Both models build upon our prior work with H2O-Danube language models, extending their capabilities into the visual domain. We release them under the Apache 2.0 license, making VLMs accessible to everyone, democratizing document AI and visual LLMs.
☆ All models are wrong, some are useful: Model Selection with Limited Labels
With the multitude of pretrained models available thanks to the advancements in large-scale supervised and self-supervised learning, choosing the right model is becoming increasingly pivotal in the machine learning lifecycle. However, much like the training process, choosing the best pretrained off-the-shelf model for raw, unlabeled data is a labor-intensive task. To overcome this, we introduce MODEL SELECTOR, a framework for label-efficient selection of pretrained classifiers. Given a pool of unlabeled target data, MODEL SELECTOR samples a small subset of highly informative examples for labeling, in order to efficiently identify the best pretrained model for deployment on this target dataset. Through extensive experiments, we demonstrate that MODEL SELECTOR drastically reduces the need for labeled data while consistently picking the best or near-best performing model. Across 18 model collections on 16 different datasets, comprising over 1,500 pretrained models, MODEL SELECTOR reduces the labeling cost by up to 94.15% to identify the best model compared to the cost of the strongest baseline. Our results further highlight the robustness of MODEL SELECTOR in model selection, as it reduces the labeling cost by up to 72.41% when selecting a near-best model, whose accuracy is only within 1% of the best model.
Transformer-Based Approaches for Sensor-Based Human Activity Recognition: Opportunities and Challenges
Transformers have excelled in natural language processing and computer vision, paving their way to sensor-based Human Activity Recognition (HAR). Previous studies show that transformers outperform their counterparts exclusively when they harness abundant data or employ compute-intensive optimization algorithms. However, neither of these scenarios is viable in sensor-based HAR due to the scarcity of data in this field and the frequent need to perform training and inference on resource-constrained devices. Our extensive investigation into various implementations of transformer-based versus non-transformer-based HAR using wearable sensors, encompassing more than 500 experiments, corroborates these concerns. We observe that transformer-based solutions pose higher computational demands, consistently yield inferior performance, and experience significant performance degradation when quantized to accommodate resource-constrained devices. Additionally, transformers demonstrate lower robustness to adversarial attacks, posing a potential threat to user trust in HAR.
☆ Towards Satellite Non-IID Imagery: A Spectral Clustering-Assisted Federated Learning Approach
Low Earth orbit (LEO) satellites are capable of gathering abundant Earth observation data (EOD) to enable different Internet of Things (IoT) applications. However, to accomplish an effective EOD processing mechanism, it is imperative to investigate: 1) the challenge of processing the observed data without transmitting those large-size data to the ground because the connection between the satellites and the ground stations is intermittent, and 2) the challenge of processing the non-independent and identically distributed (non-IID) satellite data. In this paper, to cope with those challenges, we propose an orbit-based spectral clustering-assisted clustered federated self-knowledge distillation (OSC-FSKD) approach for each orbit of an LEO satellite constellation, which retains the advantage of FL that the observed data does not need to be sent to the ground. Specifically, we introduce normalized Laplacian-based spectral clustering (NLSC) into federated learning (FL) to create clustered FL in each round to address the challenge resulting from non-IID data. Particularly, NLSC is adopted to dynamically group clients into several clusters based on cosine similarities calculated by model updates. In addition, self-knowledge distillation is utilized to construct each local client, where the most recent updated local model is used to guide current local model training. Experiments demonstrate that the observation accuracy obtained by the proposed method is separately 1.01x, 2.15x, 1.10x, and 1.03x higher than that of pFedSD, FedProx, FedAU, and FedALA approaches using the SAT4 dataset. The proposed method also shows superiority when using other datasets.
comment: 10 pages, 5 figures
☆ Text-Guided Multi-Property Molecular Optimization with a Diffusion Language Model
Molecular optimization (MO) is a crucial stage in drug discovery in which task-oriented generated molecules are optimized to meet practical industrial requirements. Existing mainstream MO approaches primarily utilize external property predictors to guide iterative property optimization. However, learning all molecular samples in the vast chemical space is unrealistic for predictors. As a result, errors and noise are inevitably introduced during property prediction due to the nature of approximation. This leads to discrepancy accumulation, generalization reduction and suboptimal molecular candidates. In this paper, we propose a text-guided multi-property molecular optimization method utilizing transformer-based diffusion language model (TransDLM). TransDLM leverages standardized chemical nomenclature as semantic representations of molecules and implicitly embeds property requirements into textual descriptions, thereby preventing error propagation during diffusion process. Guided by physically and chemically detailed textual descriptions, TransDLM samples and optimizes encoded source molecules, retaining core scaffolds of source molecules and ensuring structural similarities. Moreover, TransDLM enables simultaneous sampling of multiple molecules, making it ideal for scalable, efficient large-scale optimization through distributed computation on web platforms. Furthermore, our approach surpasses state-of-the-art methods in optimizing molecular structural similarity and enhancing chemical properties on the benchmark dataset. The code is available at: https://anonymous.4open.science/r/TransDLM-A901.
☆ Towards Better Performance in Incomplete LDL: Addressing Data Imbalance
Label Distribution Learning (LDL) is a novel machine learning paradigm that addresses the problem of label ambiguity and has found widespread applications. Obtaining complete label distributions in real-world scenarios is challenging, which has led to the emergence of Incomplete Label Distribution Learning (InLDL). However, the existing InLDL methods overlook a crucial aspect of LDL data: the inherent imbalance in label distributions. To address this limitation, we propose \textbf{Incomplete and Imbalance Label Distribution Learning (I\(^2\)LDL)}, a framework that simultaneously handles incomplete labels and imbalanced label distributions. Our method decomposes the label distribution matrix into a low-rank component for frequent labels and a sparse component for rare labels, effectively capturing the structure of both head and tail labels. We optimize the model using the Alternating Direction Method of Multipliers (ADMM) and derive generalization error bounds via Rademacher complexity, providing strong theoretical guarantees. Extensive experiments on 15 real-world datasets demonstrate the effectiveness and robustness of our proposed framework compared to existing InLDL methods.
☆ Sample Compression Hypernetworks: From Generalization Bounds to Meta-Learning NeurIPS 2024
Reconstruction functions are pivotal in sample compression theory, a framework for deriving tight generalization bounds. From a small sample of the training set (the compression set) and an optional stream of information (the message), they recover a predictor previously learned from the whole training set. While usually fixed, we propose to learn reconstruction functions. To facilitate the optimization and increase the expressiveness of the message, we derive a new sample compression generalization bound for real-valued messages. From this theoretical analysis, we then present a new hypernetwork architecture that outputs predictors with tight generalization guarantees when trained using an original meta-learning framework. The results of promising preliminary experiments are then reported.
comment: Accepted at the NeurIPS 2024 workshop on Compression in Machine Learning
☆ Ornstein-Uhlenbeck Adaptation as a Mechanism for Learning in Brains and Machines
Learning is a fundamental property of intelligent systems, observed across biological organisms and engineered systems. While modern intelligent systems typically rely on gradient descent for learning, the need for exact gradients and complex information flow makes its implementation in biological and neuromorphic systems challenging. This has motivated the exploration of alternative learning mechanisms that can operate locally and do not rely on exact gradients. In this work, we introduce a novel approach that leverages noise in the parameters of the system and global reinforcement signals. Using an Ornstein-Uhlenbeck process with adaptive dynamics, our method balances exploration and exploitation during learning, driven by deviations from error predictions, akin to reward prediction error. Operating in continuous time, Orstein-Uhlenbeck adaptation (OUA) is proposed as a general mechanism for learning dynamic, time-evolving environments. We validate our approach across diverse tasks, including supervised learning and reinforcement learning in feedforward and recurrent systems. Additionally, we demonstrate that it can perform meta-learning, adjusting hyper-parameters autonomously. Our results indicate that OUA provides a viable alternative to traditional gradient-based methods, with potential applications in neuromorphic computing. It also hints at a possible mechanism for noise-driven learning in the brain, where stochastic neurotransmitter release may guide synaptic adjustments.
☆ Adaptive and oblivious statistical adversaries are equivalent
We resolve a fundamental question about the ability to perform a statistical task, such as learning, when an adversary corrupts the sample. Such adversaries are specified by the types of corruption they can make and their level of knowledge about the sample. The latter distinguishes between sample-adaptive adversaries which know the contents of the sample when choosing the corruption, and sample-oblivious adversaries, which do not. We prove that for all types of corruptions, sample-adaptive and sample-oblivious adversaries are \emph{equivalent} up to polynomial factors in the sample size. This resolves the main open question introduced by \cite{BLMT22} and further explored in \cite{CHLLN23}. Specifically, consider any algorithm $A$ that solves a statistical task even when a sample-oblivious adversary corrupts its input. We show that there is an algorithm $A'$ that solves the same task when the corresponding sample-adaptive adversary corrupts its input. The construction of $A'$ is simple and maintains the computational efficiency of $A$: It requests a polynomially larger sample than $A$ uses and then runs $A$ on a uniformly random subsample. One of our main technical tools is a new structural result relating two distributions defined on sunflowers which may be of independent interest.
☆ Generative Adversarial Synthesis of Radar Point Cloud Scenes
For the validation and verification of automotive radars, datasets of realistic traffic scenarios are required, which, how ever, are laborious to acquire. In this paper, we introduce radar scene synthesis using GANs as an alternative to the real dataset acquisition and simulation-based approaches. We train a PointNet++ based GAN model to generate realistic radar point cloud scenes and use a binary classifier to evaluate the performance of scenes generated using this model against a test set of real scenes. We demonstrate that our GAN model achieves similar performance (~87%) to the real scenes test set.
comment: ICMIM 2024; 7th IEEE MTT Conference
☆ PORTAL: Scalable Tabular Foundation Models via Content-Specific Tokenization NeurIPS 2024
Self-supervised learning on tabular data seeks to apply advances from natural language and image domains to the diverse domain of tables. However, current techniques often struggle with integrating multi-domain data and require data cleaning or specific structural requirements, limiting the scalability of pre-training datasets. We introduce PORTAL (Pretraining One-Row-at-a-Time for All tabLes), a framework that handles various data modalities without the need for cleaning or preprocessing. This simple yet powerful approach can be effectively pre-trained on online-collected datasets and fine-tuned to match state-of-the-art methods on complex classification and regression tasks. This work offers a practical advancement in self-supervised learning for large-scale tabular data.
comment: Accepted at Table Representation Learning Workshop at NeurIPS 2024
☆ CERES: Critical-Event Reconstruction via Temporal Scene Graph Completion
This paper proposes a method for on-demand scenario generation in simulation, grounded on real-world data. Evaluating the behaviour of Autonomous Vehicles (AVs) in both safety-critical and regular scenarios is essential for assessing their robustness before real-world deployment. By integrating scenarios derived from real-world datasets into the simulation, we enhance the plausibility and validity of testing sets. This work introduces a novel approach that employs temporal scene graphs to capture evolving spatiotemporal relationships among scene entities from a real-world dataset, enabling the generation of dynamic scenarios in simulation through Graph Neural Networks (GNNs). User-defined action and criticality conditioning are used to ensure flexible, tailored scenario creation. Our model significantly outperforms the benchmarks in accurately predicting links corresponding to the requested scenarios. We further evaluate the validity and compatibility of our generated scenarios in an off-the-shelf simulator.
comment: 7 pages, 8 figures
☆ MathGAP: Out-of-Distribution Evaluation on Problems with Arbitrarily Complex Proofs
Large language models (LLMs) can solve arithmetic word problems with high accuracy, but little is known about how well they generalize to problems that are more complex than the ones on which they have been trained. Empirical investigations of such questions are impeded by two major flaws of current evaluations: (i) much of the evaluation data is contaminated, in the sense that it has already been seen during training, and (ii) benchmark datasets do not capture how problem proofs may be arbitrarily complex in various ways. As a step towards addressing these issues, we present a framework for evaluating LLMs on problems that have arbitrarily complex arithmetic proofs, called MathGAP. MathGAP generates problems that follow fixed proof specifications -- along with chain-of-thought reasoning annotations -- enabling systematic studies on generalization with respect to arithmetic proof complexity. We apply MathGAP to analyze how in-context learning interacts with generalization to problems that have more complex proofs. We find that among the models tested, most show a significant decrease in performance as proofs get deeper and wider. This effect is more pronounced in complex, nonlinear proof structures, which are challenging even for GPT-4o. Surprisingly, providing in-context examples from the same distribution as the test set is not always beneficial for performance. In particular, zero-shot prompting as well as demonstrating a diverse range of examples that are less complex than the test data sometimes yield similar or higher accuracies.
comment: Preprint
☆ Integrating Large Language Models and Reinforcement Learning for Non-Linear Reasoning
Large Language Models (LLMs) were shown to struggle with long-term planning, which may be caused by the limited way in which they explore the space of possible solutions. We propose an architecture where a Reinforcement Learning (RL) Agent guides an LLM's space exploration: (1) the Agent has access to domain-specific information, and can therefore make decisions about the quality of candidate solutions based on specific and relevant metrics, which were not explicitly considered by the LLM's training objective; (2) the LLM can focus on generating immediate next steps, without the need for long-term planning. We allow non-linear reasoning by exploring alternative paths and backtracking. We evaluate this architecture on the program equivalence task, and compare it against Chain of Thought (CoT) and Tree of Thoughts (ToT). We assess both the downstream task, denoting the binary classification, and the intermediate reasoning steps. Our approach compares positively against CoT and ToT.
☆ SAda-Net: A Self-Supervised Adaptive Stereo Estimation CNN For Remote Sensing Image Data ICPR2024
Stereo estimation has made many advancements in recent years with the introduction of deep-learning. However the traditional supervised approach to deep-learning requires the creation of accurate and plentiful ground-truth data, which is expensive to create and not available in many situations. This is especially true for remote sensing applications, where there is an excess of available data without proper ground truth. To tackle this problem, we propose a self-supervised CNN with self-improving adaptive abilities. In the first iteration, the created disparity map is inaccurate and noisy. Leveraging the left-right consistency check, we get a sparse but more accurate disparity map which is used as an initial pseudo ground-truth. This pseudo ground-truth is then adapted and updated after every epoch in the training step of the network. We use the sum of inconsistent points in order to track the network convergence. The code for our method is publicly available at: https://github.com/thedodo/SAda-Net}{https://github.com/thedodo/SAda-Net
comment: Will be presented at ICPR2024 in December 2024 in Kolkata, India
☆ Enhancing Text Generation in Joint NLG/NLU Learning Through Curriculum Learning, Semi-Supervised Training, and Advanced Optimization Techniques
Text generation is the automated process of producing written or spoken language using computational methods. It involves generating coherent and contextually relevant text based on predefined rules or learned patterns. However, challenges in text generation arise from maintaining coherence, ensuring diversity and creativity, and avoiding biases or inappropriate content. This research paper developed a novel approach to improve text generation in the context of joint Natural Language Generation (NLG) and Natural Language Understanding (NLU) learning. The data is prepared by gathering and preprocessing annotated datasets, including cleaning, tokenization, stemming, and stop-word removal. Feature extraction techniques such as POS tagging, Bag of words, and Term Frequency-Inverse Document Frequency (TF-IDF) are applied. Transformer-based encoders and decoders, capturing long range dependencies and improving source-target sequence modelling. Pre-trained language models like Optimized BERT are incorporated, along with a Hybrid Redfox Artificial Hummingbird Algorithm (HRAHA). Reinforcement learning with policy gradient techniques, semi-supervised training, improved attention mechanisms, and differentiable approximations like straight-through Gumbel SoftMax estimator are employed to fine-tune the models and handle complex linguistic tasks effectively. The proposed model is implemented using Python.
☆ Deep Reinforcement Learning for Online Optimal Execution Strategies
This paper tackles the challenge of learning non-Markovian optimal execution strategies in dynamic financial markets. We introduce a novel actor-critic algorithm based on Deep Deterministic Policy Gradient (DDPG) to address this issue, with a focus on transient price impact modeled by a general decay kernel. Through numerical experiments with various decay kernels, we show that our algorithm successfully approximates the optimal execution strategy. Additionally, the proposed algorithm demonstrates adaptability to evolving market conditions, where parameters fluctuate over time. Our findings also show that modern reinforcement learning algorithms can provide a solution that reduces the need for frequent and inefficient human intervention in optimal execution tasks.
☆ Novelty-based Sample Reuse for Continuous Robotics Control
In reinforcement learning, agents collect state information and rewards through environmental interactions, essential for policy refinement. This process is notably time-consuming, especially in complex robotic simulations and real-world applications. Traditional algorithms usually re-engage with the environment after processing a single batch of samples, thereby failing to fully capitalize on historical data. However, frequently observed states, with reliable value estimates, require minimal updates; in contrast, rare observed states necessitate more intensive updates for achieving accurate value estimations. To address uneven sample utilization, we propose Novelty-guided Sample Reuse (NSR). NSR provides extra updates for infrequent, novel states and skips additional updates for frequent states, maximizing sample use before interacting with the environment again. Our experiments show that NSR improves the convergence rate and success rate of algorithms without significantly increasing time consumption. Our code is publicly available at https://github.com/ppksigs/NSR-DDPG-HER.
☆ Seeing Through VisualBERT: A Causal Adventure on Memetic Landscapes EMNLP
Detecting offensive memes is crucial, yet standard deep neural network systems often remain opaque. Various input attribution-based methods attempt to interpret their behavior, but they face challenges with implicitly offensive memes and non-causal attributions. To address these issues, we propose a framework based on a Structural Causal Model (SCM). In this framework, VisualBERT is trained to predict the class of an input meme based on both meme input and causal concepts, allowing for transparent interpretation. Our qualitative evaluation demonstrates the framework's effectiveness in understanding model behavior, particularly in determining whether the model was right due to the right reason, and in identifying reasons behind misclassification. Additionally, quantitative analysis assesses the significance of proposed modelling choices, such as de-confounding, adversarial learning, and dynamic routing, and compares them with input attribution methods. Surprisingly, we find that input attribution methods do not guarantee causality within our framework, raising questions about their reliability in safety-critical applications. The project page is at: https://newcodevelop.github.io/causality_adventure/
comment: Accepted at EMNLP Findings 2024
☆ Interpreting Temporal Graph Neural Networks with Koopman Theory
Spatiotemporal graph neural networks (STGNNs) have shown promising results in many domains, from forecasting to epidemiology. However, understanding the dynamics learned by these models and explaining their behaviour is significantly more complex than for models dealing with static data. Inspired by Koopman theory, which allows a simpler description of intricate, nonlinear dynamical systems, we introduce an explainability approach for temporal graphs. We present two methods to interpret the STGNN's decision process and identify the most relevant spatial and temporal patterns in the input for the task at hand. The first relies on dynamic mode decomposition (DMD), a Koopman-inspired dimensionality reduction method. The second relies on sparse identification of nonlinear dynamics (SINDy), a popular method for discovering governing equations, which we use for the first time as a general tool for explainability. We show how our methods can correctly identify interpretable features such as infection times and infected nodes in the context of dissemination processes.
☆ Truncating Trajectories in Monte Carlo Policy Evaluation: an Adaptive Approach
Policy evaluation via Monte Carlo (MC) simulation is at the core of many MC Reinforcement Learning (RL) algorithms (e.g., policy gradient methods). In this context, the designer of the learning system specifies an interaction budget that the agent usually spends by collecting trajectories of fixed length within a simulator. However, is this data collection strategy the best option? To answer this question, in this paper, we propose as a quality index a surrogate of the mean squared error of a return estimator that uses trajectories of different lengths, i.e., \emph{truncated}. Specifically, this surrogate shows the sub-optimality of the fixed-length trajectory schedule. Furthermore, it suggests that adaptive data collection strategies that spend the available budget sequentially can allocate a larger portion of transitions in timesteps in which more accurate sampling is required to reduce the error of the final estimate. Building on these findings, we present an adaptive algorithm called Robust and Iterative Data collection strategy Optimization (RIDO). The main intuition behind RIDO is to split the available interaction budget into mini-batches. At each round, the agent determines the most convenient schedule of trajectories that minimizes an empirical and robust version of the surrogate of the estimator's error. After discussing the theoretical properties of our method, we conclude by assessing its performance across multiple domains. Our results show that RIDO can adapt its trajectory schedule toward timesteps where more sampling is required to increase the quality of the final estimation.
☆ Progressive Mixed-Precision Decoding for Efficient LLM Inference
In spite of the great potential of large language models (LLMs) across various tasks, their deployment on resource-constrained devices remains challenging due to their excessive computational and memory demands. Quantization has emerged as an effective solution by storing weights in reduced precision. However, utilizing low precisions (i.e.~2/3-bit) to substantially alleviate the memory-boundedness of LLM decoding, still suffers from prohibitive performance drop. In this work, we argue that existing approaches fail to explore the diversity in computational patterns, redundancy, and sensitivity to approximations of the different phases of LLM inference, resorting to a uniform quantization policy throughout. Instead, we propose a novel phase-aware method that selectively allocates precision during different phases of LLM inference, achieving both strong context extraction during prefill and efficient memory bandwidth utilization during decoding. To further address the memory-boundedness of the decoding phase, we introduce Progressive Mixed-Precision Decoding (PMPD), a technique that enables the gradual lowering of precision deeper in the generated sequence, together with a spectrum of precision-switching schedulers that dynamically drive the precision-lowering decisions in either task-adaptive or prompt-adaptive manner. Extensive evaluation across diverse language tasks shows that when targeting Nvidia GPUs, PMPD achieves 1.4$-$12.2$\times$ speedup in matrix-vector multiplications over fp16 models, while when targeting an LLM-optimized NPU, our approach delivers a throughput gain of 3.8$-$8.0$\times$ over fp16 models and up to 1.54$\times$ over uniform quantization approaches while preserving the output quality.
☆ Breaking the Manual Annotation Bottleneck: Creating a Comprehensive Legal Case Criticality Dataset through Semi-Automated Labeling
Predicting case criticality helps legal professionals in the court system manage large volumes of case law. This paper introduces the Criticality Prediction dataset, a new resource for evaluating the potential influence of Swiss Federal Supreme Court decisions on future jurisprudence. Unlike existing approaches that rely on resource-intensive manual annotations, we semi-automatically derive labels leading to a much larger dataset than otherwise possible. Our dataset features a two-tier labeling system: (1) the LD-Label, which identifies cases published as Leading Decisions (LD), and (2) the Citation-Label, which ranks cases by their citation frequency and recency. This allows for a more nuanced evaluation of case importance. We evaluate several multilingual models, including fine-tuned variants and large language models, and find that fine-tuned models consistently outperform zero-shot baselines, demonstrating the need for task-specific adaptation. Our contributions include the introduction of this task and the release of a multilingual dataset to the research community.
☆ Unlocking Legal Knowledge: A Multilingual Dataset for Judicial Summarization in Switzerland
Legal research is a time-consuming task that most lawyers face on a daily basis. A large part of legal research entails looking up relevant caselaw and bringing it in relation to the case at hand. Lawyers heavily rely on summaries (also called headnotes) to find the right cases quickly. However, not all decisions are annotated with headnotes and writing them is time-consuming. Automated headnote creation has the potential to make hundreds of thousands of decisions more accessible for legal research in Switzerland alone. To kickstart this, we introduce the Swiss Leading Decision Summarization ( SLDS) dataset, a novel cross-lingual resource featuring 18K court rulings from the Swiss Federal Supreme Court (SFSC), in German, French, and Italian, along with German headnotes. We fine-tune and evaluate three mT5 variants, along with proprietary models. Our analysis highlights that while proprietary models perform well in zero-shot and one-shot settings, fine-tuned smaller models still provide a strong competitive edge. We publicly release the dataset to facilitate further research in multilingual legal summarization and the development of assistive technologies for legal professionals
☆ Fast Estimation of Partial Dependence Functions using Trees
Many existing interpretation methods are based on Partial Dependence (PD) functions that, for a pre-trained machine learning model, capture how a subset of the features affects the predictions by averaging over the remaining features. Notable methods include Shapley additive explanations (SHAP) which computes feature contributions based on a game theoretical interpretation and PD plots (i.e., 1-dim PD functions) that capture average marginal main effects. Recent work has connected these approaches using a functional decomposition and argues that SHAP values can be misleading since they merge main and interaction effects into a single local effect. A major advantage of SHAP compared to other PD-based interpretations, however, has been the availability of fast estimation techniques, such as \texttt{TreeSHAP}. In this paper, we propose a new tree-based estimator, \texttt{FastPD}, which efficiently estimates arbitrary PD functions. We show that \texttt{FastPD} consistently estimates the desired population quantity -- in contrast to path-dependent \texttt{TreeSHAP} which is inconsistent when features are correlated. For moderately deep trees, \texttt{FastPD} improves the complexity of existing methods from quadratic to linear in the number of observations. By estimating PD functions for arbitrary feature subsets, \texttt{FastPD} can be used to extract PD-based interpretations such as SHAP, PD plots and higher order interaction effects.
☆ Parameter-efficient Adaptation of Multilingual Multimodal Models for Low-resource ASR
Automatic speech recognition (ASR) for low-resource languages remains a challenge due to the scarcity of labeled training data. Parameter-efficient fine-tuning and text-only adaptation are two popular methods that have been used to address such low-resource settings. In this work, we investigate how these techniques can be effectively combined using a multilingual multimodal model like SeamlessM4T. Multimodal models are able to leverage unlabeled text via text-only adaptation with further parameter-efficient ASR fine-tuning, thus boosting ASR performance. We also show cross-lingual transfer from a high-resource language, achieving up to a relative 17% WER reduction over a baseline in a zero-shot setting without any labeled speech.
☆ Similarity-Dissimilarity Loss with Supervised Contrastive Learning for Multi-label Classification
Supervised contrastive learning has been explored in making use of label information for multi-label classification, but determining positive samples in multi-label scenario remains challenging. Previous studies have examined strategies for identifying positive samples, considering label overlap proportion between anchors and samples. However, they ignore various relations between given anchors and samples, as well as how to dynamically adjust the weights in contrastive loss functions based on different relations, leading to great ambiguity. In this paper, we introduce five distinct relations between multi-label samples and propose a Similarity-Dissimilarity Loss with contrastive learning for multi-label classification. Our loss function re-weights the loss by computing the similarity and dissimilarity between positive samples and a given anchor based on the introduced relations. We mainly conduct experiments for multi-label text classification on MIMIC datasets, then further extend the evaluation on MS-COCO. The Experimental results show that our proposed loss effectively improves the performance on all encoders under supervised contrastive learning paradigm, demonstrating its effectiveness and robustness.
☆ Solving Prior Distribution Mismatch in Diffusion Models via Optimal Transport
In recent years, the knowledge surrounding diffusion models(DMs) has grown significantly, though several theoretical gaps remain. Particularly noteworthy is prior error, defined as the discrepancy between the termination distribution of the forward process and the initial distribution of the reverse process. To address these deficiencies, this paper explores the deeper relationship between optimal transport(OT) theory and DMs with discrete initial distribution. Specifically, we demonstrate that the two stages of DMs fundamentally involve computing time-dependent OT. However, unavoidable prior error result in deviation during the reverse process under quadratic transport cost. By proving that as the diffusion termination time increases, the probability flow exponentially converges to the gradient of the solution to the classical Monge-Amp\`ere equation, we establish a vital link between these fields. Therefore, static OT emerges as the most intrinsic single-step method for bridging this theoretical potential gap. Additionally, we apply these insights to accelerate sampling in both unconditional and conditional generation scenarios. Experimental results across multiple image datasets validate the effectiveness of our approach.
☆ Partially Trained Graph Convolutional Networks Resist Oversmoothing
In this work we investigate an observation made by Kipf \& Welling, who suggested that untrained GCNs can generate meaningful node embeddings. In particular, we investigate the effect of training only a single layer of a GCN, while keeping the rest of the layers frozen. We propose a basis on which the effect of the untrained layers and their contribution to the generation of embeddings can be predicted. Moreover, we show that network width influences the dissimilarity of node embeddings produced after the initial node features pass through the untrained part of the model. Additionally, we establish a connection between partially trained GCNs and oversmoothing, showing that they are capable of reducing it. We verify our theoretical results experimentally and show the benefits of using deep networks that resist oversmoothing, in a ``cold start'' scenario, where there is a lack of feature information for unlabeled nodes.
☆ RAMPA: Robotic Augmented Reality for Machine Programming and Automation
As robotics continue to enter various sectors beyond traditional industrial applications, the need for intuitive robot training and interaction systems becomes increasingly more important. This paper introduces Robotic Augmented Reality for Machine Programming (RAMPA), a system that utilizes the capabilities of state-of-the-art and commercially available AR headsets, e.g., Meta Quest 3, to facilitate the application of Programming from Demonstration (PfD) approaches on industrial robotic arms, such as Universal Robots UR10. Our approach enables in-situ data recording, visualization, and fine-tuning of skill demonstrations directly within the user's physical environment. RAMPA addresses critical challenges of PfD, such as safety concerns, programming barriers, and the inefficiency of collecting demonstrations on the actual hardware. The performance of our system is evaluated against the traditional method of kinesthetic control in teaching three different robotic manipulation tasks and analyzed with quantitative metrics, measuring task performance and completion time, trajectory smoothness, system usability, user experience, and task load using standardized surveys. Our findings indicate a substantial advancement in how robotic tasks are taught and refined, promising improvements in operational safety, efficiency, and user engagement in robotic programming.
comment: This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible
☆ MoR: Mixture of Ranks for Low-Rank Adaptation Tuning
Low-Rank Adaptation (LoRA) drives research to align its performance with full fine-tuning. However, significant challenges remain: (1) Simply increasing the rank size of LoRA does not effectively capture high-rank information, which leads to a performance bottleneck.(2) MoE-style LoRA methods substantially increase parameters and inference latency, contradicting the goals of efficient fine-tuning and ease of application. To address these challenges, we introduce Mixture of Ranks (MoR), which learns rank-specific information for different tasks based on input and efficiently integrates multi-rank information. We firstly propose a new framework that equates the integration of multiple LoRAs to expanding the rank of LoRA. Moreover, we hypothesize that low-rank LoRA already captures sufficient intrinsic information, and MoR can derive high-rank information through mathematical transformations of the low-rank components. Thus, MoR can reduces the learning difficulty of LoRA and enhances its multi-task capabilities. MoR achieves impressive results, with MoR delivering a 1.31\% performance improvement while using only 93.93\% of the parameters compared to baseline methods.
comment: 11 pages, 7 figures
☆ Predicting Breast Cancer Survival: A Survival Analysis Approach Using Log Odds and Clinical Variables
Breast cancer remains a significant global health challenge, with prognosis and treatment decisions largely dependent on clinical characteristics. Accurate prediction of patient outcomes is crucial for personalized treatment strategies. This study employs survival analysis techniques, including Cox proportional hazards and parametric survival models, to enhance the prediction of the log odds of survival in breast cancer patients. Clinical variables such as tumor size, hormone receptor status, HER2 status, age, and treatment history were analyzed to assess their impact on survival outcomes. Data from 1557 breast cancer patients were obtained from a publicly available dataset provided by the University College Hospital, Ibadan, Nigeria. This dataset was preprocessed and analyzed using both univariate and multivariate approaches to evaluate survival outcomes. Kaplan-Meier survival curves were generated to visualize survival probabilities, while the Cox proportional hazards model identified key risk factors influencing mortality. The results showed that older age, larger tumor size, and HER2-positive status were significantly associated with an increased risk of mortality. In contrast, estrogen receptor positivity and breast-conserving surgery were linked to better survival outcomes. The findings suggest that integrating these clinical variables into predictive models improvesthe accuracy of survival predictions, helping to identify high-risk patients who may benefit from more aggressive interventions. This study demonstrates the potential of survival analysis in optimizing breast cancer care, particularly in resource-limited settings. Future research should focus on integrating genomic data and real-world clinical outcomes to further refine these models.
comment: 17 pages
☆ A Self-Constructing Multi-Expert Fuzzy System for High-dimensional Data Classification
Fuzzy Neural Networks (FNNs) are effective machine learning models for classification tasks, commonly based on the Takagi-Sugeno-Kang (TSK) fuzzy system. However, when faced with high-dimensional data, especially with noise, FNNs encounter challenges such as vanishing gradients, excessive fuzzy rules, and limited access to prior knowledge. To address these challenges, we propose a novel fuzzy system, the Self-Constructing Multi-Expert Fuzzy System (SOME-FS). It combines two learning strategies: mixed structure learning and multi-expert advanced learning. The former enables each base classifier to effectively determine its structure without requiring prior knowledge, while the latter tackles the issue of vanishing gradients by enabling each rule to focus on its local region, thereby enhancing the robustness of the fuzzy classifiers. The overall ensemble architecture enhances the stability and prediction performance of the fuzzy system. Our experimental results demonstrate that the proposed SOME-FS is effective in high-dimensional tabular data, especially in dealing with uncertainty. Moreover, our stable rule mining process can identify concise and core rules learned by the SOME-FS.
☆ Learning Counterfactual Distributions via Kernel Nearest Neighbors
Consider a setting with multiple units (e.g., individuals, cohorts, geographic locations) and outcomes (e.g., treatments, times, items), where the goal is to learn a multivariate distribution for each unit-outcome entry, such as the distribution of a user's weekly spend and engagement under a specific mobile app version. A common challenge is the prevalence of missing not at random data, where observations are available only for certain unit-outcome combinations and the observation availability can be correlated with the properties of distributions themselves, i.e., there is unobserved confounding. An additional challenge is that for any observed unit-outcome entry, we only have a finite number of samples from the underlying distribution. We tackle these two challenges by casting the problem into a novel distributional matrix completion framework and introduce a kernel based distributional generalization of nearest neighbors to estimate the underlying distributions. By leveraging maximum mean discrepancies and a suitable factor model on the kernel mean embeddings of the underlying distributions, we establish consistent recovery of the underlying distributions even when data is missing not at random and positivity constraints are violated. Furthermore, we demonstrate that our nearest neighbors approach is robust to heteroscedastic noise, provided we have access to two or more measurements for the observed unit-outcome entries, a robustness not present in prior works on nearest neighbors with single measurements.
comment: 33 pages, 2 figures
☆ Data-Augmented Predictive Deep Neural Network: Enhancing the extrapolation capabilities of non-intrusive surrogate models
Numerically solving a large parametric nonlinear dynamical system is challenging due to its high complexity and the high computational costs. In recent years, machine-learning-aided surrogates are being actively researched. However, many methods fail in accurately generalizing in the entire time interval $[0, T]$, when the training data is available only in a training time interval $[0, T_0]$, with $T_0
☆ Addressing Heterogeneity and Heterophily in Graphs: A Heterogeneous Heterophilic Spectral Graph Neural Network
Graph Neural Networks (GNNs) have garnered significant scholarly attention for their powerful capabilities in modeling graph structures. Despite this, two primary challenges persist: heterogeneity and heterophily. Existing studies often address heterogeneous and heterophilic graphs separately, leaving a research gap in the understanding of heterogeneous heterophilic graphs-those that feature diverse node or relation types with dissimilar connected nodes. To address this gap, we investigate the application of spectral graph filters within heterogeneous graphs. Specifically, we propose a Heterogeneous Heterophilic Spectral Graph Neural Network (H2SGNN), which employs a dual-module approach: local independent filtering and global hybrid filtering. The local independent filtering module applies polynomial filters to each subgraph independently to adapt to different homophily, while the global hybrid filtering module captures interactions across different subgraphs. Extensive empirical evaluations on four real-world datasets demonstrate the superiority of H2SGNN compared to state-of-the-art methods.
☆ Statistical testing on generative AI anomaly detection tools in Alzheimer's Disease diagnosis
Alzheimer's Disease is challenging to diagnose due to our limited understanding of its mechanism and large heterogeneity among patients. Neurodegeneration is studied widely as a biomarker for clinical diagnosis, which can be measured from time series MRI progression. On the other hand, generative AI has shown promise in anomaly detection in medical imaging and used for tasks including tumor detection. However, testing the reliability of such data-driven methods is non-trivial due to the issue of double-dipping in hypothesis testing. In this work, we propose to solve this issue with selective inference and develop a reliable generative AI method for Alzheimer's prediction. We show that compared to traditional statistical methods with highly inflated p-values, selective inference successfully controls the false discovery rate under the desired alpha level while retaining statistical power. In practice, our pipeline could assist clinicians in Alzheimer's diagnosis and early intervention.
☆ Remember, Retrieve and Generate: Understanding Infinite Visual Concepts as Your Personalized Assistant
The development of large language models (LLMs) has significantly enhanced the capabilities of multimodal LLMs (MLLMs) as general assistants. However, lack of user-specific knowledge still restricts their application in human's daily life. In this paper, we introduce the Retrieval Augmented Personalization (RAP) framework for MLLMs' personalization. Starting from a general MLLM, we turn it into a personalized assistant in three steps. (a) Remember: We design a key-value database to store user-related information, e.g., user's name, avatar and other attributes. (b) Retrieve: When the user initiates a conversation, RAP will retrieve relevant information from the database using a multimodal retriever. (c) Generate: The input query and retrieved concepts' information are fed into MLLMs to generate personalized, knowledge-augmented responses. Unlike previous methods, RAP allows real-time concept editing via updating the external database. To further improve generation quality and alignment with user-specific information, we design a pipeline for data collection and create a specialized dataset for personalized training of MLLMs. Based on the dataset, we train a series of MLLMs as personalized multimodal assistants. By pretraining on large-scale dataset, RAP-MLLMs can generalize to infinite visual concepts without additional finetuning. Our models demonstrate outstanding flexibility and generation quality across a variety of tasks, such as personalized image captioning, question answering and visual recognition. The code, data and models are available at https://github.com/Hoar012/RAP-MLLM.
☆ Representation Learning of Structured Data for Medical Foundation Models NeurIPS 2024
Large Language Models (LLMs) have demonstrated remarkable performance across various domains, including healthcare. However, their ability to effectively represent structured non-textual data, such as the alphanumeric medical codes used in records like ICD-10 or SNOMED-CT, is limited and has been particularly exposed in recent research. This paper examines the challenges LLMs face in processing medical codes due to the shortcomings of current tokenization methods. As a result, we introduce the UniStruct architecture to design a multimodal medical foundation model of unstructured text and structured data, which addresses these challenges by adapting subword tokenization techniques specifically for the structured medical codes. Our approach is validated through model pre-training on both an extensive internal medical database and a public repository of structured medical records. Trained on over 1 billion tokens on the internal medical database, the proposed model achieves up to a 23% improvement in evaluation metrics, with around 2% gain attributed to our proposed tokenization. Additionally, when evaluated on the EHRSHOT public benchmark with a 1/1000 fraction of the pre-training data, the UniStruct model improves performance on over 42% of the downstream tasks. Our approach not only enhances the representation and generalization capabilities of patient-centric models but also bridges a critical gap in representation learning models' ability to handle complex structured medical data, alongside unstructured text.
comment: NeurIPS 2024 Workshop on Unifying Representations in Neural Models (UniReps 2024)
☆ Do LLMs Overcome Shortcut Learning? An Evaluation of Shortcut Challenges in Large Language Models
Large Language Models (LLMs) have shown remarkable capabilities in various natural language processing tasks. However, LLMs may rely on dataset biases as shortcuts for prediction, which can significantly impair their robustness and generalization capabilities. This paper presents Shortcut Suite, a comprehensive test suite designed to evaluate the impact of shortcuts on LLMs' performance, incorporating six shortcut types, five evaluation metrics, and four prompting strategies. Our extensive experiments yield several key findings: 1) LLMs demonstrate varying reliance on shortcuts for downstream tasks, significantly impairing their performance. 2) Larger LLMs are more likely to utilize shortcuts under zero-shot and few-shot in-context learning prompts. 3) Chain-of-thought prompting notably reduces shortcut reliance and outperforms other prompting strategies, while few-shot prompts generally underperform compared to zero-shot prompts. 4) LLMs often exhibit overconfidence in their predictions, especially when dealing with datasets that contain shortcuts. 5) LLMs generally have a lower explanation quality in shortcut-laden datasets, with errors falling into three types: distraction, disguised comprehension, and logical fallacy. Our findings offer new insights for evaluating robustness and generalization in LLMs and suggest potential directions for mitigating the reliance on shortcuts. The code is available at \url {https://github.com/yyhappier/ShortcutSuite.git}.
☆ Limits to scalable evaluation at the frontier: LLM as Judge won't beat twice the data
High quality annotations are increasingly a bottleneck in the explosively growing machine learning ecosystem. Scalable evaluation methods that avoid costly annotation have therefore become an important research ambition. Many hope to use strong existing models in lieu of costly labels to provide cheap model evaluations. Unfortunately, this method of using models as judges introduces biases, such as self-preferencing, that can distort model comparisons. An emerging family of debiasing tools promises to fix these issues by using a few high quality labels to debias a large number of model judgments. In this paper, we study how far such debiasing methods, in principle, can go. Our main result shows that when the judge is no more accurate than the evaluated model, no debiasing method can decrease the required amount of ground truth labels by more than half. Our result speaks to the severe limitations of the LLM-as-a-judge paradigm at the evaluation frontier where the goal is to assess newly released models that are possibly better than the judge. Through an empirical evaluation, we demonstrate that the sample size savings achievable in practice are even more modest than what our theoretical limit suggests. Along the way, our work provides new observations about debiasing methods for model evaluation, and points out promising avenues for future work.
comment: 22 pages, 5 figures
☆ DiffImp: Efficient Diffusion Model for Probabilistic Time Series Imputation with Bidirectional Mamba Backbone
Probabilistic time series imputation has been widely applied in real-world scenarios due to its ability to estimate uncertainty of imputation results. Meanwhile, denoising diffusion probabilistic models (DDPMs) have achieved great success in probabilistic time series imputation tasks with its power to model complex distributions. However, current DDPM-based probabilistic time series imputation methodologies are confronted with two types of challenges: 1)~\textit{~The backbone modules of the denoising parts are not capable of achieving sequence modeling with low time complexity.} 2)~\textit{The architecture of denoising modules can not handle the inter-variable and bidirectional dependencies in the time series imputation problem effectively.} To address the first challenge, we integrate the computational efficient state space model, namely Mamba, as the backbone denosing module for DDPMs. To tackle the second challenge, we carefully devise several SSM-based blocks for bidirectional modeling and inter-variable relation understanding. Experimental results demonstrate that our approach can achieve state-of-the-art time series imputation results on multiple datasets, different missing scenarios and missing ratios.
comment: 25 pages, 14 figures
☆ Do LLMs Have Political Correctness? Analyzing Ethical Biases and Jailbreak Vulnerabilities in AI Systems
Although large language models (LLMs) demonstrate impressive proficiency in various tasks, they present potential safety risks, such as `jailbreaks', where malicious inputs can coerce LLMs into generating harmful content. To address these issues, many LLM developers have implemented various safety measures to align these models. This alignment involves several techniques, including data filtering during pre-training, supervised fine-tuning, reinforcement learning from human feedback, and red-teaming exercises. These methods often introduce deliberate and intentional biases similar to Political Correctness (PC) to ensure the ethical behavior of LLMs. In this paper, we delve into the intentional biases injected into LLMs for safety purposes and examine methods to circumvent these safety alignment techniques. Notably, these intentional biases result in a jailbreaking success rate in GPT-4o models that differs by 20% between non-binary and cisgender keywords and by 16% between white and black keywords, even when the other parts of the prompts are identical. We introduce the concept of PCJailbreak, highlighting the inherent risks posed by these safety-induced biases. Additionally, we propose an efficient defense method PCDefense, which prevents jailbreak attempts by injecting defense prompts prior to generation. PCDefense stands as an appealing alternative to Guard Models, such as Llama-Guard, that require additional inference cost after text generation. Our findings emphasize the urgent need for LLM developers to adopt a more responsible approach when designing and implementing safety measures.
☆ Improving Discrete Optimisation Via Decoupled Straight-Through Gumbel-Softmax
Discrete representations play a crucial role in many deep learning architectures, yet their non-differentiable nature poses significant challenges for gradient-based optimization. To address this issue, various gradient estimators have been developed, including the Straight-Through Gumbel-Softmax (ST-GS) estimator, which combines the Straight-Through Estimator (STE) and the Gumbel-based reparameterization trick. However, the performance of ST-GS is highly sensitive to temperature, with its selection often compromising gradient fidelity. In this work, we propose a simple yet effective extension to ST-GS by employing decoupled temperatures for forward and backward passes, which we refer to as "Decoupled ST-GS". We show that our approach significantly enhances the original ST-GS through extensive experiments across multiple tasks and datasets. We further investigate the impact of our method on gradient fidelity from multiple perspectives, including the gradient gap and the bias-variance trade-off of estimated gradients. Our findings contribute to the ongoing effort to improve discrete optimization in deep learning, offering a practical solution that balances simplicity and effectiveness.
☆ Precipitation Nowcasting Using Diffusion Transformer with Causal Attention
Short-term precipitation forecasting remains challenging due to the difficulty in capturing long-term spatiotemporal dependencies. Current deep learning methods fall short in establishing effective dependencies between conditions and forecast results, while also lacking interpretability. To address this issue, we propose a Precipitation Nowcasting Using Diffusion Transformer with Causal Attention model. Our model leverages Transformer and combines causal attention mechanisms to establish spatiotemporal queries between conditional information (causes) and forecast results (results). This design enables the model to effectively capture long-term dependencies, allowing forecast results to maintain strong causal relationships with input conditions over a wide range of time and space. We explore four variants of spatiotemporal information interactions for DTCA, demonstrating that global spatiotemporal labeling interactions yield the best performance. In addition, we introduce a Channel-To-Batch shift operation to further enhance the model's ability to represent complex rainfall dynamics. We conducted experiments on two datasets. Compared to state-of-the-art U-Net-based methods, our approach improved the CSI (Critical Success Index) for predicting heavy precipitation by approximately 15% and 8% respectively, achieving state-of-the-art performance.
☆ Active inference and deep generative modeling for cognitive ultrasound
Ultrasound (US) has the unique potential to offer access to medical imaging to anyone, everywhere. Devices have become ultra-portable and cost-effective, akin to the stethoscope. Nevertheless US image quality and diagnostic efficacy are still highly operator- and patient-dependent. In difficult-to-image patients, image quality is often insufficient for reliable diagnosis. In this paper, we put forth that US imaging systems can be recast as information-seeking agents that engage in reciprocal interactions with their anatomical environment. Such agents autonomously adapt their transmit-receive sequences to fully personalize imaging and actively maximize information gain in-situ. To that end, we will show that the sequence of pulse-echo experiments that a US system performs can be interpreted as a perception-action loop: the action is the data acquisition, probing tissue with acoustic waves and recording reflections at the detection array, and perception is the inference of the anatomical and or functional state, potentially including associated diagnostic quantities. We then equip systems with a mechanism to actively reduce uncertainty and maximize diagnostic value across a sequence of experiments, treating action and perception jointly using Bayesian inference given generative models of the environment and action-conditional pulse-echo observations. Since the representation capacity of the generative models dictates both the quality of inferred anatomical states and the effectiveness of inferred sequences of future imaging actions, we will be greatly leveraging the enormous advances in deep generative modelling that are currently disrupting many fields and society at large. Finally, we show some examples of cognitive, closed-loop, US systems that perform active beamsteering and adaptive scanline selection, based on deep generative models that track anatomical belief states.
☆ Hiformer: Hybrid Frequency Feature Enhancement Inverted Transformer for Long-Term Wind Power Prediction
The increasing severity of climate change necessitates an urgent transition to renewable energy sources, making the large-scale adoption of wind energy crucial for mitigating environmental impact. However, the inherent uncertainty of wind power poses challenges for grid stability, underscoring the need for accurate wind energy prediction models to enable effective power system planning and operation. While many existing studies on wind power prediction focus on short-term forecasting, they often overlook the importance of long-term predictions. Long-term wind power forecasting is essential for effective power grid dispatch and market transactions, as it requires careful consideration of weather features such as wind speed and direction, which directly influence power output. Consequently, methods designed for short-term predictions may lead to inaccurate results and high computational costs in long-term settings. To adress these limitations, we propose a novel approach called Hybrid Frequency Feature Enhancement Inverted Transformer (Hiformer). Hiformer introduces a unique structure that integrates signal decomposition technology with weather feature extraction technique to enhance the modeling of correlations between meteorological conditions and wind power generation. Additionally, Hiformer employs an encoder-only architecture, which reduces the computational complexity associated with long-term wind power forecasting. Compared to the state-of-the-art methods, Hiformer: (i) can improve the prediction accuracy by up to 52.5\%; and (ii) can reduce computational time by up to 68.5\%.
☆ A theoretical perspective on mode collapse in variational inference
While deep learning has expanded the possibilities for highly expressive variational families, the practical benefits of these tools for variational inference (VI) are often limited by the minimization of the traditional Kullback-Leibler objective, which can yield suboptimal solutions. A major challenge in this context is \emph{mode collapse}: the phenomenon where a model concentrates on a few modes of the target distribution during training, despite being statistically capable of expressing them all. In this work, we carry a theoretical investigation of mode collapse for the gradient flow on Gaussian mixture models. We identify the key low-dimensional statistics characterizing the flow, and derive a closed set of low-dimensional equations governing their evolution. Leveraging this compact description, we show that mode collapse is present even in statistically favorable scenarios, and identify two key mechanisms driving it: mean alignment and vanishing weight. Our theoretical findings are consistent with the implementation of VI using normalizing flows, a class of popular generative models, thereby offering practical insights.
☆ LLM-Rank: A Graph Theoretical Approach to Pruning Large Language Models
The evolving capabilities of large language models are accompanied by growing sizes and deployment costs, necessitating effective inference optimisation techniques. We propose a novel pruning method utilising centrality measures from graph theory, reducing both the computational requirements and the memory footprint of these models. Specifically, we devise a method for creating a weighted directed acyclical graph representation of multilayer perceptrons to which we apply a modified version of the weighted PageRank centrality measure to compute node importance scores. In combination with uniform pruning this leads to structured sparsity. We call this pruning method MLPRank. Furthermore we introduce an extension to decoder-only transformer models and call it LLMRank. For both variants we demonstrate a strong performance. With MLPRank on average leading to 6.09 % higher accuracy retention than three popular baselines and 13.42 % with LLMRank compared to two popular baselines.
☆ Fairness-Enhancing Ensemble Classification in Water Distribution Networks
As relevant examples such as the future criminal detection software [1] show, fairness of AI-based and social domain affecting decision support tools constitutes an important area of research. In this contribution, we investigate the applications of AI to socioeconomically relevant infrastructures such as those of water distribution networks (WDNs), where fairness issues have yet to gain a foothold. To establish the notion of fairness in this domain, we propose an appropriate definition of protected groups and group fairness in WDNs as an extension of existing definitions. We demonstrate that typical methods for the detection of leakages in WDNs are unfair in this sense. Further, we thus propose a remedy to increase the fairness which can be applied even to non-differentiable ensemble classification methods as used in this context.
☆ PiLocNet: Physics-informed neural network on 3D localization with rotating point spread function
For the 3D localization problem using point spread function (PSF) engineering, we propose a novel enhancement of our previously introduced localization neural network, LocNet. The improved network is a physics-informed neural network (PINN) that we call PiLocNet. Previous works on the localization problem may be categorized separately into model-based optimization and neural network approaches. Our PiLocNet combines the unique strengths of both approaches by incorporating forward-model-based information into the network via a data-fitting loss term that constrains the neural network to yield results that are physically sensible. We additionally incorporate certain regularization terms from the variational method, which further improves the robustness of the network in the presence of image noise, as we show for the Poisson and Gaussian noise models. This framework accords interpretability to the neural network, and the results we obtain show its superiority. Although the paper focuses on the use of single-lobe rotating PSF to encode the full 3D source location, we expect the method to be widely applicable to other PSFs and imaging problems that are constrained by known forward processes.
comment: 25 pages, 4 figures
☆ SBI-RAG: Enhancing Math Word Problem Solving for Students through Schema-Based Instruction and Retrieval-Augmented Generation NeurIPS'24
Many students struggle with math word problems (MWPs), often finding it difficult to identify key information and select the appropriate mathematical operations.Schema-based instruction (SBI) is an evidence-based strategy that helps students categorize problems based on their structure, improving problem-solving accuracy. Building on this, we propose a Schema-Based Instruction Retrieval-Augmented Generation (SBI-RAG) framework that incorporates a large language model (LLM).Our approach emphasizes step-by-step reasoning by leveraging schemas to guide solution generation. We evaluate its performance on the GSM8K dataset, comparing it with GPT-4 and GPT-3.5 Turbo, and introduce a "reasoning score" metric to assess solution quality. Our findings suggest that SBI-RAG enhances reasoning clarity and problem-solving accuracy, potentially providing educational benefits for students
comment: Accepted to the 4th MATH-AI Workshop at NeurIPS'24
☆ An Online Learning Approach to Prompt-based Selection of Generative Models
Selecting a sample generation scheme from multiple text-based generative models is typically addressed by choosing the model that maximizes an averaged evaluation score. However, this score-based selection overlooks the possibility that different models achieve the best generation performance for different types of text prompts. An online identification of the best generation model for various input prompts can reduce the costs associated with querying sub-optimal models. In this work, we explore the possibility of varying rankings of text-based generative models for different text prompts and propose an online learning framework to predict the best data generation model for a given input prompt. The proposed framework adapts the kernelized contextual bandit (CB) methodology to a CB setting with shared context variables across arms, utilizing the generated data to update a kernel-based function that predicts which model will achieve the highest score for unseen text prompts. Additionally, we apply random Fourier features (RFF) to the kernelized CB algorithm to accelerate the online learning process and establish a $\widetilde{\mathcal{O}}(\sqrt{T})$ regret bound for the proposed RFF-based CB algorithm over T iterations. Our numerical experiments on real and simulated text-to-image and image-to-text generative models show RFF-UCB performs successfully in identifying the best generation model across different sample types.
☆ A Human-in-the-Loop Fairness-Aware Model Selection Framework for Complex Fairness Objective Landscapes
Fairness-aware Machine Learning (FairML) applications are often characterized by complex social objectives and legal requirements, frequently involving multiple, potentially conflicting notions of fairness. Despite the well-known Impossibility Theorem of Fairness and extensive theoretical research on the statistical and socio-technical trade-offs between fairness metrics, many FairML tools still optimize or constrain for a single fairness objective. However, this one-sided optimization can inadvertently lead to violations of other relevant notions of fairness. In this socio-technical and empirical study, we frame fairness as a many-objective (MaO) problem by treating fairness metrics as conflicting objectives. We introduce ManyFairHPO, a human-in-the-loop, fairness-aware model selection framework that enables practitioners to effectively navigate complex and nuanced fairness objective landscapes. ManyFairHPO aids in the identification, evaluation, and balancing of fairness metric conflicts and their related social consequences, leading to more informed and socially responsible model-selection decisions. Through a comprehensive empirical evaluation and a case study on the Law School Admissions problem, we demonstrate the effectiveness of ManyFairHPO in balancing multiple fairness objectives, mitigating risks such as self-fulfilling prophecies, and providing interpretable insights to guide stakeholders in making fairness-aware modeling decisions.
☆ Learning to Route with Confidence Tokens
Large language models (LLMs) have demonstrated impressive performance on several tasks and are increasingly deployed in real-world applications. However, especially in high-stakes settings, it becomes vital to know when the output of an LLM may be unreliable. Depending on whether an answer is trustworthy, a system can then choose to route the question to another expert, or otherwise fall back on a safe default behavior. In this work, we study the extent to which LLMs can reliably indicate confidence in their answers, and how this notion of confidence can translate into downstream accuracy gains. We propose Self-REF, a lightweight training strategy to teach LLMs to express confidence in whether their answers are correct in a reliable manner. Self-REF introduces confidence tokens into the LLM, from which a confidence score can be extracted. Compared to conventional approaches such as verbalizing confidence and examining token probabilities, we demonstrate empirically that confidence tokens show significant improvements in downstream routing and rejection learning tasks.
☆ Inductive Gradient Adjustment For Spectral Bias In Implicit Neural Representations
Implicit Neural Representations (INRs), as a versatile representation paradigm, have achieved success in various computer vision tasks. Due to the spectral bias of the vanilla multi-layer perceptrons (MLPs), existing methods focus on designing MLPs with sophisticated architectures or repurposing training techniques for highly accurate INRs. In this paper, we delve into the linear dynamics model of MLPs and theoretically identify the empirical Neural Tangent Kernel (eNTK) matrix as a reliable link between spectral bias and training dynamics. Based on eNTK matrix, we propose a practical inductive gradient adjustment method, which could purposefully improve the spectral bias via inductive generalization of eNTK-based gradient transformation matrix. We evaluate our method on different INRs tasks with various INR architectures and compare to existing training techniques. The superior representation performance clearly validates the advantage of our proposed method. Armed with our gradient adjustment method, better INRs with more enhanced texture details and sharpened edges can be learned from data by tailored improvements on spectral bias.
comment: 28 pages, 12 figures
☆ The Latent Road to Atoms: Backmapping Coarse-grained Protein Structures with Latent Diffusion
Coarse-grained(CG) molecular dynamics simulations offer computational efficiency for exploring protein conformational ensembles and thermodynamic properties. Though coarse representations enable large-scale simulations across extended temporal and spatial ranges, the sacrifice of atomic-level details limits their utility in tasks such as ligand docking and protein-protein interaction prediction. Backmapping, the process of reconstructing all-atom structures from coarse-grained representations, is crucial for recovering these fine details. While recent machine learning methods have made strides in protein structure generation, challenges persist in reconstructing diverse atomistic conformations that maintain geometric accuracy and chemical validity. In this paper, we present Latent Diffusion Backmapping (LDB), a novel approach leveraging denoising diffusion within latent space to address these challenges. By combining discrete latent encoding with diffusion, LDB bypasses the need for equivariant and internal coordinate manipulation, significantly simplifying the training and sampling processes as well as facilitating better and wider exploration in configuration space. We evaluate LDB's state-of-the-art performance on three distinct protein datasets, demonstrating its ability to efficiently reconstruct structures with high structural accuracy and chemical validity. Moreover, LDB shows exceptional versatility in capturing diverse protein ensembles, highlighting its capability to explore intricate conformational spaces. Our results position LDB as a powerful and scalable approach for backmapping, effectively bridging the gap between CG simulations and atomic-level analyses in computational biology.
comment: Paper under review
☆ A Simplifying and Learnable Graph Convolutional Attention Network for Unsupervised Knowledge Graphs Alignment
The success of current Entity Alignment (EA) task depends largely on the supervision information provided by labeled data. Considering the cost of labeled data, most supervised methods are difficult to apply in practical scenarios. Therefore, more and more works based on contrastive learning, active learning or other deep learning techniques have been developed, to solve the performance bottleneck caused by the lack of labeled data. However, the existing unsupervised EA methods still have some limitations, either their modeling complexity is high or they cannot balance the effectiveness and practicality of alignment. To overcome these issues, we propose a Simplifying and Learnable graph convolutional attention network for Unsupervised Knowledge Graphs alignment method (SLU). Specifically, we first introduce LCAT, a new and simple framework as the backbone network to model the graph structure of two KGs. Then we design a reconstruction method of relation structure based on potential matching relations for efficiently filtering invalid neighborhood information of aligned entities, to improve the usability and scalability of SLU. Impressively, a similarity function based on consistency is proposed to better measure the similarity of candidate entity pairs. Finally, we conduct extensive experiments on three datasets of different sizes (15K and 100K) and different types (cross-lingual and monolingual) to verify the superiority of SLU. Experimental results show that SLU significantly improves alignment accuracy, outperforming 25 supervised or unsupervised methods, and improving 6.4% in Hits@1 over the best baseline in the best case.
comment: 14 pages, 3 figures
☆ scFusionTTT: Single-cell transcriptomics and proteomics fusion with Test-Time Training layers
Single-cell multi-omics (scMulti-omics) refers to the paired multimodal data, such as Cellular Indexing of Transcriptomes and Epitopes by Sequencing (CITE-seq), where the regulation of each cell was measured from different modalities, i.e. genes and proteins. scMulti-omics can reveal heterogeneity inside tumors and understand the distinct genetic properties of diverse cell types, which is crucial to targeted therapy. Currently, deep learning methods based on attention structures in the bioinformatics area face two challenges. The first challenge is the vast number of genes in a single cell. Traditional attention-based modules struggled to effectively leverage all gene information due to their limited capacity for long-context learning and high-complexity computing. The second challenge is that genes in the human genome are ordered and influence each other's expression. Most of the methods ignored this sequential information. The recently introduced Test-Time Training (TTT) layer is a novel sequence modeling approach, particularly suitable for handling long contexts like genomics data because TTT layer is a linear complexity sequence modeling structure and is better suited to data with sequential relationships. In this paper, we propose scFusionTTT, a novel method for Single-Cell multimodal omics Fusion with TTT-based masked autoencoder. Of note, we combine the order information of genes and proteins in the human genome with the TTT layer, fuse multimodal omics, and enhance unimodal omics analysis. Finally, the model employs a three-stage training strategy, which yielded the best performance across most metrics in four multimodal omics datasets and four unimodal omics datasets, demonstrating the superior performance of our model. The dataset and code will be available on https://github.com/DM0815/scFusionTTT.
☆ FDF: Flexible Decoupled Framework for Time Series Forecasting with Conditional Denoising and Polynomial Modeling
Time series forecasting is vital in numerous web applications, influencing critical decision-making across industries. While diffusion models have recently gained increasing popularity for this task, we argue they suffer from a significant drawback: indiscriminate noise addition to the original time series followed by denoising, which can obscure underlying dynamic evolving trend and complicate forecasting. To address this limitation, we propose a novel flexible decoupled framework (FDF) that learns high-quality time series representations for enhanced forecasting performance. A key characteristic of our approach leverages the inherent inductive bias of time series data by decomposing it into trend and seasonal components, each modeled separately to enable decoupled analysis and modeling. Specifically, we propose an innovative Conditional Denoising Seasonal Module (CDSM) within the diffusion model, which leverages statistical information from the historical window to conditionally model the complex seasonal component. Notably, we incorporate a Polynomial Trend Module (PTM) to effectively capture the smooth trend component, thereby enhancing the model's ability to represent temporal dependencies. Extensive experiments validate the effectiveness of our framework, demonstrating superior performance over existing methods and higlighting its flexibility in time series forecasting. We hope our work can bring a new perspective for time series forecasting. We intend to make our code publicly available as open-source in the future.
☆ Disentangling Likes and Dislikes in Personalized Generative Explainable Recommendation
Recent research on explainable recommendation generally frames the task as a standard text generation problem, and evaluates models simply based on the textual similarity between the predicted and ground-truth explanations. However, this approach fails to consider one crucial aspect of the systems: whether their outputs accurately reflect the users' (post-purchase) sentiments, i.e., whether and why they would like and/or dislike the recommended items. To shed light on this issue, we introduce new datasets and evaluation methods that focus on the users' sentiments. Specifically, we construct the datasets by explicitly extracting users' positive and negative opinions from their post-purchase reviews using an LLM, and propose to evaluate systems based on whether the generated explanations 1) align well with the users' sentiments, and 2) accurately identify both positive and negative opinions of users on the target items. We benchmark several recent models on our datasets and demonstrate that achieving strong performance on existing metrics does not ensure that the generated explanations align well with the users' sentiments. Lastly, we find that existing models can provide more sentiment-aware explanations when the users' (predicted) ratings for the target items are directly fed into the models as input. We will release our code and datasets upon acceptance.
☆ Quamba: A Post-Training Quantization Recipe for Selective State Space Models
State Space Models (SSMs) have emerged as an appealing alternative to Transformers for large language models, achieving state-of-the-art accuracy with constant memory complexity which allows for holding longer context lengths than attention-based networks. The superior computational efficiency of SSMs in long sequence modeling positions them favorably over Transformers in many scenarios. However, improving the efficiency of SSMs on request-intensive cloud-serving and resource-limited edge applications is still a formidable task. SSM quantization is a possible solution to this problem, making SSMs more suitable for wide deployment, while still maintaining their accuracy. Quantization is a common technique to reduce the model size and to utilize the low bit-width acceleration features on modern computing units, yet existing quantization techniques are poorly suited for SSMs. Most notably, SSMs have highly sensitive feature maps within the selective scan mechanism (i.e., linear recurrence) and massive outliers in the output activations which are not present in the output of token-mixing in the self-attention modules. To address this issue, we propose a static 8-bit per-tensor SSM quantization method which suppresses the maximum values of the input activations to the selective SSM for finer quantization precision and quantizes the output activations in an outlier-free space with Hadamard transform. Our 8-bit weight-activation quantized Mamba 2.8B SSM benefits from hardware acceleration and achieves a 1.72x lower generation latency on an Nvidia Orin Nano 8G, with only a 0.9% drop in average accuracy on zero-shot tasks. The experiments demonstrate the effectiveness and practical applicability of our approach for deploying SSM-based models of all sizes on both cloud and edge platforms.
☆ From PINNs to PIKANs: Recent Advances in Physics-Informed Machine Learning
Physics-Informed Neural Networks (PINNs) have emerged as a key tool in Scientific Machine Learning since their introduction in 2017, enabling the efficient solution of ordinary and partial differential equations using sparse measurements. Over the past few years, significant advancements have been made in the training and optimization of PINNs, covering aspects such as network architectures, adaptive refinement, domain decomposition, and the use of adaptive weights and activation functions. A notable recent development is the Physics-Informed Kolmogorov-Arnold Networks (PIKANS), which leverage a representation model originally proposed by Kolmogorov in 1957, offering a promising alternative to traditional PINNs. In this review, we provide a comprehensive overview of the latest advancements in PINNs, focusing on improvements in network design, feature expansion, optimization techniques, uncertainty quantification, and theoretical insights. We also survey key applications across a range of fields, including biomedicine, fluid and solid mechanics, geophysics, dynamical systems, heat transfer, chemical engineering, and beyond. Finally, we review computational frameworks and software tools developed by both academia and industry to support PINN research and applications.
comment: physics-informed neural networks, Kolmogorov-Arnold networks, optimization algorithms, separable PINNs, self-adaptive weights, uncertainty quantification
☆ MixEHR-Nest: Identifying Subphenotypes within Electronic Health Records through Hierarchical Guided-Topic Modeling
Automatic subphenotyping from electronic health records (EHRs)provides numerous opportunities to understand diseases with unique subgroups and enhance personalized medicine for patients. However, existing machine learning algorithms either focus on specific diseases for better interpretability or produce coarse-grained phenotype topics without considering nuanced disease patterns. In this study, we propose a guided topic model, MixEHR-Nest, to infer sub-phenotype topics from thousands of disease using multi-modal EHR data. Specifically, MixEHR-Nest detects multiple subtopics from each phenotype topic, whose prior is guided by the expert-curated phenotype concepts such as Phenotype Codes (PheCodes) or Clinical Classification Software (CCS) codes. We evaluated MixEHR-Nest on two EHR datasets: (1) the MIMIC-III dataset consisting of over 38 thousand patients from intensive care unit (ICU) from Beth Israel Deaconess Medical Center (BIDMC) in Boston, USA; (2) the healthcare administrative database PopHR, comprising 1.3 million patients from Montreal, Canada. Experimental results demonstrate that MixEHR-Nest can identify subphenotypes with distinct patterns within each phenotype, which are predictive for disease progression and severity. Consequently, MixEHR-Nest distinguishes between type 1 and type 2 diabetes by inferring subphenotypes using CCS codes, which do not differentiate these two subtype concepts. Additionally, MixEHR-Nest not only improved the prediction accuracy of short-term mortality of ICU patients and initial insulin treatment in diabetic patients but also revealed the contributions of subphenotypes. For longitudinal analysis, MixEHR-Nest identified subphenotypes of distinct age prevalence under the same phenotypes, such as asthma, leukemia, epilepsy, and depression. The MixEHR-Nest software is available at GitHub: https://github.com/li-lab-mcgill/MixEHR-Nest.
☆ Balancing Label Quantity and Quality for Scalable Elicitation
Scalable oversight studies methods of training and evaluating AI systems in domains where human judgement is unreliable or expensive, such as scientific research and software engineering in complex codebases. Recent work in this area by Burns et al. (2023) suggests that Language Models (LMs) pretrained on internet-scale corpora exhibit an inductive bias toward producing correct answers, even when finetuned on error-prone labels produced by a smaller language model. This suggests that massive pretraining combined with finetuning on imperfect human labels may be a solid baseline method for scalable oversight. In the real world, however, label quality is not fixed: practitioners face a quantity-quality tradeoff when generating finetuning data. In this paper, we explore the microeconomics of the quantity-quality tradeoff on binary NLP classification tasks used in Burns et al. (2023). We find that there are three regimes of eliciting classification knowledge from pretrained models using supervised finetuning: quantity-dominant, quality-dominant, and a mixed regime involving the use of low- and high-quality data together to attain higher accuracy at a lower cost than using either alone. We explore sample-efficient elicitation methods that make use of two datasets of differing qualities, and establish a Pareto frontier of scalable elicitation methods that optimally trade off labeling cost and classifier performance.
☆ LLMOPT: Learning to Define and Solve General Optimization Problems from Scratch
Optimization problems are prevalent across various scenarios. Formulating and then solving optimization problems described by natural language often requires highly specialized human expertise, which could block the widespread application of optimization-based decision making. To make problem formulating and solving automated, leveraging large language models (LLMs) has emerged as a potential way. However, this kind of way suffers from the issue of optimization generalization. Namely, the accuracy of most current LLM-based methods and the generality of optimization problem types that they can model are still limited. In this paper, we propose a unified learning-based framework called LLMOPT to boost optimization generalization. Starting from the natural language descriptions of optimization problems and a pre-trained LLM, LLMOPT constructs the introduced five-element formulation as a universal model for learning to define diverse optimization problem types. Then, LLMOPT employs the multi-instruction tuning to enhance both problem formalization and solver code generation accuracy and generality. After that, to prevent hallucinations in LLMs, such as sacrificing solving accuracy to avoid execution errors, model alignment and self-correction mechanism are adopted in LLMOPT. We evaluate the optimization generalization ability of LLMOPT and compared methods across six real-world datasets covering roughly 20 fields such as health, environment, energy and manufacturing, etc. Extensive experiment results show that LLMOPT is able to model various optimization problem types such as linear/nonlinear programming, mixed integer programming and combinatorial optimization, and achieves a notable 11.08% average solving accuracy improvement compared with the state-of-the-art methods. The code is available at https://github.com/caigaojiang/LLMOPT.
☆ AsymKV: Enabling 1-Bit Quantization of KV Cache with Layer-Wise Asymmetric Quantization Configurations
Large language models have shown exceptional capabilities in a wide range of tasks, such as text generation and video generation, among others. However, due to their massive parameter count, these models often require substantial storage space, imposing significant constraints on the machines deploying LLMs. To overcome this limitation, one research direction proposes to compress the models using integer replacements for floating-point numbers, in a process known as Quantization. Some recent studies suggest quantizing the key and value cache (KV Cache) of LLMs, and designing quantization techniques that treat the key and value matrices equivalently. This work delves deeper into the asymmetric structural roles of KV Cache, a phenomenon where the transformer's output loss is more sensitive to the quantization of key matrices. We conduct a systematic examination of the attention output error resulting from key and value quantization. The phenomenon inspires us to propose an asymmetric quantization strategy. Our approach allows for 1-bit quantization of the KV cache by implementing distinct configurations for key and value matrices. We carry out experiments across a variety of datasets, demonstrating that our proposed model allows for the quantization of up to 75% decoder layers with 1 bit, while simultaneously maintaining performance levels comparable to those of the models with floating parameters.
comment: 12 pages, 4 figures
☆ Estimating the Probabilities of Rare Outputs in Language Models
We consider the problem of low probability estimation: given a machine learning model and a formally-specified input distribution, how can we estimate the probability of a binary property of the model's output, even when that probability is too small to estimate by random sampling? This problem is motivated by the need to improve worst-case performance, which distribution shift can make much more likely. We study low probability estimation in the context of argmax sampling from small transformer language models. We compare two types of methods: importance sampling, which involves searching for inputs giving rise to the rare output, and activation extrapolation, which involves extrapolating a probability distribution fit to the model's logits. We find that importance sampling outperforms activation extrapolation, but both outperform naive sampling. Finally, we explain how minimizing the probability estimate of an undesirable behavior generalizes adversarial training, and argue that new methods for low probability estimation are needed to provide stronger guarantees about worst-case performance.
comment: 27 pages, 9 figures
☆ TabSeq: A Framework for Deep Learning on Tabular Data via Sequential Ordering ICPR 2024
Effective analysis of tabular data still poses a significant problem in deep learning, mainly because features in tabular datasets are often heterogeneous and have different levels of relevance. This work introduces TabSeq, a novel framework for the sequential ordering of features, addressing the vital necessity to optimize the learning process. Features are not always equally informative, and for certain deep learning models, their random arrangement can hinder the model's learning capacity. Finding the optimum sequence order for such features could improve the deep learning models' learning process. The novel feature ordering technique we provide in this work is based on clustering and incorporates both local ordering and global ordering. It is designed to be used with a multi-head attention mechanism in a denoising autoencoder network. Our framework uses clustering to align comparable features and improve data organization. Multi-head attention focuses on essential characteristics, whereas the denoising autoencoder highlights important aspects by rebuilding from distorted inputs. This method improves the capability to learn from tabular data while lowering redundancy. Our research, demonstrating improved performance through appropriate feature sequence rearrangement using raw antibody microarray and two other real-world biomedical datasets, validates the impact of feature ordering. These results demonstrate that feature ordering can be a viable approach to improved deep learning of tabular data.
comment: This paper has been accepted for presentation at the 26th International Conference on Pattern Recognition (ICPR 2024) in Kolkata, India
☆ Meta-DiffuB: A Contextualized Sequence-to-Sequence Text Diffusion Model with Meta-Exploration
The diffusion model, a new generative modeling paradigm, has achieved significant success in generating images, audio, video, and text. It has been adapted for sequence-to-sequence text generation (Seq2Seq) through DiffuSeq, termed S2S Diffusion. Existing S2S-Diffusion models predominantly rely on fixed or hand-crafted rules to schedule noise during the diffusion and denoising processes. However, these models are limited by non-contextualized noise, which fails to fully consider the characteristics of Seq2Seq tasks. In this paper, we propose the Meta-DiffuB framework - a novel scheduler-exploiter S2S-Diffusion paradigm designed to overcome the limitations of existing S2S-Diffusion models. We employ Meta-Exploration to train an additional scheduler model dedicated to scheduling contextualized noise for each sentence. Our exploiter model, an S2S-Diffusion model, leverages the noise scheduled by our scheduler model for updating and generation. Meta-DiffuB achieves state-of-the-art performance compared to previous S2S-Diffusion models and fine-tuned pre-trained language models (PLMs) across four Seq2Seq benchmark datasets. We further investigate and visualize the impact of Meta-DiffuB's noise scheduling on the generation of sentences with varying difficulties. Additionally, our scheduler model can function as a "plug-and-play" model to enhance DiffuSeq without the need for fine-tuning during the inference stage.
☆ Context-Enhanced Multi-View Trajectory Representation Learning: Bridging the Gap through Self-Supervised Models
Modeling trajectory data with generic-purpose dense representations has become a prevalent paradigm for various downstream applications, such as trajectory classification, travel time estimation and similarity computation. However, existing methods typically rely on trajectories from a single spatial view, limiting their ability to capture the rich contextual information that is crucial for gaining deeper insights into movement patterns across different geospatial contexts. To this end, we propose MVTraj, a novel multi-view modeling method for trajectory representation learning. MVTraj integrates diverse contextual knowledge, from GPS to road network and points-of-interest to provide a more comprehensive understanding of trajectory data. To align the learning process across multiple views, we utilize GPS trajectories as a bridge and employ self-supervised pretext tasks to capture and distinguish movement patterns across different spatial views. Following this, we treat trajectories from different views as distinct modalities and apply a hierarchical cross-modal interaction module to fuse the representations, thereby enriching the knowledge derived from multiple sources. Extensive experiments on real-world datasets demonstrate that MVTraj significantly outperforms existing baselines in tasks associated with various spatial views, validating its effectiveness and practical utility in spatio-temporal modeling.
☆ Golyadkin's Torment: Doppelgängers and Adversarial Vulnerability
Many machine learning (ML) classifiers are claimed to outperform humans, but they still make mistakes that humans do not. The most notorious examples of such mistakes are adversarial visual metamers. This paper aims to define and investigate the phenomenon of adversarial Doppelgangers (AD), which includes adversarial visual metamers, and to compare the performance and robustness of ML classifiers to human performance. We find that AD are inputs that are close to each other with respect to a perceptual metric defined in this paper. AD are qualitatively different from the usual adversarial examples. The vast majority of classifiers are vulnerable to AD and robustness-accuracy trade-offs may not improve them. Some classification problems may not admit any AD robust classifiers because the underlying classes are ambiguous. We provide criteria that can be used to determine whether a classification problem is well defined or not; describe the structure and attributes of an AD-robust classifier; introduce and explore the notions of conceptual entropy and regions of conceptual ambiguity for classifiers that are vulnerable to AD attacks, along with methods to bound the AD fooling rate of an attack. We define the notion of classifiers that exhibit hypersensitive behavior, that is, classifiers whose only mistakes are adversarial Doppelgangers. Improving the AD robustness of hyper-sensitive classifiers is equivalent to improving accuracy. We identify conditions guaranteeing that all classifiers with sufficiently high accuracy are hyper-sensitive. Our findings are aimed at significant improvements in the reliability and security of machine learning systems.
☆ CohEx: A Generalized Framework for Cohort Explanation
eXplainable Artificial Intelligence (XAI) has garnered significant attention for enhancing transparency and trust in machine learning models. However, the scopes of most existing explanation techniques focus either on offering a holistic view of the explainee model (global explanation) or on individual instances (local explanation), while the middle ground, i.e., cohort-based explanation, is less explored. Cohort explanations offer insights into the explainee's behavior on a specific group or cohort of instances, enabling a deeper understanding of model decisions within a defined context. In this paper, we discuss the unique challenges and opportunities associated with measuring cohort explanations, define their desired properties, and create a generalized framework for generating cohort explanations based on supervised clustering.
♻ ☆ Towards Multilingual LLM Evaluation for European Languages
The rise of Large Language Models (LLMs) has revolutionized natural language processing across numerous languages and tasks. However, evaluating LLM performance in a consistent and meaningful way across multiple European languages remains challenging, especially due to the scarcity of language-parallel multilingual benchmarks. We introduce a multilingual evaluation approach tailored for European languages. We employ translated versions of five widely-used benchmarks to assess the capabilities of 40 LLMs across 21 European languages. Our contributions include examining the effectiveness of translated benchmarks, assessing the impact of different translation services, and offering a multilingual evaluation framework for LLMs that includes newly created datasets: EU20-MMLU, EU20-HellaSwag, EU20-ARC, EU20-TruthfulQA, and EU20-GSM8K. The benchmarks and results are made publicly available to encourage further research in multilingual LLM evaluation.
♻ ☆ Retrieval Augmented Generation or Long-Context LLMs? A Comprehensive Study and Hybrid Approach EMNLP 2024
Retrieval Augmented Generation (RAG) has been a powerful tool for Large Language Models (LLMs) to efficiently process overly lengthy contexts. However, recent LLMs like Gemini-1.5 and GPT-4 show exceptional capabilities to understand long contexts directly. We conduct a comprehensive comparison between RAG and long-context (LC) LLMs, aiming to leverage the strengths of both. We benchmark RAG and LC across various public datasets using three latest LLMs. Results reveal that when resourced sufficiently, LC consistently outperforms RAG in terms of average performance. However, RAG's significantly lower cost remains a distinct advantage. Based on this observation, we propose Self-Route, a simple yet effective method that routes queries to RAG or LC based on model self-reflection. Self-Route significantly reduces the computation cost while maintaining a comparable performance to LC. Our findings provide a guideline for long-context applications of LLMs using RAG and LC.
comment: Accepted to EMNLP 2024 industry track
♻ ☆ Many-Shot In-Context Learning NeurIPS
Large language models (LLMs) excel at few-shot in-context learning (ICL) -- learning from a few examples provided in context at inference, without any weight updates. Newly expanded context windows allow us to investigate ICL with hundreds or thousands of examples -- the many-shot regime. Going from few-shot to many-shot, we observe significant performance gains across a wide variety of generative and discriminative tasks. While promising, many-shot ICL can be bottlenecked by the available amount of human-generated examples. To mitigate this limitation, we explore two new settings: Reinforced and Unsupervised ICL. Reinforced ICL uses model-generated chain-of-thought rationales in place of human examples. Unsupervised ICL removes rationales from the prompt altogether, and prompts the model only with domain-specific questions. We find that both Reinforced and Unsupervised ICL can be quite effective in the many-shot regime, particularly on complex reasoning tasks. Finally, we demonstrate that, unlike few-shot learning, many-shot learning is effective at overriding pretraining biases, can learn high-dimensional functions with numerical inputs, and performs comparably to fine-tuning. We also find that inference cost increases linearly in the many-shot regime, and frontier LLMs benefit from many-shot ICL to varying degrees. Our analysis also reveals the limitations of next-token prediction loss as an indicator of downstream ICL performance.
comment: NeurIPS (Spotlight)
♻ ☆ Data-Driven Estimation of Heterogeneous Treatment Effects
Estimating how a treatment affects different individuals, known as heterogeneous treatment effect estimation, is an important problem in empirical sciences. In the last few years, there has been a considerable interest in adapting machine learning algorithms to the problem of estimating heterogeneous effects from observational and experimental data. However, these algorithms often make strong assumptions about the observed features in the data and ignore the structure of the underlying causal model, which can lead to biased estimation. At the same time, the underlying causal mechanism is rarely known in real-world datasets, making it hard to take it into consideration. In this work, we provide a survey of state-of-the-art data-driven methods for heterogeneous treatment effect estimation using machine learning, broadly categorizing them as methods that focus on counterfactual prediction and methods that directly estimate the causal effect. We also provide an overview of a third category of methods which rely on structural causal models and learn the model structure from data. Our empirical evaluation under various underlying structural model mechanisms shows the advantages and deficiencies of existing estimators and of the metrics for measuring their performance.
♻ ☆ Dynamic Topic Language Model on Heterogeneous Children's Mental Health Clinical Notes
Mental health diseases affect children's lives and well-beings which have received increased attention since the COVID-19 pandemic. Analyzing psychiatric clinical notes with topic models is critical to evaluating children's mental status over time. However, few topic models are built for longitudinal settings, and most existing approaches fail to capture temporal trajectories for each document. To address these challenges, we develop a dynamic topic model with consistent topics and individualized temporal dependencies on the evolving document metadata. Our model preserves the semantic meaning of discovered topics over time and incorporates heterogeneity among documents. In particular, when documents can be categorized, we propose a classifier-free approach to maximize topic heterogeneity across different document groups. We also present an efficient variational optimization procedure adapted for the multistage longitudinal setting. In this case study, we apply our method to the psychiatric clinical notes from a large tertiary pediatric hospital in Southern California and achieve a 38% increase in the overall coherence of extracted topics. Our real data analysis reveals that children tend to express more negative emotions during state shutdowns and more positive when schools reopen. Furthermore, it suggests that sexual and gender minority (SGM) children display more pronounced reactions to major COVID-19 events and a greater sensitivity to vaccine-related news than non-SGM children. This study examines children's mental health progression during the pandemic and offers clinicians valuable insights to recognize disparities in children's mental health related to their sexual and gender identities.
♻ ☆ Achieving Exponential Asymptotic Optimality in Average-Reward Restless Bandits without Global Attractor Assumption
We consider the infinite-horizon average-reward restless bandit problem. We propose a novel \emph{two-set policy} that maintains two dynamic subsets of arms: one subset of arms has a nearly optimal state distribution and takes actions according to an Optimal Local Control routine; the other subset of arms is driven towards the optimal state distribution and gradually merged into the first subset. We show that our two-set policy is asymptotically optimal with an $O(\exp(-C N))$ optimality gap for an $N$-armed problem, under the mild assumptions of aperiodic-unichain, non-degeneracy, and local stability. Our policy is the first to achieve \emph{exponential asymptotic optimality} under the above set of easy-to-verify assumptions, whereas prior work either requires a strong \emph{global attractor} assumption or only achieves an $O(1/\sqrt{N})$ optimality gap. We further discuss obstacles in weakening the assumptions by demonstrating examples where exponential asymptotic optimality is not achievable when any of the three assumptions is violated. Notably, we prove a lower bound for a large class of locally unstable restless bandits, showing that local stability is particularly fundamental for exponential asymptotic optimality. Finally, we use simulations to demonstrate that the two-set policy outperforms previous policies on certain RB problems and performs competitively overall.
comment: 55 pages, 4 figures. In this version we included simulations
♻ ☆ Guided Multi-objective Generative AI to Enhance Structure-based Drug Design
Generative AI has the potential to revolutionize drug discovery. Yet, despite recent advances in deep learning, existing models cannot generate molecules that satisfy all desired physicochemical properties. Herein, we describe IDOLpro, a generative chemistry AI combining diffusion with multi-objective optimization for structure-based drug design. Differentiable scoring functions guide the latent variables of the diffusion model to explore uncharted chemical space and generate novel ligands in silico, optimizing a plurality of target physicochemical properties. We demonstrate our platform's effectiveness by generating ligands with optimized binding affinity and synthetic accessibility on two benchmark sets. IDOLpro produces ligands with binding affinities over 10%-20% better than the next best state-of-the-art method on each test set, producing more drug-like molecules with generally better synthetic accessibility scores than other methods. We do a head-to-head comparison of IDOLpro against a classic virtual screen of a large database of drug-like molecules. We show that IDOLpro can generate molecules for a range of important disease-related targets with better binding affinity and synthetic accessibility than any molecule found in the virtual screen while being over 100x faster and less expensive to run. On a test set of experimental complexes, IDOLpro is the first to produce molecules with better binding affinities than experimentally observed ligands. IDOLpro can accommodate other scoring functions (e.g. ADME-Tox) to accelerate hit-finding, hit-to-lead, and lead optimization for drug discovery.
♻ ☆ Stage-Aware Learning for Dynamic Treatments
Recent advances in dynamic treatment regimes (DTRs) facilitate the search for optimal treatments, which are tailored to individuals' specific needs and able to maximize their expected clinical benefits. However, existing algorithms relying on consistent trajectories, such as inverse probability weighting estimators (IPWEs), could suffer from insufficient sample size under optimal treatments and a growing number of decision-making stages, particularly in the context of chronic diseases. To address these challenges, we propose a novel individualized learning method which estimates the DTR with a focus on prioritizing alignment between the observed treatment trajectory and the one obtained by the optimal regime across decision stages. By relaxing the restriction that the observed trajectory must be fully aligned with the optimal treatments, our approach substantially improves the sample efficiency and stability of IPWE-based methods. In particular, the proposed learning scheme builds a more general framework which includes the popular outcome weighted learning framework as a special case of ours. Moreover, we introduce the notion of stage importance scores along with an attention mechanism to explicitly account for heterogeneity among decision stages. We establish the theoretical properties of the proposed approach, including the Fisher consistency and finite-sample performance bound. Empirically, we evaluate the proposed method in extensive simulated environments and a real case study for the COVID-19 pandemic.
♻ ☆ Corrective Machine Unlearning
Machine Learning models increasingly face data integrity challenges due to the use of large-scale training datasets drawn from the Internet. We study what model developers can do if they detect that some data was manipulated or incorrect. Such manipulated data can cause adverse effects including vulnerability to backdoored samples, systemic biases, and reduced accuracy on certain input domains. Realistically, all manipulated training samples cannot be identified, and only a small, representative subset of the affected data can be flagged. We formalize Corrective Machine Unlearning as the problem of mitigating the impact of data affected by unknown manipulations on a trained model, only having identified a subset of the corrupted data. We demonstrate that the problem of corrective unlearning has significantly different requirements from traditional privacy-oriented unlearning. We find most existing unlearning methods, including retraining-from-scratch without the deletion set, require most of the manipulated data to be identified for effective corrective unlearning. However, one approach, Selective Synaptic Dampening, achieves limited success, unlearning adverse effects with just a small portion of the manipulated samples in our setting, which shows encouraging signs for future progress. We hope our work spurs research towards developing better methods for corrective unlearning and offers practitioners a new strategy to handle data integrity challenges arising from web-scale training. Code is available at https://github.com/drimpossible/corrective-unlearning-bench.
comment: Published in Transactions of Machine Learning Research (TMLR), 17 pages, 7 figures
♻ ☆ GPTreeO: An R package for continual regression with dividing local Gaussian processes
We introduce GPTreeO, a flexible R package for scalable Gaussian process (GP) regression, particularly tailored to continual learning problems. GPTreeO builds upon the Dividing Local Gaussian Processes (DLGP) algorithm, in which a binary tree of local GP regressors is dynamically constructed using a continual stream of input data. In GPTreeO we extend the original DLGP algorithm by allowing continual optimisation of the GP hyperparameters, incorporating uncertainty calibration, and introducing new strategies for how the local partitions are created. Moreover, the modular code structure allows users to interface their favourite GP library to perform the local GP regression in GPTreeO. The flexibility of GPTreeO gives the user fine-grained control of the balance between computational speed, accuracy, stability and smoothness. We conduct a sensitivity analysis to show how GPTreeO's configurable features impact the regression performance in a continual learning setting.
comment: Updated the bibliography, and is now equivalent to the journal submission
♻ ☆ Moments of Clarity: Streamlining Latent Spaces in Machine Learning using Moment Pooling
Many machine learning applications involve learning a latent representation of data, which is often high-dimensional and difficult to directly interpret. In this work, we propose "Moment Pooling", a natural extension of Deep Sets networks which drastically decrease latent space dimensionality of these networks while maintaining or even improving performance. Moment Pooling generalizes the summation in Deep Sets to arbitrary multivariate moments, which enables the model to achieve a much higher effective latent dimensionality for a fixed latent dimension. We demonstrate Moment Pooling on the collider physics task of quark/gluon jet classification by extending Energy Flow Networks (EFNs) to Moment EFNs. We find that Moment EFNs with latent dimensions as small as 1 perform similarly to ordinary EFNs with higher latent dimension. This small latent dimension allows for the internal representation to be directly visualized and interpreted, which in turn enables the learned internal jet representation to be extracted in closed form.
comment: 15+7 pages, 14 figures, 7 tables. Code available at https://github.com/athiso/moment and https://github.com/rikab/MomentAnalysis; v2: Updated to match journal version
♻ ☆ Machine-learning prediction of tipping with applications to the Atlantic Meridional Overturning Circulation
Anticipating a tipping point, a transition from one stable steady state to another, is a problem of broad relevance due to the ubiquity of the phenomenon in diverse fields. The steady-state nature of the dynamics about a tipping point makes its prediction significantly more challenging than predicting other types of critical transitions from oscillatory or chaotic dynamics. Exploiting the benefits of noise, we develop a general data-driven and machine-learning approach to predicting potential future tipping in nonautonomous dynamical systems and validate the framework using examples from different fields. As an application, we address the problem of predicting the potential collapse of the Atlantic Meridional Overturning Circulation (AMOC), possibly driven by climate-induced changes in the freshwater input to the North Atlantic. Our predictions based on synthetic and currently available empirical data place a potential collapse window spanning from 2040 to 2065, in consistency with the results in the current literature.
comment: 11 pages, 7 figures
♻ ☆ LieRE: Generalizing Rotary Position Encodings
While Rotary Position Embeddings (RoPE) for large language models have become widely adopted, their application for other modalities has been slower. Here, we introduce Lie group Relative position Encodings (LieRE) that goes beyond RoPE in supporting n-dimensional inputs. We evaluate the performance of LieRE on 2D and 3D image classification tasks and observe that LieRE leads to marked relative improvements in performance (up to 9.7% for 2D and up to 25.5% for 3D), training efficiency (3.5x reduction), data efficiency (30%) compared to the baselines of DeiT III, RoPE-Mixed and Vision-Llama. https://github.com/Stanford-AIMI/LieRE
♻ ☆ ShadowLLM: Predictor-based Contextual Sparsity for Large Language Models EMNLP 2024
The high power consumption and latency-sensitive deployments of large language models (LLMs) have motivated efficiency techniques like quantization and sparsity. Contextual sparsity, where the sparsity pattern is input-dependent, is crucial in LLMs because the permanent removal of attention heads or neurons from LLMs can significantly degrade accuracy. Prior work has attempted to model contextual sparsity using neural networks trained to predict activation magnitudes, which can be used to dynamically prune structures with low predicted activation magnitude. In this paper, we look beyond magnitude-based pruning criteria to assess attention head and neuron importance in LLMs. We develop a novel predictor called ShadowLLM, which can shadow the LLM behavior and enforce better sparsity patterns, resulting in over 15% improvement in end-to-end accuracy compared to prior methods. In addition, ShadowLLM achieves up to a 20% speed-up over the state-of-the-art DejaVu framework. These enhancements are validated on Llama-2 and OPT models with up to 30 billion parameters. Our code is available at \href{https://github.com/abdelfattah-lab/shadow_llm/}{ShadowLLM}.
comment: Accepted to EMNLP 2024 (Main, Long Paper)
♻ ☆ FlashTex: Fast Relightable Mesh Texturing with LightControlNet
Manually creating textures for 3D meshes is time-consuming, even for expert visual content creators. We propose a fast approach for automatically texturing an input 3D mesh based on a user-provided text prompt. Importantly, our approach disentangles lighting from surface material/reflectance in the resulting texture so that the mesh can be properly relit and rendered in any lighting environment. We introduce LightControlNet, a new text-to-image model based on the ControlNet architecture, which allows the specification of the desired lighting as a conditioning image to the model. Our text-to-texture pipeline then constructs the texture in two stages. The first stage produces a sparse set of visually consistent reference views of the mesh using LightControlNet. The second stage applies a texture optimization based on Score Distillation Sampling (SDS) that works with LightControlNet to increase the texture quality while disentangling surface material from lighting. Our algorithm is significantly faster than previous text-to-texture methods, while producing high-quality and relightable textures.
comment: Project page: https://flashtex.github.io/
♻ ☆ Block-Attention for Efficient RAG
We introduce Block-Attention, an attention mechanism designed to address the increased inference latency and cost in Retrieval-Augmented Generation (RAG) scenarios. Traditional approaches often encode the entire context. Instead, Block-Attention divides retrieved documents into discrete blocks, with each block independently calculating key-value (KV) states except for the final block. In RAG scenarios, by defining each passage as a block, Block-Attention enables us to reuse the KV states of passages that have been seen before, thereby significantly reducing the latency and the computation overhead during inference. The implementation of Block-Attention involves block segmentation, position re-encoding, and fine-tuning the LLM to adapt to the Block-Attention mechanism. Experiments on four RAG benchmarks demonstrate that after block fine-tuning, the Block-Attention model achieves performance comparable to self-attention models (68.4\% vs 67.9\% on Llama3) or even superior performance (62.8\% vs 59.6\% on Mistral). Notably, Block-Attention significantly reduces the time to first token (TTFT) and floating point operations (FLOPs) to a very low level. It only takes 45 ms to output the first token for an input sequence with a total length of 32K. Compared to the self-attention models, the time consumption and corresponding FLOPs are reduced by 98.7\% and 99.8\%, respectively.
♻ ☆ Prompt-SAW: Leveraging Relation-Aware Graphs for Textual Prompt Compression
Large Language Models (LLMs) have shown exceptional abilities for multiple different natural language processing tasks. While prompting is a crucial tool for LLM inference, we observe that there is a significant cost associated with exceedingly lengthy prompts. Existing attempts to compress lengthy prompts lead to substandard results in terms of readability/interpretability of the compressed prompt, with a detrimental impact on prompt utility. To address this, we propose PromptSAW: Prompt compresSion via Relation AWare graphs, an effective strategy for prompt compression over task-agnostic and task-aware prompts. Prompt-SAW uses the prompt's textual information to build a graph and later extracts key information elements in the graph to come up with the compressed prompt. We also propose GSM8K-aug, i.e., an extended version of the existing GSM8K benchmark for task-agnostic prompts in order to provide a comprehensive evaluation platform. Experimental evaluation using benchmark datasets shows that prompts compressed by Prompt-SAW are not only better in terms of readability, but they also outperform the best-performing baseline models by up to 10.1 and 77.1, respectively, for task-agnostic and task-aware settings while compressing the original prompt text by 34.9 and 56.7.
comment: 16 pages
♻ ☆ Expected Sliced Transport Plans
The optimal transport (OT) problem has gained significant traction in modern machine learning for its ability to: (1) provide versatile metrics, such as Wasserstein distances and their variants, and (2) determine optimal couplings between probability measures. To reduce the computational complexity of OT solvers, methods like entropic regularization and sliced optimal transport have been proposed. The sliced OT framework improves efficiency by comparing one-dimensional projections (slices) of high-dimensional distributions. However, despite their computational efficiency, sliced-Wasserstein approaches lack a transportation plan between the input measures, limiting their use in scenarios requiring explicit coupling. In this paper, we address two key questions: Can a transportation plan be constructed between two probability measures using the sliced transport framework? If so, can this plan be used to define a metric between the measures? We propose a "lifting" operation to extend one-dimensional optimal transport plans back to the original space of the measures. By computing the expectation of these lifted plans, we derive a new transportation plan, termed expected sliced transport (EST) plans. We prove that using the EST plan to weight the sum of the individual Euclidean costs for moving from one point to another results in a valid metric between the input discrete probability measures. We demonstrate the connection between our approach and the recently proposed min-SWGG, along with illustrative numerical examples that support our theoretical findings.
♻ ☆ Learning Contrastive Feature Representations for Facial Action Unit Detection
Facial action unit (AU) detection has long encountered the challenge of detecting subtle feature differences when AUs activate. Existing methods often rely on encoding pixel-level information of AUs, which not only encodes additional redundant information but also leads to increased model complexity and limited generalizability. Additionally, the accuracy of AU detection is negatively impacted by the class imbalance issue of each AU type, and the presence of noisy and false AU labels. In this paper, we introduce a novel contrastive learning framework aimed for AU detection that incorporates both self-supervised and supervised signals, thereby enhancing the learning of discriminative features for accurate AU detection. To tackle the class imbalance issue, we employ a negative sample re-weighting strategy that adjusts the step size of updating parameters for minority and majority class samples. Moreover, to address the challenges posed by noisy and false AU labels, we employ a sampling technique that encompasses three distinct types of positive sample pairs. This enables us to inject self-supervised signals into the supervised signal, effectively mitigating the adverse effects of noisy labels. Our experimental assessments, conducted on four widely-utilized benchmark datasets (BP4D, DISFA, GFT and Aff-Wild2), underscore the superior performance of our approach compared to state-of-the-art methods of AU detection. Our code is available at \url{https://github.com/Ziqiao-Shang/AUNCE}.
comment: 35 pages, 18 figures, submitted to Pattern Recognition (PR)
♻ ☆ LayerSkip: Enabling Early Exit Inference and Self-Speculative Decoding ACL 2024
We present LayerSkip, an end-to-end solution to speed-up inference of large language models (LLMs). First, during training we apply layer dropout, with low dropout rates for earlier layers and higher dropout rates for later layers, and an early exit loss where all transformer layers share the same exit. Second, during inference, we show that this training recipe increases the accuracy of early exit at earlier layers, without adding any auxiliary layers or modules to the model. Third, we present a novel self-speculative decoding solution where we exit at early layers and verify and correct with remaining layers of the model. Our proposed self-speculative decoding approach has less memory footprint than other speculative decoding approaches and benefits from shared compute and activations of the draft and verification stages. We run experiments on different Llama model sizes on different types of training: pretraining from scratch, continual pretraining, finetuning on specific data domain, and finetuning on specific task. We implement our inference solution and show speedups of up to 2.16x on summarization for CNN/DM documents, 1.82x on coding, and 2.0x on TOPv2 semantic parsing task. We open source our code and checkpoints at https://github.com/facebookresearch/LayerSkip.
comment: ACL 2024
♻ ☆ LLM-based Cognitive Models of Students with Misconceptions
Accurately modeling student cognition is crucial for developing effective AI-driven educational technologies. A key challenge is creating realistic student models that satisfy two essential properties: (1) accurately replicating specific misconceptions, and (2) correctly solving problems where these misconceptions are not applicable. This dual requirement reflects the complex nature of student understanding, where misconceptions coexist with correct knowledge. This paper investigates whether Large Language Models (LLMs) can be instruction-tuned to meet this dual requirement and effectively simulate student thinking in algebra. We introduce MalAlgoPy, a novel Python library that generates datasets reflecting authentic student solution patterns through a graph-based representation of algebraic problem-solving. Utilizing MalAlgoPy, we define and examine Cognitive Student Models (CSMs) - LLMs instruction tuned to faithfully emulate realistic student behavior. Our findings reveal that LLMs trained on misconception examples can efficiently learn to replicate errors. However, the training diminishes the model's ability to solve problems correctly, particularly for problem types where the misconceptions are not applicable, thus failing to satisfy second property of CSMs. We demonstrate that by carefully calibrating the ratio of correct to misconception examples in the training data - sometimes as low as 0.25 - it is possible to develop CSMs that satisfy both properties. Our insights enhance our understanding of AI-based student models and pave the way for effective adaptive learning systems.
♻ ☆ MuJo: Multimodal Joint Feature Space Learning for Human Activity Recognition
Human Activity Recognition (HAR) is a longstanding problem in AI with applications in a broad range of areas, including healthcare, sports and fitness, security, and more. The performance of HAR in real-world settings is strongly dependent on the type and quality of the input signal that can be acquired. Given an unobstructed, high-quality camera view of a scene, computer vision systems, in particular in conjunction with foundation models, can today fairly reliably distinguish complex activities. On the other hand, recognition using modalities such as wearable sensors (which are often more broadly available, e.g., in mobile phones and smartwatches) is a more difficult problem, as the signals often contain less information and labeled training data is more difficult to acquire. To alleviate the need for labeled data, we introduce our comprehensive Fitness Multimodal Activity Dataset (FiMAD) in this work, which can be used with the proposed pre-training method MuJo (Multimodal Joint Feature Space Learning) to enhance HAR performance across various modalities. FiMAD was created using YouTube fitness videos and contains parallel video, language, pose, and simulated IMU sensor data. MuJo utilizes this dataset to learn a joint feature space for these modalities. We show that classifiers pre-trained on FiMAD can increase the performance on real HAR datasets such as MM-Fit, MyoGym, MotionSense, and MHEALTH. For instance, on MM-Fit, we achieve an Macro F1-Score of up to 0.855 when fine-tuning on only 2% of the training data and 0.942 when utilizing the full training set for classification tasks. We have compared our approach to other self-supervised ones and showed that, unlike them, ours can consistently improve on the baseline network performance as well as provide a better data-efficiency.
♻ ☆ Improving Generalization on the ProcGen Benchmark with Simple Architectural Changes and Scale
We demonstrate that recent advances in reinforcement learning (RL) combined with simple architectural changes significantly improves generalization on the ProcGen benchmark. These changes are frame stacking, replacing 2D convolutional layers with 3D convolutional layers, and scaling up the number of convolutional kernels per layer. Experimental results using a single set of hyperparameters across all environments show a 37.9\% reduction in the optimality gap compared to the baseline (from 0.58 to 0.36). This performance matches or exceeds current state-of-the-art methods. The proposed changes are largely orthogonal and therefore complementary to the existing approaches for improving generalization in RL, and our results suggest that further exploration in this direction could yield substantial improvements in addressing generalization challenges in deep reinforcement learning.
♻ ☆ Automatic Mapping of Anatomical Landmarks from Free-Text Using Large Language Models: Insights from Llama-2
Anatomical landmarks are vital in medical imaging for navigation and anomaly detection. Modern large language models (LLMs), like Llama-2, offer promise for automating the mapping of these landmarks in free-text radiology reports to corresponding positions in image data. Recent studies propose LLMs may develop coherent representations of generative processes. Motivated by these insights, we investigated whether LLMs accurately represent the spatial positions of anatomical landmarks. Through experiments with Llama-2 models, we found that they can linearly represent anatomical landmarks in space with considerable robustness to different prompts. These results underscore the potential of LLMs to enhance the efficiency and accuracy of medical imaging workflows.
comment: 6 pages, 2 figures, 1 table
♻ ☆ Efficient PAC Learning of Halfspaces with Constant Malicious Noise Rate
Understanding noise tolerance of learning algorithms under certain conditions is a central quest in learning theory. In this work, we study the problem of computationally efficient PAC learning of halfspaces in the presence of malicious noise, where an adversary can corrupt both instances and labels of training samples. The best-known noise tolerance either depends on a target error rate under distributional assumptions or on a margin parameter under large-margin conditions. In this work, we show that when both types of conditions are satisfied, it is possible to achieve {\em constant} noise tolerance by minimizing a reweighted hinge loss. Our key ingredients include: 1) an efficient algorithm that finds weights to control the gradient deterioration from corrupted samples, and 2) a new analysis on the robustness of the hinge loss equipped with such weights.
comment: author list in contribution order
♻ ☆ Generalization Error of the Tilted Empirical Risk
The generalization error (risk) of a supervised statistical learning algorithm quantifies its prediction ability on previously unseen data. Inspired by exponential tilting, Li et al. (2021) proposed the tilted empirical risk as a non-linear risk metric for machine learning applications such as classification and regression problems. In this work, we examine the generalization error of the tilted empirical risk. In particular, we provide uniform and information-theoretic bounds on the tilted generalization error, defined as the difference between the population risk and the tilted empirical risk, with a convergence rate of $O(1/\sqrt{n})$ where $n$ is the number of training samples. Furthermore, we study the solution to the KL-regularized expected tilted empirical risk minimization problem and derive an upper bound on the expected tilted generalization error with a convergence rate of $O(1/n)$.
comment: New results are added
♻ ☆ RECOVAR: Representation Covariances on Deep Latent Spaces for Seismic Event Detection
While modern deep learning methods have shown great promise in the problem of earthquake detection, the most successful methods so far have been based on supervised learning, which requires large datasets with ground-truth labels. The curation of such datasets is both time consuming and prone to systematic biases, which result in difficulties with cross-dataset generalization, hindering general applicability. In this paper, we develop an unsupervised method for earthquake detection that learns to detect earthquakes from raw waveforms, without access to ground truth labels. The performance is comparable to, and in some cases better than, some state-of-the-art supervised methods. Moreover, the method has strong \emph{cross-dataset generalization} performance. The algorithm utilizes deep autoencoders that learn to reproduce the waveforms after a data-compressive bottleneck and uses a simple, cross-covariance-based triggering algorithm at the bottleneck for labeling. The approach has the potential to be useful for time series datasets from other domains.
♻ ☆ Scaling Laws and Compute-Optimal Training Beyond Fixed Training Durations NeurIPS 2024
Scale has become a main ingredient in obtaining strong machine learning models. As a result, understanding a model's scaling properties is key to effectively designing both the right training setup as well as future generations of architectures. In this work, we argue that scale and training research has been needlessly complex due to reliance on the cosine schedule, which prevents training across different lengths for the same model size. We investigate the training behavior of a direct alternative -- constant learning rate and cooldowns -- and find that it scales predictably and reliably similar to cosine. Additionally, we show that stochastic weight averaging yields improved performance along the training trajectory, without additional training costs, across different scales. Importantly, with these findings we demonstrate that scaling experiments can be performed with significantly reduced compute and GPU hours by utilizing fewer but reusable training runs. Our code is available at \url{https://github.com/epfml/schedules-and-scaling/}.
comment: Spotlight at NeurIPS 2024
♻ ☆ Recursive deep learning framework for forecasting the decadal world economic outlook
The gross domestic product (GDP) is the most widely used indicator in macroeconomics and the main tool for measuring a country's economic output. Due to the diversity and complexity of the world economy, a wide range of models have been used, but there are challenges in making decadal GDP forecasts given unexpected changes such as emergence of catastrophic world events including pandemics and wars. Deep learning models are well suited for modelling temporal sequences and time series forecasting. In this paper, we develop a deep learning framework to forecast the GDP growth rate of the world economy over a decade. We use the Penn World Table as the data source featuring 13 countries prior to the COVID-19 pandemic, such as Australia, China, India, and the United States. We present a recursive deep learning framework to predict the GDP growth rate in the next ten years. We test prominent deep learning models and compare their results with traditional econometric models for selected developed and developing countries. Our decadal forecasts reveal that that most of the developed countries would experience economic growth slowdown, stagnation and even recession within five years (2020-2024). Furthermore, our model forecasts show that only China, France, and India would experience stable GDP growth.
♻ ☆ Targeted Vaccine: Safety Alignment for Large Language Models against Harmful Fine-Tuning via Layer-wise Perturbation
Harmful fine-tuning attack poses a serious threat to the online fine-tuning service. Vaccine, a recent alignment-stage defense, applies uniform perturbation to all layers of embedding to make the model robust to the simulated embedding drift. However, applying layer-wise uniform perturbation may lead to excess perturbations for some particular safety-irrelevant layers, resulting in defense performance degradation and unnecessary memory consumption. To address this limitation, we propose Targeted Vaccine (T-Vaccine), a memory-efficient safety alignment method that applies perturbation to only selected layers of the model. T-Vaccine follows two core steps: First, it uses gradient norm as a statistical metric to identify the safety-critical layers. Second, instead of applying uniform perturbation across all layers, T-Vaccine only applies perturbation to the safety-critical layers while keeping other layers frozen during training. Results show that T-Vaccine outperforms Vaccine in terms of both defense effectiveness and resource efficiency. Comparison with other defense baselines, e.g., RepNoise and TAR also demonstrate the superiority of T-Vaccine. Notably, T-Vaccine is the first defense that can address harmful fine-tuning issues for a 7B pre-trained models trained on consumer GPUs with limited memory (e.g., RTX 4090). Our code is available at https://github.com/Lslland/T-Vaccine.
♻ ☆ MirrorCheck: Efficient Adversarial Defense for Vision-Language Models
Vision-Language Models (VLMs) are becoming increasingly vulnerable to adversarial attacks as various novel attack strategies are being proposed against these models. While existing defenses excel in unimodal contexts, they currently fall short in safeguarding VLMs against adversarial threats. To mitigate this vulnerability, we propose a novel, yet elegantly simple approach for detecting adversarial samples in VLMs. Our method leverages Text-to-Image (T2I) models to generate images based on captions produced by target VLMs. Subsequently, we calculate the similarities of the embeddings of both input and generated images in the feature space to identify adversarial samples. Empirical evaluations conducted on different datasets validate the efficacy of our approach, outperforming baseline methods adapted from image classification domains. Furthermore, we extend our methodology to classification tasks, showcasing its adaptability and model-agnostic nature. Theoretical analyses and empirical findings also show the resilience of our approach against adaptive attacks, positioning it as an excellent defense mechanism for real-world deployment against adversarial threats.
♻ ☆ Beyond Thumbs Up/Down: Untangling Challenges of Fine-Grained Feedback for Text-to-Image Generation
Human feedback plays a critical role in learning and refining reward models for text-to-image generation, but the optimal form the feedback should take for learning an accurate reward function has not been conclusively established. This paper investigates the effectiveness of fine-grained feedback which captures nuanced distinctions in image quality and prompt-alignment, compared to traditional coarse-grained feedback (for example, thumbs up/down or ranking between a set of options). While fine-grained feedback holds promise, particularly for systems catering to diverse societal preferences, we show that demonstrating its superiority to coarse-grained feedback is not automatic. Through experiments on real and synthetic preference data, we surface the complexities of building effective models due to the interplay of model choice, feedback type, and the alignment between human judgment and computational interpretation. We identify key challenges in eliciting and utilizing fine-grained feedback, prompting a reassessment of its assumed benefits and practicality. Our findings -- e.g., that fine-grained feedback can lead to worse models for a fixed budget, in some settings; however, in controlled settings with known attributes, fine grained rewards can indeed be more helpful -- call for careful consideration of feedback attributes and potentially beckon novel modeling approaches to appropriately unlock the potential value of fine-grained feedback in-the-wild.
♻ ☆ Structure-Preserving Network Compression Via Low-Rank Induced Training Through Linear Layers Composition
Deep Neural Networks (DNNs) have achieved remarkable success in addressing many previously unsolvable tasks. However, the storage and computational requirements associated with DNNs pose a challenge for deploying these trained models on resource-limited devices. Therefore, a plethora of compression and pruning techniques have been proposed in recent years. Low-rank decomposition techniques are among the approaches most utilized to address this problem. Compared to post-training compression, compression-promoted training is still under-explored. In this paper, we present a theoretically-justified technique termed Low-Rank Induced Training (LoRITa), that promotes low-rankness through the composition of linear layers and compresses by using singular value truncation. This is achieved without the need to change the structure at inference time or require constrained and/or additional optimization, other than the standard weight decay regularization. Moreover, LoRITa eliminates the need to (i) initialize with pre-trained models, (ii) specify rank selection prior to training, and (iii) compute SVD in each iteration. Our experimental results (i) demonstrate the effectiveness of our approach using MNIST on Fully Connected Networks, CIFAR10 on Vision Transformers, and CIFAR10/100 and ImageNet on Convolutional Neural Networks, and (ii) illustrate that we achieve either competitive or state-of-the-art results when compared to leading structured pruning and low-rank training methods in terms of FLOPs and parameters drop. Our code is available at \url{https://github.com/XitongSystem/LoRITa/tree/main}.
♻ ☆ CL3: A Collaborative Learning Framework for the Medical Data Ensuring Data Privacy in the Hyperconnected Environment
In a hyperconnected environment, medical institutions are particularly concerned with data privacy when sharing and transmitting sensitive patient information due to the risk of data breaches, where malicious actors could intercept sensitive information. A collaborative learning framework, including transfer, federated, and incremental learning, can generate efficient, secure, and scalable models while requiring less computation, maintaining patient data privacy, and ensuring an up-to-date model. This study aims to address the detection of COVID-19 using chest X-ray images through a proposed collaborative learning framework called CL3. Initially, transfer learning is employed, leveraging knowledge from a pre-trained model as the starting global model. Local models from different medical institutes are then integrated, and a new global model is constructed to adapt to any data drift observed in the local models. Additionally, incremental learning is considered, allowing continuous adaptation to new medical data without forgetting previously learned information. Experimental results demonstrate that the CL3 framework achieved a global accuracy of 89.99% when using Xception with a batch size of 16 after being trained for six federated communication rounds. A demo of the CL3 framework is available at https://github.com/zavidparvez/CL3-Collaborative-Approach to ensure reproducibility.
♻ ☆ G2D: From Global to Dense Radiography Representation Learning via Vision-Language Pre-training NeurIPS2024
Recently, medical vision-language pre-training (VLP) has reached substantial progress to learn global visual representation from medical images and their paired radiology reports. However, medical imaging tasks in real world usually require finer granularity in visual features. These tasks include visual localization tasks (e.g., semantic segmentation, object detection) and visual grounding task. Yet, current medical VLP methods face challenges in learning these fine-grained features, as they primarily focus on brute-force alignment between image patches and individual text tokens for local visual feature learning, which is suboptimal for downstream dense prediction tasks. In this work, we propose a new VLP framework, named \textbf{G}lobal to \textbf{D}ense level representation learning (G2D) that achieves significantly improved granularity and more accurate grounding for the learned features, compared to existing medical VLP approaches. In particular, G2D learns dense and semantically-grounded image representations via a pseudo segmentation task parallel with the global vision-language alignment. Notably, generating pseudo segmentation targets does not incur extra trainable parameters: they are obtained on the fly during VLP with a parameter-free processor. G2D achieves superior performance across 6 medical imaging tasks and 25 diseases, particularly in semantic segmentation, which necessitates fine-grained, semantically-grounded image features. In this task, G2D surpasses peer models even when fine-tuned with just 1\% of the training data, compared to the 100\% used by these models. The code will be released upon acceptance.
comment: Accepted by NeurIPS2024
♻ ☆ FLOPS: Forward Learning with OPtimal Sampling
Given the limitations of backpropagation, perturbation-based gradient computation methods have recently gained focus for learning with only forward passes, also referred to as queries. Conventional forward learning consumes enormous queries on each data point for accurate gradient estimation through Monte Carlo sampling, which hinders the scalability of those algorithms. However, not all data points deserve equal queries for gradient estimation. In this paper, we study the problem of improving the forward learning efficiency from a novel perspective: how to reduce the gradient estimation variance with minimum cost? For this, we propose to allocate the optimal number of queries over each data in one batch during training to achieve a good balance between estimation accuracy and computational efficiency. Specifically, with a simplified proxy objective and a reparameterization technique, we derive a novel plug-and-play query allocator with minimal parameters. Theoretical results are carried out to verify its optimality. We conduct extensive experiments for fine-tuning Vision Transformers on various datasets and further deploy the allocator to two black-box applications: prompt tuning and multimodal alignment for foundation models. All findings demonstrate that our proposed allocator significantly enhances the scalability of forward-learning algorithms, paving the way for real-world applications.
♻ ☆ t-READi: Transformer-Powered Robust and Efficient Multimodal Inference for Autonomous Driving
Given the wide adoption of multimodal sensors (e.g., camera, lidar, radar) by autonomous vehicles (AVs), deep analytics to fuse their outputs for a robust perception become imperative. However, existing fusion methods often make two assumptions rarely holding in practice: i) similar data distributions for all inputs and ii) constant availability for all sensors. Because, for example, lidars have various resolutions and failures of radars may occur, such variability often results in significant performance degradation in fusion. To this end, we present tREADi, an adaptive inference system that accommodates the variability of multimodal sensory data and thus enables robust and efficient perception. t-READi identifies variation-sensitive yet structure-specific model parameters; it then adapts only these parameters while keeping the rest intact. t-READi also leverages a cross-modality contrastive learning method to compensate for the loss from missing modalities. Both functions are implemented to maintain compatibility with existing multimodal deep fusion methods. The extensive experiments evidently demonstrate that compared with the status quo approaches, t-READi not only improves the average inference accuracy by more than 6% but also reduces the inference latency by almost 15x with the cost of only 5% extra memory overhead in the worst case under realistic data and modal variations.
comment: 14 pages, 16 figures
♻ ☆ Online-to-PAC Conversions: Generalization Bounds via Regret Analysis
We present a new framework for deriving bounds on the generalization bound of statistical learning algorithms from the perspective of online learning. Specifically, we construct an online learning game called the "generalization game", where an online learner is trying to compete with a fixed statistical learning algorithm in predicting the sequence of generalization gaps on a training set of i.i.d. data points. We establish a connection between the online and statistical learning setting by showing that the existence of an online learning algorithm with bounded regret in this game implies a bound on the generalization error of the statistical learning algorithm, up to a martingale concentration term that is independent of the complexity of the statistical learning method. This technique allows us to recover several standard generalization bounds including a range of PAC-Bayesian and information-theoretic guarantees, as well as generalizations thereof.
♻ ☆ Robust Fast Adaptation from Adversarially Explicit Task Distribution Generation
Meta-learning is a practical learning paradigm to transfer skills across tasks from a few examples. Nevertheless, the existence of task distribution shifts tends to weaken meta-learners' generalization capability, particularly when the task distribution is naively hand-crafted or based on simple priors that fail to cover typical scenarios sufficiently. Here, we consider explicitly generative modeling task distributions placed over task identifiers and propose robustifying fast adaptation from adversarial training. Our approach, which can be interpreted as a model of a Stackelberg game, not only uncovers the task structure during problem-solving from an explicit generative model but also theoretically increases the adaptation robustness in worst cases. This work has practical implications, particularly in dealing with task distribution shifts in meta-learning, and contributes to theoretical insights in the field. Our method demonstrates its robustness in the presence of task subpopulation shifts and improved performance over SOTA baselines in extensive experiments. The project is available at https://sites.google.com/view/ar-metalearn.
comment: The project is available at https://sites.google.com/view/ar-metalearn
♻ ☆ Leveraging Invariant Principle for Heterophilic Graph Structure Distribution Shifts
Heterophilic Graph Neural Networks (HGNNs) have shown promising results for semi-supervised learning tasks on graphs. Notably, most real-world heterophilic graphs are composed of a mixture of nodes with different neighbor patterns, exhibiting local node-level homophilic and heterophilic structures. However, existing works are only devoted to designing better HGNN backbones or architectures for node classification tasks on heterophilic and homophilic graph benchmarks simultaneously, and their analyses of HGNN performance with respect to nodes are only based on the determined data distribution without exploring the effect caused by this structural difference between training and testing nodes. How to learn invariant node representations on heterophilic graphs to handle this structure difference or distribution shifts remains unexplored. In this paper, we first discuss the limitations of previous graph-based invariant learning methods from the perspective of data augmentation. Then, we propose \textbf{HEI}, a framework capable of generating invariant node representations through incorporating heterophily information to infer latent environments without augmentation, which are then used for invariant prediction, under heterophilic graph structure distribution shifts. We theoretically show that our proposed method can achieve guaranteed performance under heterophilic graph structure distribution shifts. Extensive experiments on various benchmarks and backbones can also demonstrate the effectiveness of our method compared with existing state-of-the-art baselines.
comment: 15 pages, 7 figures
♻ ☆ Advancing RVFL networks: Robust classification with the HawkEye loss function
Random vector functional link (RVFL), a variant of single-layer feedforward neural network (SLFN), has garnered significant attention due to its lower computational cost and robustness to overfitting. Despite its advantages, the RVFL network's reliance on the square error loss function makes it highly sensitive to outliers and noise, leading to degraded model performance in real-world applications. To remedy it, we propose the incorporation of the HawkEye loss (H-loss) function into the RVFL framework. The H-loss function features nice mathematical properties, including smoothness and boundedness, while simultaneously incorporating an insensitive zone. Each characteristic brings its own advantages: 1) Boundedness limits the impact of extreme errors, enhancing robustness against outliers; 2) Smoothness facilitates the use of gradient-based optimization algorithms, ensuring stable and efficient convergence; and 3) The insensitive zone mitigates the effect of minor discrepancies and noise. Leveraging the H-loss function, we embed it into the RVFL framework and develop a novel robust RVFL model termed H-RVFL. Notably, this work addresses a significant gap, as no bounded loss function has been incorporated into RVFL to date. The non-convex optimization of the proposed H-RVFL is effectively addressed by the Nesterov accelerated gradient (NAG) algorithm, whose computational complexity is also discussed. The proposed H-RVFL model's effectiveness is validated through extensive experiments on $40$ benchmark datasets from UCI and KEEL repositories, with and without label noise. The results highlight significant improvements in robustness and efficiency, establishing the H-RVFL model as a powerful tool for applications in noisy and outlier-prone environments.
♻ ☆ Reducing Bias in Federated Class-Incremental Learning with Hierarchical Generative Prototypes
Federated Learning (FL) aims at unburdening the training of deep models by distributing computation across multiple devices (clients) while safeguarding data privacy. On top of that, Federated Continual Learning (FCL) also accounts for data distribution evolving over time, mirroring the dynamic nature of real-world environments. In this work, we shed light on the Incremental and Federated biases that naturally emerge in FCL. While the former is a known problem in Continual Learning, stemming from the prioritization of recently introduced classes, the latter (i.e., the bias towards local distributions) remains relatively unexplored. Our proposal constrains both biases in the last layer by efficiently fine-tuning a pre-trained backbone using learnable prompts, resulting in clients that produce less biased representations and more biased classifiers. Therefore, instead of solely relying on parameter aggregation, we also leverage generative prototypes to effectively balance the predictions of the global model. Our method improves on the current State Of The Art, providing an average increase of +7.9% in accuracy.
♻ ☆ Decision Mamba Architectures
Recent advancements in imitation learning have been largely fueled by the integration of sequence models, which provide a structured flow of information to effectively mimic task behaviours. Currently, Decision Transformer (DT) and subsequently, the Hierarchical Decision Transformer (HDT), presented Transformer-based approaches to learn task policies. Recently, the Mamba architecture has shown to outperform Transformers across various task domains. In this work, we introduce two novel methods, Decision Mamba (DM) and Hierarchical Decision Mamba (HDM), aimed at enhancing the performance of the Transformer models. Through extensive experimentation across diverse environments such as OpenAI Gym and D4RL, leveraging varying demonstration data sets, we demonstrate the superiority of Mamba models over their Transformer counterparts in a majority of tasks. Results show that DM outperforms other methods in most settings. The code can be found at https://github.com/meowatthemoon/DecisionMamba.
♻ ☆ SmoothGNN: Smoothing-aware GNN for Unsupervised Node Anomaly Detection
The smoothing issue in graph learning leads to indistinguishable node representations, posing significant challenges for graph-related tasks. However, our experiments reveal that this problem can uncover underlying properties of node anomaly detection (NAD) that previous research has missed. We introduce Individual Smoothing Patterns (ISP) and Neighborhood Smoothing Patterns (NSP), which indicate that the representations of anomalous nodes are harder to smooth than those of normal ones. In addition, we explore the theoretical implications of these patterns, demonstrating the potential benefits of ISP and NSP for NAD tasks. Motivated by these findings, we propose SmoothGNN, a novel unsupervised NAD framework. First, we design a learning component to explicitly capture ISP for detecting node anomalies. Second, we design a spectral graph neural network to implicitly learn ISP to enhance detection. Third, we design an effective coefficient based on our findings that NSP can serve as coefficients for node representations, aiding in the identification of anomalous nodes. Furthermore, we devise a novel anomaly measure to calculate loss functions and anomalous scores for nodes, reflecting the properties of NAD using ISP and NSP. Extensive experiments on 9 real datasets show that SmoothGNN outperforms the best rival by an average of 14.66% in AUC and 7.28% in Average Precision, with 75x running time speedup, validating the effectiveness and efficiency of our framework.
♻ ☆ HC-GLAD: Dual Hyperbolic Contrastive Learning for Unsupervised Graph-Level Anomaly Detection
Unsupervised graph-level anomaly detection (UGAD) has garnered increasing attention in recent years due to its significance. Most existing methods that rely on traditional GNNs mainly consider pairwise relationships between first-order neighbors, which is insufficient to capture the complex high-order dependencies often associated with anomalies. This limitation underscores the necessity of exploring high-order node interactions in UGAD. In addition, most previous works ignore the underlying properties (e.g., hierarchy and power-law structure) which are common in real-world graph datasets and therefore are indispensable factors in the UGAD task. In this paper, we propose a novel Dual Hyperbolic Contrastive Learning for Unsupervised Graph-Level Anomaly Detection (HC-GLAD in short). To exploit high-order node group information, we construct hypergraphs based on pre-designed gold motifs and subsequently perform hypergraph convolution. Furthermore, to preserve the hierarchy of real-world graphs, we introduce hyperbolic geometry into this field and conduct both graph and hypergraph embedding learning in hyperbolic space with the hyperboloid model. To the best of our knowledge, this is the first work to simultaneously apply hypergraph with node group information and hyperbolic geometry in this field. Extensive experiments on 13 real-world datasets of different fields demonstrate the superiority of HC-GLAD on the UGAD task. The code is available at https://github.com/Yali-F/HC-GLAD.
♻ ☆ LLoCO: Learning Long Contexts Offline EMNLP 2024
Processing long contexts remains a challenge for large language models (LLMs) due to the quadratic computational and memory overhead of the self-attention mechanism and the substantial KV cache sizes during generation. We propose LLoCO, a novel approach to address this problem by learning contexts offline through context compression and in-domain parameter-efficient finetuning with LoRA. Our method enables an LLM to create a concise representation of the original context and efficiently retrieve relevant information to answer questions accurately. Our approach extends the effective context window of a 4k token LLaMA2-7B model to handle up to 128k tokens. We evaluate our approach on several long-context question-answering datasets, demonstrating that LLoCO significantly outperforms in-context learning while using $30\times$ fewer tokens during inference. LLoCO achieves up to $7.62\times$ speed-up during inference and $11.52\times$ higher throughput during finetuning, substantially reduces the cost of long document question answering. This makes it a promising solution for efficient long context processing. Our code is publicly available on https://github.com/jeffreysijuntan/lloco.
comment: EMNLP 2024. The first two authors contributed equally to this work
♻ ☆ Continuous-time q-Learning for Jump-Diffusion Models under Tsallis Entropy
This paper studies the continuous-time reinforcement learning in jump-diffusion models by featuring the q-learning (the continuous-time counterpart of Q-learning) under Tsallis entropy regularization. Contrary to the Shannon entropy, the general form of Tsallis entropy renders the optimal policy not necessary a Gibbs measure, where the Lagrange and KKT multipliers naturally arise from some constraints to ensure the learnt policy to be a probability density function. As a consequence, the characterization of the optimal policy using the q-function also involves a Lagrange multiplier. In response, we establish the martingale characterization of the q-function under Tsallis entropy and devise two q-learning algorithms depending on whether the Lagrange multiplier can be derived explicitly or not. In the latter case, we need to consider different parameterizations of the optimal q-function and the optimal policy and update them alternatively in an Actor-Critic manner. We also study two financial applications, namely, an optimal portfolio liquidation problem and a non-LQ control problem. It is interesting to see therein that the optimal policies under the Tsallis entropy regularization can be characterized explicitly, which are distributions concentrated on some compact support. The satisfactory performance of our q-learning algorithms is illustrated in each example.
♻ ☆ Stochastic Concept Bottleneck Models NeurIPS 2024
Concept Bottleneck Models (CBMs) have emerged as a promising interpretable method whose final prediction is based on intermediate, human-understandable concepts rather than the raw input. Through time-consuming manual interventions, a user can correct wrongly predicted concept values to enhance the model's downstream performance. We propose Stochastic Concept Bottleneck Models (SCBMs), a novel approach that models concept dependencies. In SCBMs, a single-concept intervention affects all correlated concepts, thereby improving intervention effectiveness. Unlike previous approaches that model the concept relations via an autoregressive structure, we introduce an explicit, distributional parameterization that allows SCBMs to retain the CBMs' efficient training and inference procedure. Additionally, we leverage the parameterization to derive an effective intervention strategy based on the confidence region. We show empirically on synthetic tabular and natural image datasets that our approach improves intervention effectiveness significantly. Notably, we showcase the versatility and usability of SCBMs by examining a setting with CLIP-inferred concepts, alleviating the need for manual concept annotations.
comment: Published at 38th Conference on Neural Information Processing Systems (NeurIPS 2024)
♻ ☆ Don't Label Twice: Quantity Beats Quality when Comparing Binary Classifiers on a Budget ICML 2024
We study how to best spend a budget of noisy labels to compare the accuracy of two binary classifiers. It's common practice to collect and aggregate multiple noisy labels for a given data point into a less noisy label via a majority vote. We prove a theorem that runs counter to conventional wisdom. If the goal is to identify the better of two classifiers, we show it's best to spend the budget on collecting a single label for more samples. Our result follows from a non-trivial application of Cram\'er's theorem, a staple in the theory of large deviations. We discuss the implications of our work for the design of machine learning benchmarks, where they overturn some time-honored recommendations. In addition, our results provide sample size bounds superior to what follows from Hoeffding's bound.
comment: 34 pages, 3 Figures, Published at ICML 2024
♻ ☆ Steerable Conditional Diffusion for Out-of-Distribution Adaptation in Medical Image Reconstruction
Denoising diffusion models have emerged as the go-to generative framework for solving inverse problems in imaging. A critical concern regarding these models is their performance on out-of-distribution tasks, which remains an under-explored challenge. Using a diffusion model on an out-of-distribution dataset, realistic reconstructions can be generated, but with hallucinating image features that are uniquely present in the training dataset. To address this discrepancy during train-test time and improve reconstruction accuracy, we introduce a novel sampling framework called Steerable Conditional Diffusion. Specifically, this framework adapts the diffusion model, concurrently with image reconstruction, based solely on the information provided by the available measurement. Utilising our proposed method, we achieve substantial enhancements in out-of-distribution performance across diverse imaging modalities, advancing the robust deployment of denoising diffusion models in real-world applications.
♻ ☆ Leveraging LLM Embeddings for Cross Dataset Label Alignment and Zero Shot Music Emotion Prediction
In this work, we present a novel method for music emotion recognition that leverages Large Language Model (LLM) embeddings for label alignment across multiple datasets and zero-shot prediction on novel categories. First, we compute LLM embeddings for emotion labels and apply non-parametric clustering to group similar labels, across multiple datasets containing disjoint labels. We use these cluster centers to map music features (MERT) to the LLM embedding space. To further enhance the model, we introduce an alignment regularization that enables dissociation of MERT embeddings from different clusters. This further enhances the model's ability to better adaptation to unseen datasets. We demonstrate the effectiveness of our approach by performing zero-shot inference on a new dataset, showcasing its ability to generalize to unseen labels without additional training.
♻ ☆ Generative Model for Constructing Reaction Path from Initial to Final States
Mapping the chemical reaction pathways and their corresponding activation barriers is a significant challenge in molecular simulation. Given the inherent complexities of 3D atomic geometries, even generating an initial guess of these paths can be difficult for humans. This paper presents an innovative approach that utilizes neural networks to generate initial guesses for reaction pathways based on the initial state and learning from a database of low-energy transition paths. The proposed method is initiated by inputting the coordinates of the initial state, followed by progressive alterations to its structure. This iterative process culminates in the generation of the guess reaction path and the coordinates of the final state. The method does not require one-the-fly computation of the actual potential energy surface, and is therefore fast-acting. The application of this geometry-based method extends to complex reaction pathways illustrated by organic reactions. Training was executed on the Transition1x dataset of organic reaction pathways. The results revealed the generation of reactions that bore substantial similarities with the test set of chemical reaction paths. The method's flexibility allows for reactions to be generated either to conform to predetermined conditions or in a randomized manner.
♻ ☆ Feature learning in finite-width Bayesian deep linear networks with multiple outputs and convolutional layers
Deep linear networks have been extensively studied, as they provide simplified models of deep learning. However, little is known in the case of finite-width architectures with multiple outputs and convolutional layers. In this manuscript, we provide rigorous results for the statistics of functions implemented by the aforementioned class of networks, thus moving closer to a complete characterization of feature learning in the Bayesian setting. Our results include: (i) an exact and elementary non-asymptotic integral representation for the joint prior distribution over the outputs, given in terms of a mixture of Gaussians; (ii) an analytical formula for the posterior distribution in the case of squared error loss function (Gaussian likelihood); (iii) a quantitative description of the feature learning infinite-width regime, using large deviation theory. From a physical perspective, deep architectures with multiple outputs or convolutional layers represent different manifestations of kernel shape renormalization, and our work provides a dictionary that translates this physics intuition and terminology into rigorous Bayesian statistics.
♻ ☆ Benchmarking Data Efficiency in $Δ$-ML and Multifidelity Models for Quantum Chemistry
The development of machine learning (ML) methods has made quantum chemistry (QC) calculations more accessible by reducing the compute cost incurred in conventional QC methods. This has since been translated into the overhead cost of generating training data. Increased work in reducing the cost of generating training data resulted in the development of $\Delta$-ML and multifidelity machine learning methods which use data at more than one QC level of accuracy, or fidelity. This work compares the data costs associated with $\Delta$-ML, multifidelity machine learning (MFML), and optimized MFML (o-MFML) in contrast with a newly introduced Multifidelity$\Delta$-Machine Learning (MF$\Delta$ML) method for the prediction of ground state energies over the multifidelity benchmark dataset QeMFi. This assessment is made on the basis of training data generation cost associated with each model and is compared with the single fidelity kernel ridge regression (KRR) case. The results indicate that the use of multifidelity methods surpasses the standard $\Delta$-ML approaches in cases of a large number of predictions. For cases, where $\Delta$-ML method might be favored, such as small test set regimes, the MF$\Delta$-ML method is shown to be more efficient than conventional $\Delta$-ML.
comment: Supplementary information (sections S1,S2, and figure S1) included; v2: fixed some refs in the SI. Results and main text remain unchanged
♻ ☆ ASTM :Autonomous Smart Traffic Management System Using Artificial Intelligence CNN and LSTM
In the modern world, the development of Artificial Intelligence (AI) has contributed to improvements in various areas, including automation, computer vision, fraud detection, and more. AI can be leveraged to enhance the efficiency of Autonomous Smart Traffic Management (ASTM) systems and reduce traffic congestion rates. This paper presents an Autonomous Smart Traffic Management (STM) system that uses AI to improve traffic flow rates. The system employs the YOLO V5 Convolutional Neural Network to detect vehicles in traffic management images. Additionally, it predicts the number of vehicles for the next 12 hours using a Recurrent Neural Network with Long Short-Term Memory (RNN-LSTM). The Smart Traffic Management Cycle Length Analysis manages the traffic cycle length based on these vehicle predictions, aided by AI. From the results of the RNN-LSTM model for predicting vehicle numbers over the next 12 hours, we observe that the model predicts traffic with a Mean Squared Error (MSE) of 4.521 vehicles and a Root Mean Squared Error (RMSE) of 2.232 vehicles. After simulating the STM system in the CARLA simulation environment, we found that the Traffic Management Congestion Flow Rate with ASTM (21 vehicles per minute) is 50\% higher than the rate without STM (around 15 vehicles per minute). Additionally, the Traffic Management Vehicle Pass Delay with STM (5 seconds per vehicle) is 70\% lower than without STM (around 12 seconds per vehicle). These results demonstrate that the STM system using AI can increase traffic flow by 50\% and reduce vehicle pass delays by 70\%.
comment: In process to IEEE Intelligent Vehicle Symposium 2025
♻ ☆ LinFusion: 1 GPU, 1 Minute, 16K Image
Modern diffusion models, particularly those utilizing a Transformer-based UNet for denoising, rely heavily on self-attention operations to manage complex spatial relationships, thus achieving impressive generation performance. However, this existing paradigm faces significant challenges in generating high-resolution visual content due to its quadratic time and memory complexity with respect to the number of spatial tokens. To address this limitation, we aim at a novel linear attention mechanism as an alternative in this paper. Specifically, we begin our exploration from recently introduced models with linear complexity, e.g., Mamba2, RWKV6, Gated Linear Attention, etc, and identify two key features--attention normalization and non-causal inference--that enhance high-resolution visual generation performance. Building on these insights, we introduce a generalized linear attention paradigm, which serves as a low-rank approximation of a wide spectrum of popular linear token mixers. To save the training cost and better leverage pre-trained models, we initialize our models and distill the knowledge from pre-trained StableDiffusion (SD). We find that the distilled model, termed LinFusion, achieves performance on par with or superior to the original SD after only modest training, while significantly reducing time and memory complexity. Extensive experiments on SD-v1.5, SD-v2.1, and SD-XL demonstrate that LinFusion enables satisfactory and efficient zero-shot cross-resolution generation, accommodating ultra-resolution images like 16K on a single GPU. Moreover, it is highly compatible with pre-trained SD components and pipelines, such as ControlNet, IP-Adapter, DemoFusion, DistriFusion, etc, requiring no adaptation efforts. Codes are available at https://github.com/Huage001/LinFusion.
comment: Work in Progress. Codes are available at https://github.com/Huage001/LinFusion
♻ ☆ Grounding Large Language Models in Interactive Environments with Online Reinforcement Learning
Recent works successfully leveraged Large Language Models' (LLM) abilities to capture abstract knowledge about world's physics to solve decision-making problems. Yet, the alignment between LLMs' knowledge and the environment can be wrong and limit functional competence due to lack of grounding. In this paper, we study an approach (named GLAM) to achieve this alignment through functional grounding: we consider an agent using an LLM as a policy that is progressively updated as the agent interacts with the environment, leveraging online Reinforcement Learning to improve its performance to solve goals. Using an interactive textual environment designed to study higher-level forms of functional grounding, and a set of spatial and navigation tasks, we study several scientific questions: 1) Can LLMs boost sample efficiency for online learning of various RL tasks? 2) How can it boost different forms of generalization? 3) What is the impact of online learning? We study these questions by functionally grounding several variants (size, architecture) of FLAN-T5.
♻ ☆ A Cost-Efficient FPGA Implementation of Tiny Transformer Model using Neural ODE
Transformer has been adopted to image recognition tasks and shown to outperform CNNs and RNNs while it suffers from high training cost and computational complexity. To address these issues, a hybrid approach has become a recent research trend, which replaces a part of ResNet with an MHSA (Multi-Head Self-Attention). In this paper, we propose a lightweight hybrid model which uses Neural ODE (Ordinary Differential Equation) as a backbone instead of ResNet so that we can increase the number of iterations of building blocks while reusing the same parameters, mitigating the increase in parameter size per iteration. The proposed model is deployed on a modest-sized FPGA device for edge computing. The model is further quantized by QAT (Quantization Aware Training) scheme to reduce FPGA resource utilization while suppressing the accuracy loss. The quantized model achieves 79.68% top-1 accuracy for STL10 dataset that contains 96$\times$96 pixel images. The weights of the feature extraction network are stored on-chip to minimize the memory transfer overhead, allowing faster inference. By eliminating the overhead of memory transfers, inference can be executed seamlessly, leading to accelerated inference. The proposed FPGA implementation accelerates the backbone and MHSA parts by 34.01$\times$, and achieves an overall 9.85$\times$ speedup when taking into account the software pre- and post-processing. The FPGA acceleration leads to 7.10$\times$ better energy efficiency compared to the ARM Cortex-A53 CPU. The proposed lightweight Transformer model is demonstrated on Xilinx ZCU104 board for the image recognition of 96$\times$96 pixel images in this paper and can be applied to different image sizes by modifying the pre-processing layer.
♻ ☆ Loss Landscape Characterization of Neural Networks without Over-Parametrization
Optimization methods play a crucial role in modern machine learning, powering the remarkable empirical achievements of deep learning models. These successes are even more remarkable given the complex non-convex nature of the loss landscape of these models. Yet, ensuring the convergence of optimization methods requires specific structural conditions on the objective function that are rarely satisfied in practice. One prominent example is the widely recognized Polyak-Lojasiewicz (PL) inequality, which has gained considerable attention in recent years. However, validating such assumptions for deep neural networks entails substantial and often impractical levels of over-parametrization. In order to address this limitation, we propose a novel class of functions that can characterize the loss landscape of modern deep models without requiring extensive over-parametrization and can also include saddle points. Crucially, we prove that gradient-based optimizers possess theoretical guarantees of convergence under this assumption. Finally, we validate the soundness of our new function class through both theoretical analysis and empirical experimentation across a diverse range of deep learning models.
♻ ☆ ClickAgent: Enhancing UI Location Capabilities of Autonomous Agents
With the growing reliance on digital devices equipped with graphical user interfaces (GUIs), such as computers and smartphones, the need for effective automation tools has become increasingly important. While multimodal large language models (MLLMs) like GPT-4V excel in many areas, they struggle with GUI interactions, limiting their effectiveness in automating everyday tasks. In this paper, we introduce ClickAgent, a novel framework for building autonomous agents. In ClickAgent, the MLLM handles reasoning and action planning, while a separate UI location model (e.g., SeeClick) identifies the relevant UI elements on the screen. This approach addresses a key limitation of current-generation MLLMs: their difficulty in accurately locating UI elements. ClickAgent outperforms other prompt-based autonomous agents (CogAgent, AppAgent) on the AITW benchmark. Our evaluation was conducted on both an Android smartphone emulator and an actual Android smartphone, using the task success rate as the key metric for measuring agent performance.
comment: The code for ClickAgent is available at github.com/Samsung/ClickAgent
♻ ☆ Clustering and Data Augmentation to Improve Accuracy of Sleep Assessment and Sleep Individuality Analysis
Recently, growing health awareness, novel methods allow individuals to monitor sleep at home. Utilizing sleep sounds offers advantages over conventional methods like smartwatches, being non-intrusive, and capable of detecting various physiological activities. This study aims to construct a machine learning-based sleep assessment model providing evidence-based assessments, such as poor sleep due to frequent movement during sleep onset. Extracting sleep sound events, deriving latent representations using VAE, clustering with GMM, and training LSTM for subjective sleep assessment achieved a high accuracy of 94.8% in distinguishing sleep satisfaction. Moreover, TimeSHAP revealed differences in impactful sound event types and timings for different individuals.
♻ ☆ Evolutionary Computation and Explainable AI: A Roadmap to Understandable Intelligent Systems
Artificial intelligence methods are being increasingly applied across various domains, but their often opaque nature has raised concerns about accountability and trust. In response, the field of explainable AI (XAI) has emerged to address the need for human-understandable AI systems. Evolutionary computation (EC), a family of powerful optimization and learning algorithms, offers significant potential to contribute to XAI, and vice versa. This paper provides an introduction to XAI and reviews current techniques for explaining machine learning models. We then explore how EC can be leveraged in XAI and examine existing XAI approaches that incorporate EC techniques. Furthermore, we discuss the application of XAI principles within EC itself, investigating how these principles can illuminate the behavior and outcomes of EC algorithms, their (automatic) configuration, and the underlying problem landscapes they optimize. Finally, we discuss open challenges in XAI and highlight opportunities for future research at the intersection of XAI and EC. Our goal is to demonstrate EC's suitability for addressing current explainability challenges and to encourage further exploration of these methods, ultimately contributing to the development of more understandable and trustworthy ML models and EC algorithms.
comment: 24 pages, 4 figures. arXiv admin note: substantial text overlap with arXiv:2306.14786
♻ ☆ Context-Aware Adapter Tuning for Few-Shot Relation Learning in Knowledge Graphs EMNLP 2024
Knowledge graphs (KGs) are instrumental in various real-world applications, yet they often suffer from incompleteness due to missing relations. To predict instances for novel relations with limited training examples, few-shot relation learning approaches have emerged, utilizing techniques such as meta-learning. However, the assumption is that novel relations in meta-testing and base relations in meta-training are independently and identically distributed, which may not hold in practice. To address the limitation, we propose RelAdapter, a context-aware adapter for few-shot relation learning in KGs designed to enhance the adaptation process in meta-learning. First, RelAdapter is equipped with a lightweight adapter module that facilitates relation-specific, tunable adaptation of meta-knowledge in a parameter-efficient manner. Second, RelAdapter is enriched with contextual information about the target relation, enabling enhanced adaptation to each distinct relation. Extensive experiments on three benchmark KGs validate the superiority of RelAdapter over state-of-the-art methods.
comment: Accepted by EMNLP 2024
♻ ☆ Temporally Consistent Factuality Probing for Large Language Models
The prolific use of Large Language Models (LLMs) as an alternate knowledge base requires them to be factually consistent, necessitating both correctness and consistency traits for paraphrased queries. Recently, significant attempts have been made to benchmark datasets and metrics to evaluate LLMs for these traits. However, structural simplicity (subject-relation-object) and contemporary association in their query formulation limit the broader definition of factuality and consistency. In this study, we introduce TeCFaP, a novel Temporally Consistent Factuality Probe task to expand the consistent factuality probe in the temporal dimension. To this end, we propose TEMP-COFAC, a high-quality dataset of prefix-style English query paraphrases. Subsequently, we extend the definitions of existing metrics to represent consistent factuality across temporal dimension. We experiment with a diverse set of LLMs and find most of them performing poorly on TeCFaP. Next, we propose a novel solution CoTSeLF (Consistent-Time-Sensitive Learning Framework) combining multi-task instruction tuning (MT-IT) with consistent-time-sensitive reinforcement learning (CTSRL) to improve temporally consistent factuality in LLMs. Our experiments demonstrate the efficacy of CoTSeLF over several baselines.
♻ ☆ Memory-Efficient Optimization with Factorized Hamiltonian Descent
Modern deep learning heavily depends on adaptive optimizers such as Adam and its variants, which are renowned for their capacity to handle model scaling and streamline hyperparameter tuning. However, these algorithms typically experience high memory overhead caused by the accumulation of optimization states, leading to a critical challenge in training large-scale network models. In this study, we introduce a novel adaptive optimizer, H-Fac, which incorporates a memory-efficient factorization approach to address this challenge. By employing a rank-1 parameterization for both momentum and scaling parameter estimators, H-Fac reduces memory costs to a sublinear level while maintaining competitive performance across a wide range of architectures. We develop our algorithms based on principles derived from Hamiltonian dynamics, providing robust theoretical underpinnings in optimization dynamics and convergence guarantees. These optimization algorithms are designed to be both straightforward and adaptable, facilitating easy implementation in diverse settings.
comment: 16 pages, 5 figures
♻ ☆ Variational Source-Channel Coding for Semantic Communication
Semantic communication technology emerges as a pivotal bridge connecting AI with classical communication. The current semantic communication systems are generally modeled as an Auto-Encoder (AE). AE lacks a deep integration of AI principles with communication strategies due to its inability to effectively capture channel dynamics. This gap makes it difficult to justify the need for joint source-channel coding (JSCC) and to explain why performance improves. This paper begins by exploring lossless and lossy communication, highlighting that the inclusion of data distortion distinguishes semantic communication from classical communication. It breaks the conditions for the separation theorem to hold and explains why the amount of data transferred by semantic communication is less. Therefore, employing JSCC becomes imperative for achieving optimal semantic communication. Moreover, a Variational Source-Channel Coding (VSCC) method is proposed for constructing semantic communication systems based on data distortion theory, integrating variational inference and channel characteristics. Using a deep learning network, we develop a semantic communication system employing the VSCC method and demonstrate its capability for semantic transmission. We also establish semantic communication systems of equivalent complexity employing the AE method and the VAE method. Experimental results reveal that the VSCC model offers superior interpretability compared to AE model, as it clearly captures the semantic features of the transmitted data, represented as the variance of latent variables in our experiments. In addition, VSCC model exhibits superior semantic transmission capabilities compared to VAE model. At the same level of data distortion evaluated by PSNR, VSCC model exhibits stronger human interpretability, which can be partially assessed by SSIM.
♻ ☆ DAQ: Density-Aware Post-Training Weight-Only Quantization For LLMs
Large language models (LLMs) excel in various tasks but face deployment challenges due to hardware constraints. We propose density-aware post-training weight-only quantization (DAQ), which has two stages: 1) density-centric alignment, which identifies the center of high-density weights and centers the dynamic range on this point to align high-density weight regions with floating-point high-precision regions; 2) learnable dynamic range adjustment, which adjusts the dynamic range by optimizing quantization parameters (i.e., scale and zero-point) based on the impact of weights on the model output. Experiments on LLaMA and LLaMA-2 show that DAQ consistently outperforms the best baseline method, reducing perplexity loss by an average of 22.8% on LLaMA and 19.6% on LLaMA-2. Our code is available at https://github.com/LuoYingSong/DAQ.
comment: 9 pages, 4 figures
♻ ☆ Fairness-aware Federated Minimax Optimization with Convergence Guarantee
Federated learning (FL) has garnered considerable attention due to its privacy-preserving feature. Nonetheless, the lack of freedom in managing user data can lead to group fairness issues, where models are biased towards sensitive factors such as race or gender. To tackle this issue, this paper proposes a novel algorithm, fair federated averaging with augmented Lagrangian method (FFALM), designed explicitly to address group fairness issues in FL. Specifically, we impose a fairness constraint on the training objective and solve the minimax reformulation of the constrained optimization problem. Then, we derive the theoretical upper bound for the convergence rate of FFALM. The effectiveness of FFALM in improving fairness is shown empirically on CelebA and UTKFace datasets in the presence of severe statistical heterogeneity.
♻ ☆ Context Matters: Leveraging Contextual Features for Time Series Forecasting
Time series forecasts are often influenced by exogenous contextual features in addition to their corresponding history. For example, in financial settings, it is hard to accurately predict a stock price without considering public sentiments and policy decisions in the form of news articles, tweets, etc. Though this is common knowledge, the current state-of-the-art (SOTA) forecasting models fail to incorporate such contextual information, owing to its heterogeneity and multimodal nature. To address this, we introduce ContextFormer, a novel plug-and-play method to surgically integrate multimodal contextual information into existing pre-trained forecasting models. ContextFormer effectively distills forecast-specific information from rich multimodal contexts, including categorical, continuous, time-varying, and even textual information, to significantly enhance the performance of existing base forecasters. ContextFormer outperforms SOTA forecasting models by up to 30% on a range of real-world datasets spanning energy, traffic, environmental, and financial domains.
♻ ☆ Multi-CATE: Multi-Accurate Conditional Average Treatment Effect Estimation Robust to Unknown Covariate Shifts
Estimating heterogeneous treatment effects is important to tailor treatments to those individuals who would most likely benefit. However, conditional average treatment effect predictors may often be trained on one population but possibly deployed on different, possibly unknown populations. We use methodology for learning multi-accurate predictors to post-process CATE T-learners (differenced regressions) to become robust to unknown covariate shifts at the time of deployment. The method works in general for pseudo-outcome regression, such as the DR-learner. We show how this approach can combine (large) confounded observational and (smaller) randomized datasets by learning a confounded predictor from the observational dataset, and auditing for multi-accuracy on the randomized controlled trial. We show improvements in bias and mean squared error in simulations with increasingly larger covariate shift, and on a semi-synthetic case study of a parallel large observational study and smaller randomized controlled experiment. Overall, we establish a connection between methods developed for multi-distribution learning and achieve appealing desiderata (e.g. external validity) in causal inference and machine learning.
♻ ☆ Faster Diffusion Sampling with Randomized Midpoints: Sequential and Parallel
Sampling algorithms play an important role in controlling the quality and runtime of diffusion model inference. In recent years, a number of works~\cite{chen2023sampling,chen2023ode,benton2023error,lee2022convergence} have proposed schemes for diffusion sampling with provable guarantees; these works show that for essentially any data distribution, one can approximately sample in polynomial time given a sufficiently accurate estimate of its score functions at different noise levels. In this work, we propose a new scheme inspired by Shen and Lee's randomized midpoint method for log-concave sampling~\cite{ShenL19}. We prove that this approach achieves the best known dimension dependence for sampling from arbitrary smooth distributions in total variation distance ($\widetilde O(d^{5/12})$ compared to $\widetilde O(\sqrt{d})$ from prior work). We also show that our algorithm can be parallelized to run in only $\widetilde O(\log^2 d)$ parallel rounds, constituting the first provable guarantees for parallel sampling with diffusion models. As a byproduct of our methods, for the well-studied problem of log-concave sampling in total variation distance, we give an algorithm and simple analysis achieving dimension dependence $\widetilde O(d^{5/12})$ compared to $\widetilde O(\sqrt{d})$ from prior work.
♻ ☆ MagR: Weight Magnitude Reduction for Enhancing Post-Training Quantization NeurIPS 2024
In this paper, we present a simple optimization-based preprocessing technique called Weight Magnitude Reduction (MagR) to improve the performance of post-training quantization. For each linear layer, we adjust the pre-trained floating-point weights by solving an $\ell_\infty$-regularized optimization problem. This process greatly diminishes the maximum magnitude of the weights and smooths out outliers, while preserving the layer's output. The preprocessed weights are centered more towards zero, which facilitates the subsequent quantization process. To implement MagR, we address the $\ell_\infty$-regularization by employing an efficient proximal gradient descent algorithm. Unlike existing preprocessing methods that involve linear transformations and subsequent post-processing steps, which can introduce significant overhead at inference time, MagR functions as a non-linear transformation, eliminating the need for any additional post-processing. This ensures that MagR introduces no overhead whatsoever during inference. Our experiments demonstrate that MagR achieves state-of-the-art performance on the Llama family of models. For example, we achieve a Wikitext2 perplexity of 5.95 on the LLaMA2-70B model for per-channel INT2 weight quantization without incurring any inference overhead.
comment: Accepted by NeurIPS 2024
♻ ☆ Tables as Texts or Images: Evaluating the Table Reasoning Ability of LLMs and MLLMs ACL 2024
In this paper, we investigate the effectiveness of various LLMs in interpreting tabular data through different prompting strategies and data formats. Our analyses extend across six benchmarks for table-related tasks such as question-answering and fact-checking. We introduce for the first time the assessment of LLMs' performance on image-based table representations. Specifically, we compare five text-based and three image-based table representations, demonstrating the role of representation and prompting on LLM performance. Our study provides insights into the effective use of LLMs on table-related tasks.
comment: Accepted to ACL 2024 Findings; Naihao and Zhenjie contributed equally to the project; Data available at: https://github.com/dnaihao/Tables-as-Texts-or-Images
♻ ☆ Global convergence of gradient descent for phase retrieval
We propose a tensor-based criterion for benign landscape in phase retrieval and establish boundedness of gradient trajectories. This implies that gradient descent will converge to a global minimum for almost every initial point.
♻ ☆ Depth-supervised NeRF: Fewer Views and Faster Training for Free DSN
A commonly observed failure mode of Neural Radiance Field (NeRF) is fitting incorrect geometries when given an insufficient number of input views. One potential reason is that standard volumetric rendering does not enforce the constraint that most of a scene's geometry consist of empty space and opaque surfaces. We formalize the above assumption through DS-NeRF (Depth-supervised Neural Radiance Fields), a loss for learning radiance fields that takes advantage of readily-available depth supervision. We leverage the fact that current NeRF pipelines require images with known camera poses that are typically estimated by running structure-from-motion (SFM). Crucially, SFM also produces sparse 3D points that can be used as "free" depth supervision during training: we add a loss to encourage the distribution of a ray's terminating depth matches a given 3D keypoint, incorporating depth uncertainty. DS-NeRF can render better images given fewer training views while training 2-3x faster. Further, we show that our loss is compatible with other recently proposed NeRF methods, demonstrating that depth is a cheap and easily digestible supervisory signal. And finally, we find that DS-NeRF can support other types of depth supervision such as scanned depth sensors and RGB-D reconstruction outputs.
comment: Project page: http://www.cs.cmu.edu/~dsnerf/ GitHub: https://github.com/dunbar12138/DSNeRF
Multimedia 8
☆ MixEval-X: Any-to-Any Evaluations from Real-World Data Mixtures
Perceiving and generating diverse modalities are crucial for AI models to effectively learn from and engage with real-world signals, necessitating reliable evaluations for their development. We identify two major issues in current evaluations: (1) inconsistent standards, shaped by different communities with varying protocols and maturity levels; and (2) significant query, grading, and generalization biases. To address these, we introduce MixEval-X, the first any-to-any real-world benchmark designed to optimize and standardize evaluations across input and output modalities. We propose multi-modal benchmark mixture and adaptation-rectification pipelines to reconstruct real-world task distributions, ensuring evaluations generalize effectively to real-world use cases. Extensive meta-evaluations show our approach effectively aligns benchmark samples with real-world task distributions and the model rankings correlate strongly with that of crowd-sourced real-world evaluations (up to 0.98). We provide comprehensive leaderboards to rerank existing models and organizations and offer insights to enhance understanding of multi-modal evaluations and inform future research.
☆ Improving Multi-modal Large Language Model through Boosting Vision Capabilities
We focus on improving the visual understanding capability for boosting the vision-language models. We propose \textbf{Arcana}, a multiModal language model, which introduces two crucial techniques. First, we present Multimodal LoRA (MM-LoRA), a module designed to enhance the decoder. Unlike traditional language-driven decoders, MM-LoRA consists of two parallel LoRAs -- one for vision and one for language -- each with its own parameters. This disentangled parameters design allows for more specialized learning in each modality and better integration of multimodal information. Second, we introduce the Query Ladder adapter (QLadder) to improve the visual encoder. QLadder employs a learnable ``\textit{ladder}'' structure to deeply aggregates the intermediate representations from the frozen pretrained visual encoder (e.g., CLIP image encoder). This enables the model to learn new and informative visual features, as well as remaining the powerful capabilities of the pretrained visual encoder. These techniques collectively enhance Arcana's visual perception power, enabling it to leverage improved visual information for more accurate and contextually relevant outputs across various multimodal scenarios. Extensive experiments and ablation studies demonstrate the effectiveness and generalization capability of our Arcana. The code and re-annotated data are available at \url{https://arcana-project-page.github.io}.
☆ Multimodal growth and development assessment model
With the development of social economy and the improvement of people's attention to health, the growth and development of children and adolescents has become an important indicator to measure the level of national health. Therefore, accurate and timely assessment of children's growth and development has become increasingly important. At the same time, global health inequalities, especially child malnutrition and stunting in developing countries, urgently require effective assessment tools to monitor and intervene. In recent years, the rapid development of technologies such as big data, artificial intelligence, and cloud computing, and the cross-integration of multiple disciplines such as biomedicine, statistics, and computer science have promoted the rapid development of large-scale models for growth and development assessment. However, there are still problems such as too single evaluation factors, inaccurate diagnostic results, and inability to give accurate and reasonable recommendations. The multi-modal growth and development assessment model uses the public data set of RSNA ( North American College of Radiology ) as the training set, and the data set of the Department of Pediatrics of Huaibei People's Hospital as the open source test set. The embedded ICL module enables the model to quickly adapt and identify the tasks that need to be done to ensure that under the premise of considering multiple evaluation factors, accurate diagnosis results and reasonable medical recommendations are given, so as to provide solutions to the above problems and promote the development of the medical field.
comment: 7 Pages 7 Figures
☆ MeloTrans: A Text to Symbolic Music Generation Model Following Human Composition Habit
At present, neural network models show powerful sequence prediction ability and are used in many automatic composition models. In comparison, the way humans compose music is very different from it. Composers usually start by creating musical motifs and then develop them into music through a series of rules. This process ensures that the music has a specific structure and changing pattern. However, it is difficult for neural network models to learn these composition rules from training data, which results in a lack of musicality and diversity in the generated music. This paper posits that integrating the learning capabilities of neural networks with human-derived knowledge may lead to better results. To archive this, we develop the POP909$\_$M dataset, the first to include labels for musical motifs and their variants, providing a basis for mimicking human compositional habits. Building on this, we propose MeloTrans, a text-to-music composition model that employs principles of motif development rules. Our experiments demonstrate that MeloTrans excels beyond existing music generation models and even surpasses Large Language Models (LLMs) like ChatGPT-4. This highlights the importance of merging human insights with neural network capabilities to achieve superior symbolic music generation.
☆ Remember, Retrieve and Generate: Understanding Infinite Visual Concepts as Your Personalized Assistant
The development of large language models (LLMs) has significantly enhanced the capabilities of multimodal LLMs (MLLMs) as general assistants. However, lack of user-specific knowledge still restricts their application in human's daily life. In this paper, we introduce the Retrieval Augmented Personalization (RAP) framework for MLLMs' personalization. Starting from a general MLLM, we turn it into a personalized assistant in three steps. (a) Remember: We design a key-value database to store user-related information, e.g., user's name, avatar and other attributes. (b) Retrieve: When the user initiates a conversation, RAP will retrieve relevant information from the database using a multimodal retriever. (c) Generate: The input query and retrieved concepts' information are fed into MLLMs to generate personalized, knowledge-augmented responses. Unlike previous methods, RAP allows real-time concept editing via updating the external database. To further improve generation quality and alignment with user-specific information, we design a pipeline for data collection and create a specialized dataset for personalized training of MLLMs. Based on the dataset, we train a series of MLLMs as personalized multimodal assistants. By pretraining on large-scale dataset, RAP-MLLMs can generalize to infinite visual concepts without additional finetuning. Our models demonstrate outstanding flexibility and generation quality across a variety of tasks, such as personalized image captioning, question answering and visual recognition. The code, data and models are available at https://github.com/Hoar012/RAP-MLLM.
♻ ☆ Beyond Coarse-Grained Matching in Video-Text Retrieval ACCV 2024
Video-text retrieval has seen significant advancements, yet the ability of models to discern subtle differences in captions still requires verification. In this paper, we introduce a new approach for fine-grained evaluation. Our approach can be applied to existing datasets by automatically generating hard negative test captions with subtle single-word variations across nouns, verbs, adjectives, adverbs, and prepositions. We perform comprehensive experiments using four state-of-the-art models across two standard benchmarks (MSR-VTT and VATEX) and two specially curated datasets enriched with detailed descriptions (VLN-UVO and VLN-OOPS), resulting in a number of novel insights: 1) our analyses show that the current evaluation benchmarks fall short in detecting a model's ability to perceive subtle single-word differences, 2) our fine-grained evaluation highlights the difficulty models face in distinguishing such subtle variations. To enhance fine-grained understanding, we propose a new baseline that can be easily combined with current methods. Experiments on our fine-grained evaluations demonstrate that this approach enhances a model's ability to understand fine-grained differences.
comment: Accepted to ACCV 2024
♻ ☆ SaMoye: Zero-shot Singing Voice Conversion Model Based on Feature Disentanglement and Enhancement
Singing voice conversion (SVC) aims to convert a singer's voice to another singer's from a reference audio while keeping the original semantics. However, existing SVC methods can hardly perform zero-shot due to incomplete feature disentanglement or dependence on the speaker look-up table. We propose the first open-source high-quality zero-shot SVC model SaMoye that can convert singing to human and non-human timbre. SaMoye disentangles the singing voice's features into content, timbre, and pitch features, where we combine multiple ASR models and compress the content features to reduce timbre leaks. Besides, we enhance the timbre features by unfreezing the speaker encoder and mixing the speaker embedding with top-3 similar speakers. We also establish an unparalleled large-scale dataset to guarantee zero-shot performance, which comprises more than 1,815 hours of pure singing voice and 6,367 speakers. We conduct objective and subjective experiments to find that SaMoye outperforms other models in zero-shot SVC tasks even under extreme conditions like converting singing to animals' timbre. The code and weight of SaMoye are available on https://github.com/CarlWangChina/SaMoye-SVC. The weights, code, dataset, and documents of SaMoye are publicly available on \url{https://github.com/CarlWangChina/SaMoye-SVC}.
comment: 7 pages, 4 figures
♻ ☆ Leveraging LLM Embeddings for Cross Dataset Label Alignment and Zero Shot Music Emotion Prediction
In this work, we present a novel method for music emotion recognition that leverages Large Language Model (LLM) embeddings for label alignment across multiple datasets and zero-shot prediction on novel categories. First, we compute LLM embeddings for emotion labels and apply non-parametric clustering to group similar labels, across multiple datasets containing disjoint labels. We use these cluster centers to map music features (MERT) to the LLM embedding space. To further enhance the model, we introduce an alignment regularization that enables dissociation of MERT embeddings from different clusters. This further enhances the model's ability to better adaptation to unseen datasets. We demonstrate the effectiveness of our approach by performing zero-shot inference on a new dataset, showcasing its ability to generalize to unseen labels without additional training.
Computation and Language 183
☆ Context is Key(NMF): Modelling Topical Information Dynamics in Chinese Diaspora Media
Does the People's Republic of China (PRC) interfere with European elections through ethnic Chinese diaspora media? This question forms the basis of an ongoing research project exploring how PRC narratives about European elections are represented in Chinese diaspora media, and thus the objectives of PRC news media manipulation. In order to study diaspora media efficiently and at scale, it is necessary to use techniques derived from quantitative text analysis, such as topic modelling. In this paper, we present a pipeline for studying information dynamics in Chinese media. Firstly, we present KeyNMF, a new approach to static and dynamic topic modelling using transformer-based contextual embedding models. We provide benchmark evaluations to demonstrate that our approach is competitive on a number of Chinese datasets and metrics. Secondly, we integrate KeyNMF with existing methods for describing information dynamics in complex systems. We apply this pipeline to data from five news sites, focusing on the period of time leading up to the 2024 European parliamentary elections. Our methods and results demonstrate the effectiveness of KeyNMF for studying information dynamics in Chinese media and lay groundwork for further work addressing the broader research questions.
comment: Accepted to the 2024 Computational Humanities Research Conference (CHR)
☆ Meta-Chunking: Learning Efficient Text Segmentation via Logical Perception
Retrieval-Augmented Generation (RAG), while serving as a viable complement to large language models (LLMs), often overlooks the crucial aspect of text chunking within its pipeline, which impacts the quality of knowledge-intensive tasks. This paper introduces the concept of Meta-Chunking, which refers to a granularity between sentences and paragraphs, consisting of a collection of sentences within a paragraph that have deep linguistic logical connections. To implement Meta-Chunking, we designed two strategies based on LLMs: Margin Sampling Chunking and Perplexity Chunking. The former employs LLMs to perform binary classification on whether consecutive sentences need to be segmented, making decisions based on the probability difference obtained from margin sampling. The latter precisely identifies text chunk boundaries by analyzing the characteristics of perplexity distribution. Additionally, considering the inherent complexity of different texts, we propose a strategy that combines Meta-Chunking with dynamic merging to achieve a balance between fine-grained and coarse-grained text chunking. Experiments conducted on eleven datasets demonstrate that Meta-Chunking can more efficiently improve the performance of single-hop and multi-hop question answering based on RAG. For instance, on the 2WikiMultihopQA dataset, it outperforms similarity chunking by 1.32 while only consuming 45.8% of the time. Our code is available at https://github.com/IAAR-Shanghai/Meta-Chunking.
☆ JudgeBench: A Benchmark for Evaluating LLM-based Judges
LLM-based judges have emerged as a scalable alternative to human evaluation and are increasingly used to assess, compare, and improve models. However, the reliability of LLM-based judges themselves is rarely scrutinized. As LLMs become more advanced, their responses grow more sophisticated, requiring stronger judges to evaluate them. Existing benchmarks primarily focus on a judge's alignment with human preferences, but often fail to account for more challenging tasks where crowdsourced human preference is a poor indicator of factual and logical correctness. To address this, we propose a novel evaluation framework to objectively evaluate LLM-based judges. Based on this framework, we propose JudgeBench, a benchmark for evaluating LLM-based judges on challenging response pairs spanning knowledge, reasoning, math, and coding. JudgeBench leverages a novel pipeline for converting existing difficult datasets into challenging response pairs with preference labels reflecting objective correctness. Our comprehensive evaluation on a collection of prompted judges, fine-tuned judges, multi-agent judges, and reward models shows that JudgeBench poses a significantly greater challenge than previous benchmarks, with many strong models (e.g., GPT-4o) performing just slightly better than random guessing. Overall, JudgeBench offers a reliable platform for assessing increasingly advanced LLM-based judges. Data and code are available at https://github.com/ScalerLab/JudgeBench .
comment: preprint
☆ In-Context Learning Enables Robot Action Prediction in LLMs
Recently, Large Language Models (LLMs) have achieved remarkable success using in-context learning (ICL) in the language domain. However, leveraging the ICL capabilities within LLMs to directly predict robot actions remains largely unexplored. In this paper, we introduce RoboPrompt, a framework that enables off-the-shelf text-only LLMs to directly predict robot actions through ICL without training. Our approach first heuristically identifies keyframes that capture important moments from an episode. Next, we extract end-effector actions from these keyframes as well as the estimated initial object poses, and both are converted into textual descriptions. Finally, we construct a structured template to form ICL demonstrations from these textual descriptions and a task instruction. This enables an LLM to directly predict robot actions at test time. Through extensive experiments and analysis, RoboPrompt shows stronger performance over zero-shot and ICL baselines in simulated and real-world settings.
☆ Meta-Unlearning on Diffusion Models: Preventing Relearning Unlearned Concepts
With the rapid progress of diffusion-based content generation, significant efforts are being made to unlearn harmful or copyrighted concepts from pretrained diffusion models (DMs) to prevent potential model misuse. However, it is observed that even when DMs are properly unlearned before release, malicious finetuning can compromise this process, causing DMs to relearn the unlearned concepts. This occurs partly because certain benign concepts (e.g., "skin") retained in DMs are related to the unlearned ones (e.g., "nudity"), facilitating their relearning via finetuning. To address this, we propose meta-unlearning on DMs. Intuitively, a meta-unlearned DM should behave like an unlearned DM when used as is; moreover, if the meta-unlearned DM undergoes malicious finetuning on unlearned concepts, the related benign concepts retained within it will be triggered to self-destruct, hindering the relearning of unlearned concepts. Our meta-unlearning framework is compatible with most existing unlearning methods, requiring only the addition of an easy-to-implement meta objective. We validate our approach through empirical experiments on meta-unlearning concepts from Stable Diffusion models (SD-v1-4 and SDXL), supported by extensive ablation studies. Our code is available at https://github.com/sail-sg/Meta-Unlearning.
☆ Identifying Task Groupings for Multi-Task Learning Using Pointwise V-Usable Information
The success of multi-task learning can depend heavily on which tasks are grouped together. Naively grouping all tasks or a random set of tasks can result in negative transfer, with the multi-task models performing worse than single-task models. Though many efforts have been made to identify task groupings and to measure the relatedness among different tasks, it remains a challenging research topic to define a metric to identify the best task grouping out of a pool of many potential task combinations. We propose a metric of task relatedness based on task difficulty measured by pointwise V-usable information (PVI). PVI is a recently proposed metric to estimate how much usable information a dataset contains given a model. We hypothesize that tasks with not statistically different PVI estimates are similar enough to benefit from the joint learning process. We conduct comprehensive experiments to evaluate the feasibility of this metric for task grouping on 15 NLP datasets in the general, biomedical, and clinical domains. We compare the results of the joint learners against single learners, existing baseline methods, and recent large language models, including Llama 2 and GPT-4. The results show that by grouping tasks with similar PVI estimates, the joint learners yielded competitive results with fewer total parameters, with consistent performance across domains.
comment: main paper 12 pages, Appendix 7 pages, 1 figure, 18 tables
☆ Unitary Multi-Margin BERT for Robust Natural Language Processing
Recent developments in adversarial attacks on deep learning leave many mission-critical natural language processing (NLP) systems at risk of exploitation. To address the lack of computationally efficient adversarial defense methods, this paper reports a novel, universal technique that drastically improves the robustness of Bidirectional Encoder Representations from Transformers (BERT) by combining the unitary weights with the multi-margin loss. We discover that the marriage of these two simple ideas amplifies the protection against malicious interference. Our model, the unitary multi-margin BERT (UniBERT), boosts post-attack classification accuracies significantly by 5.3% to 73.8% while maintaining competitive pre-attack accuracies. Furthermore, the pre-attack and post-attack accuracy tradeoff can be adjusted via a single scalar parameter to best fit the design requirements for the target applications.
☆ StyleDistance: Stronger Content-Independent Style Embeddings with Synthetic Parallel Examples
Style representations aim to embed texts with similar writing styles closely and texts with different styles far apart, regardless of content. However, the contrastive triplets often used for training these representations may vary in both style and content, leading to potential content leakage in the representations. We introduce StyleDistance, a novel approach to training stronger content-independent style embeddings. We use a large language model to create a synthetic dataset of near-exact paraphrases with controlled style variations, and produce positive and negative examples across 40 distinct style features for precise contrastive learning. We assess the quality of our synthetic data and embeddings through human and automatic evaluations. StyleDistance enhances the content-independence of style embeddings, which generalize to real-world benchmarks and outperform leading style representations in downstream applications. Our model can be found at https://huggingface.co/StyleDistance/styledistance .
☆ Comparative Analysis of Extrinsic Factors for NER in French
Named entity recognition (NER) is a crucial task that aims to identify structured information, which is often replete with complex, technical terms and a high degree of variability. Accurate and reliable NER can facilitate the extraction and analysis of important information. However, NER for other than English is challenging due to limited data availability, as the high expertise, time, and expenses are required to annotate its data. In this paper, by using the limited data, we explore various factors including model structure, corpus annotation scheme and data augmentation techniques to improve the performance of a NER model for French. Our experiments demonstrate that these approaches can significantly improve the model's F1 score from original CRF score of 62.41 to 79.39. Our findings suggest that considering different extrinsic factors and combining these techniques is a promising approach for improving NER performance where the size of data is limited.
☆ CREAM: Consistency Regularized Self-Rewarding Language Models
Recent self-rewarding large language models (LLM) have successfully applied LLM-as-a-Judge to iteratively improve the alignment performance without the need of human annotations for preference data. These methods commonly utilize the same LLM to act as both the policy model (which generates responses) and the reward model (which scores and ranks those responses). The ranked responses are then used as preference pairs to train the LLM via direct alignment technologies (e.g. DPO). However, it is noteworthy that throughout this process, there is no guarantee of accuracy in the rewarding and ranking, which is critical for ensuring accurate rewards and high-quality preference data. Empirical results from relatively small LLMs (e.g., 7B parameters) also indicate that improvements from self-rewarding may diminish after several iterations in certain situations, which we hypothesize is due to accumulated bias in the reward system. This bias can lead to unreliable preference data for training the LLM. To address this issue, we first formulate and analyze the generalized iterative preference fine-tuning framework for self-rewarding language model. We then introduce the regularization to this generalized framework to mitigate the overconfident preference labeling in the self-rewarding process. Based on this theoretical insight, we propose a Consistency Regularized sElf-rewarding lAnguage Model (CREAM) that leverages the rewarding consistency across different iterations to regularize the self-rewarding training, helping the model to learn from more reliable preference data. With this explicit regularization, our empirical results demonstrate the superiority of CREAM in improving both reward consistency and alignment performance. The code is publicly available at https://github.com/Raibows/CREAM.
☆ WorldMedQA-V: a multilingual, multimodal medical examination dataset for multimodal language models evaluation
Multimodal/vision language models (VLMs) are increasingly being deployed in healthcare settings worldwide, necessitating robust benchmarks to ensure their safety, efficacy, and fairness. Multiple-choice question and answer (QA) datasets derived from national medical examinations have long served as valuable evaluation tools, but existing datasets are largely text-only and available in a limited subset of languages and countries. To address these challenges, we present WorldMedQA-V, an updated multilingual, multimodal benchmarking dataset designed to evaluate VLMs in healthcare. WorldMedQA-V includes 568 labeled multiple-choice QAs paired with 568 medical images from four countries (Brazil, Israel, Japan, and Spain), covering original languages and validated English translations by native clinicians, respectively. Baseline performance for common open- and closed-source models are provided in the local language and English translations, and with and without images provided to the model. The WorldMedQA-V benchmark aims to better match AI systems to the diverse healthcare environments in which they are deployed, fostering more equitable, effective, and representative applications.
comment: submitted for review, total of 14 pages
☆ WorldCuisines: A Massive-Scale Benchmark for Multilingual and Multicultural Visual Question Answering on Global Cuisines
Vision Language Models (VLMs) often struggle with culture-specific knowledge, particularly in languages other than English and in underrepresented cultural contexts. To evaluate their understanding of such knowledge, we introduce WorldCuisines, a massive-scale benchmark for multilingual and multicultural, visually grounded language understanding. This benchmark includes a visual question answering (VQA) dataset with text-image pairs across 30 languages and dialects, spanning 9 language families and featuring over 1 million data points, making it the largest multicultural VQA benchmark to date. It includes tasks for identifying dish names and their origins. We provide evaluation datasets in two sizes (12k and 60k instances) alongside a training dataset (1 million instances). Our findings show that while VLMs perform better with correct location context, they struggle with adversarial contexts and predicting specific regional cuisines and languages. To support future research, we release a knowledge base with annotated food entries and images along with the VQA data.
☆ Sarcasm Detection in a Less-Resourced Language
The sarcasm detection task in natural language processing tries to classify whether an utterance is sarcastic or not. It is related to sentiment analysis since it often inverts surface sentiment. Because sarcastic sentences are highly dependent on context, and they are often accompanied by various non-verbal cues, the task is challenging. Most of related work focuses on high-resourced languages like English. To build a sarcasm detection dataset for a less-resourced language, such as Slovenian, we leverage two modern techniques: a machine translation specific medium-size transformer model, and a very large generative language model. We explore the viability of translated datasets and how the size of a pretrained transformer affects its ability to detect sarcasm. We train ensembles of detection models and evaluate models' performance. The results show that larger models generally outperform smaller ones and that ensembling can slightly improve sarcasm detection performance. Our best ensemble approach achieves an $\text{F}_1$-score of 0.765 which is close to annotators' agreement in the source language.
comment: 4 pages, published in the Slovenian Conference on Artificial Intelligence
☆ VividMed: Vision Language Model with Versatile Visual Grounding for Medicine
Recent advancements in Vision Language Models (VLMs) have demonstrated remarkable promise in generating visually grounded responses. However, their application in the medical domain is hindered by unique challenges. For instance, most VLMs rely on a single method of visual grounding, whereas complex medical tasks demand more versatile approaches. Additionally, while most VLMs process only 2D images, a large portion of medical images are 3D. The lack of medical data further compounds these obstacles. To address these challenges, we present VividMed, a vision language model with versatile visual grounding for medicine. Our model supports generating both semantic segmentation masks and instance-level bounding boxes, and accommodates various imaging modalities, including both 2D and 3D data. We design a three-stage training procedure and an automatic data synthesis pipeline based on open datasets and models. Besides visual grounding tasks, VividMed also excels in other common downstream tasks, including Visual Question Answering (VQA) and report generation. Ablation studies empirically show that the integration of visual grounding ability leads to improved performance on these tasks. Our code is publicly available at https://github.com/function2-llx/MMMM.
☆ Building Better: Avoiding Pitfalls in Developing Language Resources when Data is Scarce
Language is a symbolic capital that affects people's lives in many ways (Bourdieu, 1977, 1991). It is a powerful tool that accounts for identities, cultures, traditions, and societies in general. Hence, data in a given language should be viewed as more than a collection of tokens. Good data collection and labeling practices are key to building more human-centered and socially aware technologies. While there has been a rising interest in mid- to low-resource languages within the NLP community, work in this space has to overcome unique challenges such as data scarcity and access to suitable annotators. In this paper, we collect feedback from those directly involved in and impacted by NLP artefacts for mid- to low-resource languages. We conduct a quantitative and qualitative analysis of the responses and highlight the main issues related to (1) data quality such as linguistic and cultural data suitability; and (2) the ethics of common annotation practices such as the misuse of online community services. Based on these findings, we make several recommendations for the creation of high-quality language artefacts that reflect the cultural milieu of its speakers, while simultaneously respecting the dignity and labor of data workers.
☆ Cross-Modal Safety Mechanism Transfer in Large Vision-Language Models
Vision-language alignment in Large Vision-Language Models (LVLMs) successfully enables LLMs to understand visual input. However, we find that existing vision-language alignment methods fail to transfer the existing safety mechanism for text in LLMs to vision, which leads to vulnerabilities in toxic image. To explore the cause of this problem, we give the insightful explanation of where and how the safety mechanism of LVLMs operates and conduct comparative analysis between text and vision. We find that the hidden states at the specific transformer layers play a crucial role in the successful activation of safety mechanism, while the vision-language alignment at hidden states level in current methods is insufficient. This results in a semantic shift for input images compared to text in hidden states, therefore misleads the safety mechanism. To address this, we propose a novel Text-Guided vision-language Alignment method (TGA) for LVLMs. TGA retrieves the texts related to input vision and uses them to guide the projection of vision into the hidden states space in LLMs. Experiments show that TGA not only successfully transfers the safety mechanism for text in basic LLMs to vision in vision-language alignment for LVLMs without any safety fine-tuning on the visual modality but also maintains the general performance on various vision tasks (Safe and Good).
☆ Evaluating Morphological Compositional Generalization in Large Language Models
Large language models (LLMs) have demonstrated significant progress in various natural language generation and understanding tasks. However, their linguistic generalization capabilities remain questionable, raising doubts about whether these models learn language similarly to humans. While humans exhibit compositional generalization and linguistic creativity in language use, the extent to which LLMs replicate these abilities, particularly in morphology, is under-explored. In this work, we systematically investigate the morphological generalization abilities of LLMs through the lens of compositionality. We define morphemes as compositional primitives and design a novel suite of generative and discriminative tasks to assess morphological productivity and systematicity. Focusing on agglutinative languages such as Turkish and Finnish, we evaluate several state-of-the-art instruction-finetuned multilingual models, including GPT-4 and Gemini. Our analysis shows that LLMs struggle with morphological compositional generalization particularly when applied to novel word roots, with performance declining sharply as morphological complexity increases. While models can identify individual morphological combinations better than chance, their performance lacks systematicity, leading to significant accuracy gaps compared to humans.
comment: 33 pages
☆ From Measurement Instruments to Training Data: Leveraging Theory-Driven Synthetic Training Data for Measuring Social Constructs
Computational text classification is a challenging task, especially for multi-dimensional social constructs. Recently, there has been increasing discussion that synthetic training data could enhance classification by offering examples of how these constructs are represented in texts. In this paper, we systematically examine the potential of theory-driven synthetic training data for improving the measurement of social constructs. In particular, we explore how researchers can transfer established knowledge from measurement instruments in the social sciences, such as survey scales or annotation codebooks, into theory-driven generation of synthetic data. Using two studies on measuring sexism and political topics, we assess the added value of synthetic training data for fine-tuning text classification models. Although the results of the sexism study were less promising, our findings demonstrate that synthetic data can be highly effective in reducing the need for labeled data in political topic classification. With only a minimal drop in performance, synthetic data allows for substituting large amounts of labeled data. Furthermore, theory-driven synthetic data performed markedly better than data generated without conceptual information in mind.
☆ Weak-to-Strong Generalization beyond Accuracy: a Pilot Study in Safety, Toxicity, and Legal Reasoning
As large language models (LLMs) continue to advance, ensuring their alignment with human values becomes increasingly critical. Traditional alignment methods heavily rely on human feedback to fine-tune models. With the emergence of superhuman models whose outputs may surpass human understanding, evaluating and aligning these models using human judgments poses significant challenges. To address the challenges, recent works use weak supervisors to elicit knowledge from much stronger models. However, there are important disanalogies between the empirical setup in the existing works and the genuine goal of alignment. We remark that existing works investigate the phenomenon of weak-to-strong generation in analogous setup (i.e., binary classification), rather than practical alignment-relevant tasks (e.g., safety). In this paper, we bridge this gap by extending weak-to-strong generation to the context of practical alignment. We empirically demonstrate the widespread phenomenon of weak-to-strong generation in three complicated alignment tasks: safety, toxicity, and legal reasoning}. Furthermore, we explore efficient strategies for improving alignment performance to enhance the quality of model outcomes. Lastly, we summarize and analyze the challenges and potential solutions in regard to specific alignment tasks, which we hope to catalyze the research progress on the topic of weak-to-strong generalization. Our code is released at https://github.com/yeruimeng/WTS.git.
☆ Parsing Akkadian Verbs with Prolog ACL-02
This paper describes a parsing/generation system for finite verbal forms in Akkadian, with the possible addition of suffixes, implemented in Prolog. The work described provides the framework and engine to interpret the D, N, and G stems along with accusative, dative and ventive endings.
comment: 6 pages, 9 figures, presented at ACL-02 the Association of Computational Linguistics, 2002
☆ Exploring Model Kinship for Merging Large Language Models
Model merging has become one of the key technologies for enhancing the capabilities and efficiency of Large Language Models (LLMs). However, our understanding of the expected performance gains and principles when merging any two models remains limited. In this work, we introduce model kinship, the degree of similarity or relatedness between LLMs, analogous to biological evolution. With comprehensive empirical analysis, we find that there is a certain relationship between model kinship and the performance gains after model merging, which can help guide our selection of candidate models. Inspired by this, we propose a new model merging strategy: Top-k Greedy Merging with Model Kinship, which can yield better performance on benchmark datasets. Specifically, we discover that using model kinship as a criterion can assist us in continuously performing model merging, alleviating the degradation (local optima) in model evolution, whereas model kinship can serve as a guide to escape these traps. Code is available at https://github.com/zjunlp/ModelKinship.
comment: Ongoing work
☆ Not All Votes Count! Programs as Verifiers Improve Self-Consistency of Language Models for Math Reasoning
Large language models (LLMs) have shown increasing proficiency in solving mathematical reasoning problems. However, many current open-source LLMs often still make calculation and semantic understanding errors in their intermediate reasoning steps. In this work, we propose PROVE, a simple yet effective framework that uses program-based verification as a heuristic to filter out potentially incorrect reasoning paths before aggregating the final answers. Instead of relying on vanilla majority voting, our approach rejects solutions whose corresponding program outputs are inconsistent with the generated solution, aggregating only those validated by Python programs. We conducted extensive experiments on 13 open-source LLMs from various model families and sizes, ranging from 0.5B to 13B parameters, across seven math benchmarks. We demonstrate that PROVE consistently outperforms vanilla majority voting as a heuristic for solving mathematical reasoning tasks across all datasets and model sizes. Notably, PROVE increases accuracy on the GSM8K benchmark from 48.85% to 53.83% for Qwen2-0.5B-Instruct, from 65.66% to 73.01% for Llama-3.2-1B-Instruct, from 73.39% to 79.61% for Gemma-2-2b-it, and from 41.32% to 59.51% for Llama-2-7B-chat. Our codes are available at https://github.com/declare-lab/prove.
☆ CCSBench: Evaluating Compositional Controllability in LLMs for Scientific Document Summarization
To broaden the dissemination of scientific knowledge to diverse audiences, scientific document summarization must simultaneously control multiple attributes such as length and empirical focus. However, existing research typically focuses on controlling single attributes, leaving the compositional control of multiple attributes underexplored. To address this gap, we introduce CCSBench, a benchmark for compositional controllable summarization in the scientific domain. Our benchmark enables fine-grained control over both explicit attributes (e.g., length), which are objective and straightforward, and implicit attributes (e.g., empirical focus), which are more subjective and conceptual. We conduct extensive experiments on GPT-4, LLaMA2, and other popular LLMs under various settings. Our findings reveal significant limitations in large language models' ability to balance trade-offs between control attributes, especially implicit ones that require deeper understanding and abstract reasoning.
☆ On the Risk of Evidence Pollution for Malicious Social Text Detection in the Era of LLMs
Evidence-enhanced detectors present remarkable abilities in identifying malicious social text with related evidence. However, the rise of large language models (LLMs) brings potential risks of evidence pollution to confuse detectors. This paper explores how to manipulate evidence, simulating potential misuse scenarios including basic pollution, and rephrasing or generating evidence by LLMs. To mitigate its negative impact, we propose three defense strategies from both the data and model sides, including machine-generated text detection, a mixture of experts, and parameter updating. Extensive experiments on four malicious social text detection tasks with ten datasets present that evidence pollution, especially the generate strategy, significantly compromises existing detectors. On the other hand, the defense strategies could mitigate evidence pollution, but they faced limitations for practical employment, such as the need for annotated data and huge inference costs. Further analysis illustrates that polluted evidence is of high quality, would compromise the model calibration, and could ensemble to amplify the negative impact.
☆ Can We Reverse In-Context Knowledge Edits?
In-context knowledge editing (IKE) enables efficient modification of large language model (LLM) outputs without parameter changes and at zero-cost. However, it can be misused to manipulate responses opaquely, e.g., insert misinformation or offensive content. Such malicious interventions could be incorporated into high-level wrapped APIs where the final input prompt is not shown to end-users. To address this issue, we investigate the detection and reversal of IKE-edits. First, we demonstrate that IKE-edits can be detected with high accuracy (F1 > 80\%) using only the top-10 output probabilities of the next token, even in a black-box setting, e.g. proprietary LLMs with limited output information. Further, we introduce the novel task of reversing IKE-edits using specially tuned reversal tokens. We explore using both continuous and discrete reversal tokens, achieving over 80\% accuracy in recovering original, unedited outputs across multiple LLMs. Our continuous reversal tokens prove particularly effective, with minimal impact on unedited prompts. Through analysis of output distributions, attention patterns, and token rankings, we provide insights into IKE's effects on LLMs and how reversal tokens mitigate them. This work represents a significant step towards enhancing LLM resilience against potential misuse of in-context editing, improving their transparency and trustworthiness.
☆ STRUX: An LLM for Decision-Making with Structured Explanations NAACL 2025
Countless decisions shape our daily lives, and it is paramount to understand the how and why behind these choices. In this paper, we introduce a new LLM decision-making framework called STRUX, which enhances LLM decision-making by providing structured explanations. These include favorable and adverse facts related to the decision, along with their respective strengths. STRUX begins by distilling lengthy information into a concise table of key facts. It then employs a series of self-reflection steps to determine which of these facts are pivotal, categorizing them as either favorable or adverse in relation to a specific decision. Lastly, we fine-tune an LLM to identify and prioritize these key facts to optimize decision-making. STRUX has been evaluated on the challenging task of forecasting stock investment decisions based on earnings call transcripts and demonstrated superior performance against strong baselines. It enhances decision transparency by allowing users to understand the impact of different factors, representing a meaningful step towards practical decision-making with LLMs.
comment: 10 pages, 7 figures, submitted to NAACL 2025
☆ A Claim Decomposition Benchmark for Long-form Answer Verification
The advancement of LLMs has significantly boosted the performance of complex long-form question answering tasks. However, one prominent issue of LLMs is the generated "hallucination" responses that are not factual. Consequently, attribution for each claim in responses becomes a common solution to improve the factuality and verifiability. Existing researches mainly focus on how to provide accurate citations for the response, which largely overlook the importance of identifying the claims or statements for each response. To bridge this gap, we introduce a new claim decomposition benchmark, which requires building system that can identify atomic and checkworthy claims for LLM responses. Specifically, we present the Chinese Atomic Claim Decomposition Dataset (CACDD), which builds on the WebCPM dataset with additional expert annotations to ensure high data quality. The CACDD encompasses a collection of 500 human-annotated question-answer pairs, including a total of 4956 atomic claims. We further propose a new pipeline for human annotation and describe the challenges of this task. In addition, we provide experiment results on zero-shot, few-shot and fine-tuned LLMs as baselines. The results show that the claim decomposition is highly challenging and requires further explorations. All code and data are publicly available at \url{https://github.com/FBzzh/CACDD}.
comment: Accepted by CCIR 2024
☆ LLM-based Translation Inference with Iterative Bilingual Understanding
The remarkable understanding and generation capabilities of large language models (LLMs) have greatly improved translation performance. However, incorrect understanding of the sentence to be translated can degrade translation quality. To address this issue, we proposed a novel Iterative Bilingual Understanding Translation (IBUT) method based on the cross-lingual capabilities of LLMs and the dual characteristics of translation tasks. The cross-lingual capability of LLMs enables the generation of contextual understanding for both the source and target languages separately. Furthermore, the dual characteristics allow IBUT to generate effective cross-lingual feedback, iteratively refining contextual understanding, thereby reducing errors and improving translation performance. Experimental results showed that the proposed IBUT outperforms several strong comparison methods, especially being generalized to multiple domains (e.g., news, commonsense, and cultural translation benchmarks).
comment: work in process
☆ MedAide: Towards an Omni Medical Aide via Specialized LLM-based Multi-Agent Collaboration
Large Language Model (LLM)-driven interactive systems currently show potential promise in healthcare domains. Despite their remarkable capabilities, LLMs typically lack personalized recommendations and diagnosis analysis in sophisticated medical applications, causing hallucinations and performance bottlenecks. To address these challenges, this paper proposes MedAide, an LLM-based omni medical multi-agent collaboration framework for specialized healthcare services. Specifically, MedAide first performs query rewriting through retrieval-augmented generation to accomplish accurate medical intent understanding. Immediately, we devise a contextual encoder to obtain intent prototype embeddings, which are used to recognize fine-grained intents by similarity matching. According to the intent relevance, the activated agents collaborate effectively to provide integrated decision analysis. Extensive experiments are conducted on four medical benchmarks with composite intents. Experimental results from automated metrics and expert doctor evaluations show that MedAide outperforms current LLMs and improves their medical proficiency and strategic reasoning.
comment: Under review
☆ FiRST: Finetuning Router-Selective Transformers for Input-Adaptive Latency Reduction ICLR 2025
Auto-regressive Large Language Models (LLMs) demonstrate remarkable performance across domanins such as vision and language processing. However, due to sequential processing through a stack of transformer layers, autoregressive decoding faces significant computation/latency challenges, particularly in resource constrained environments like mobile and edge devices. Existing approaches in literature that aim to improve latency via skipping layers have two distinct flavors - 1) Early exit 2) Input-agnostic heuristics where tokens exit at pre-determined layers irrespective of input sequence. Both the above strategies have limitations - the former cannot be applied to handle KV Caching necessary for speed-ups in modern framework and the latter does not capture the variation in layer importance across tasks or more generally, across input sequences. To address both limitations, we propose FIRST, an algorithm that reduces inference latency by using layer-specific routers to select a subset of transformer layers adaptively for each input sequence - the prompt (during prefill stage) decides which layers will be skipped during decoding. FIRST preserves compatibility with KV caching enabling faster inference while being quality-aware. FIRST is model-agnostic and can be easily enabled on any pre-trained LLM. We further improve performance by incorporating LoRA adapters for fine-tuning on external datasets, enhancing task-specific accuracy while maintaining latency benefits. Our approach reveals that input adaptivity is critical - indeed, different task-specific middle layers play a crucial role in evolving hidden representations depending on task. Extensive experiments show that FIRST significantly reduces latency while retaining competitive performance (as compared to baselines), making our approach an efficient solution for LLM deployment in low-resource environments.
comment: 17 pages, 6 figures, Submitted to ICLR 2025
☆ Advancing Fairness in Natural Language Processing: From Traditional Methods to Explainability
The burgeoning field of Natural Language Processing (NLP) stands at a critical juncture where the integration of fairness within its frameworks has become an imperative. This PhD thesis addresses the need for equity and transparency in NLP systems, recognizing that fairness in NLP is not merely a technical challenge but a moral and ethical necessity, requiring a rigorous examination of how these technologies interact with and impact diverse human populations. Through this lens, this thesis undertakes a thorough investigation into the development of equitable NLP methodologies and the evaluation of biases that prevail in current systems. First, it introduces an innovative algorithm to mitigate biases in multi-class classifiers, tailored for high-risk NLP applications, surpassing traditional methods in both bias mitigation and prediction accuracy. Then, an analysis of the Bios dataset reveals the impact of dataset size on discriminatory biases and the limitations of standard fairness metrics. This awareness has led to explorations in the field of explainable AI, aiming for a more complete understanding of biases where traditional metrics are limited. Consequently, the thesis presents COCKATIEL, a model-agnostic explainability method that identifies and ranks concepts in Transformer models, outperforming previous approaches in sentiment analysis tasks. Finally, the thesis contributes to bridging the gap between fairness and explainability by introducing TaCo, a novel method to neutralize bias in Transformer model embeddings. In conclusion, this thesis constitutes a significant interdisciplinary endeavor that intertwines explicability and fairness to challenge and reshape current NLP paradigms. The methodologies and critiques presented contribute to the ongoing discourse on fairness in machine learning, offering actionable solutions for more equitable and responsible AI systems.
comment: PhD Thesis, Toulouse University
☆ With a Grain of SALT: Are LLMs Fair Across Social Dimensions?
This paper presents an analysis of biases in open-source Large Language Models (LLMs) across various genders, religions, and races. We introduce a methodology for generating a bias detection dataset using seven bias triggers: General Debate, Positioned Debate, Career Advice, Story Generation, Problem-Solving, Cover-Letter Writing, and CV Generation. We use GPT-4o to generate a diverse set of prompts for each trigger across various genders, religious and racial groups. We evaluate models from Llama and Gemma family on the generated dataset. We anonymise the LLM-generated text associated with each group using GPT-4o-mini and do a pairwise comparison using GPT-4o-as-a-Judge. To quantify bias in the LLM-generated text we use the number of wins and losses in the pairwise comparison. Our analysis spans three languages, English, German, and Arabic to explore how language influences bias manifestation. Our findings reveal that LLMs exhibit strong polarization toward certain groups across each category, with a notable consistency observed across models. However, when switching languages, variations and anomalies emerge, often attributable to cultural cues and contextual differences.
☆ End-to-end Planner Training for Language Modeling
Through end-to-end training to predict the next token, LLMs have become valuable tools for various tasks. Enhancing their core training in language modeling can improve numerous downstream applications. A successful approach to enhance language modeling uses a separate planning module to predict abstract labels of future sentences and conditions the LM on these predictions. However, this method is non-differentiable, preventing joint end-to-end tuning of the planner with the LM. We propose an effective method to improve this approach by enabling joint fine-tuning of the planner and the LM. We show that a naive way of approximating the gradient of selecting a label via the straight-through estimator is not effective. Instead, we propose to use the predicted label probabilities as mixing weights to condition the LM on a weighted average of label embeddings in a differentiable manner. This not only enables joint fine-tuning of the planner and the LM, but also allows the LM to draw on the full label distribution predicted by the planner, retaining more information. Our experimental results show consistent improvements in perplexity.
comment: 14 pages
☆ Insights from the Inverse: Reconstructing LLM Training Goals Through Inverse RL
Large language models (LLMs) trained with Reinforcement Learning from Human Feedback (RLHF) have demonstrated remarkable capabilities, but their underlying reward functions and decision-making processes remain opaque. This paper introduces a novel approach to interpreting LLMs by applying inverse reinforcement learning (IRL) to recover their implicit reward functions. We conduct experiments on toxicity-aligned LLMs of varying sizes, extracting reward models that achieve up to 80.40% accuracy in predicting human preferences. Our analysis reveals key insights into the non-identifiability of reward functions, the relationship between model size and interpretability, and potential pitfalls in the RLHF process. We demonstrate that IRL-derived reward models can be used to fine-tune new LLMs, resulting in comparable or improved performance on toxicity benchmarks. This work provides a new lens for understanding and improving LLM alignment, with implications for the responsible development and deployment of these powerful systems.
comment: Preprint
☆ KcMF: A Knowledge-compliant Framework for Schema and Entity Matching with Fine-tuning-free LLMs
Schema and entity matching tasks are crucial for data integration and management. While large language models (LLMs) have shown promising results in these tasks, they suffer from hallucinations and confusion about task instructions. In this paper, we present the Knowledge-Compliant Matching Framework (KcMF), an LLM-based approach that addresses these issues without the need for domain-specific fine-tuning. KcMF employs a pseudo-code-based task decomposition strategy to adopt task-specific natural language statements that guide LLM reasoning and reduce confusion. We also propose two mechanisms, Dataset as Knowledge (DaK) and Example as Knowledge (EaK), to build domain knowledge sets when unstructured domain knowledge is lacking. Additionally, we introduce a result-ensembling strategy to leverage multiple knowledge sources and suppress poorly formatted outputs. Comprehensive evaluations on schema and entity matching tasks demonstrate that KcMF outperforms previous non-LLM state-of-the-art (SOTA) methods by an average F1 score of 22.9% and competes effectively with SOTA fine-tuned LLMs. Moreover, KcMF generalizes well across different LLMs.
☆ MlingConf: A Comprehensive Study of Multilingual Confidence Estimation on Large Language Models
The tendency of Large Language Models (LLMs) to generate hallucinations raises concerns regarding their reliability. Therefore, confidence estimations indicating the extent of trustworthiness of the generations become essential. However, current LLM confidence estimations in languages other than English remain underexplored. This paper addresses this gap by introducing a comprehensive investigation of Multilingual Confidence estimation (MlingConf) on LLMs, focusing on both language-agnostic (LA) and language-specific (LS) tasks to explore the performance and language dominance effects of multilingual confidence estimations on different tasks. The benchmark comprises four meticulously checked and human-evaluate high-quality multilingual datasets for LA tasks and one for the LS task tailored to specific social, cultural, and geographical contexts of a language. Our experiments reveal that on LA tasks English exhibits notable linguistic dominance in confidence estimations than other languages, while on LS tasks, using question-related language to prompt LLMs demonstrates better linguistic dominance in multilingual confidence estimations. The phenomena inspire a simple yet effective native-tone prompting strategy by employing language-specific prompts for LS tasks, effectively improving LLMs' reliability and accuracy on LS tasks.
☆ Retrieval-Reasoning Large Language Model-based Synthetic Clinical Trial Generation
Machine learning (ML) exhibits promise in the clinical domain. However, it is constrained by data scarcity and ethical considerations, as the generation of clinical trials presents significant challenges due to stringent privacy regulations, high costs, and the extended duration required for conducting studies with human participants. Despite the advancements of large language models (LLMs) in general generation tasks, their potential in facilitating the generation of synthetic clinical trials is under-explored. To address this gap, we introduce a novel Retrieval-Reasoning few-shot framework that leverages LLMs to generate artificial yet realistic and diverse clinical trials with binary success/failure labels. Experiments conducted on real clinical trials from the \url{ClinicalTrials.gov} database demonstrate that our synthetic data can effectively augment real datasets. Furthermore, by fine-tuning a pre-trained model as a binary classifier on synthetic clinical trial datasets, we demonstrate that this augmentation enhances model training for downstream tasks such as trial outcome prediction. Our findings suggest that LLMs for synthetic clinical trial generation hold promise for accelerating clinical research and upholding ethical standards for patient privacy. The code is publicly available at https://anonymous.4open.science/r/Retrieval_Reasoning_Clinical_Trial_Generation-3EC4.
☆ Learning to Predict Usage Options of Product Reviews with LLM-Generated Labels
Annotating large datasets can be challenging. However, crowd-sourcing is often expensive and can lack quality, especially for non-trivial tasks. We propose a method of using LLMs as few-shot learners for annotating data in a complex natural language task where we learn a standalone model to predict usage options for products from customer reviews. We also propose a new evaluation metric for this scenario, HAMS4, that can be used to compare a set of strings with multiple reference sets. Learning a custom model offers individual control over energy efficiency and privacy measures compared to using the LLM directly for the sequence-to-sequence task. We compare this data annotation approach with other traditional methods and demonstrate how LLMs can enable considerable cost savings. We find that the quality of the resulting data exceeds the level attained by third-party vendor services and that GPT-4-generated labels even reach the level of domain experts. We make the code and generated labels publicly available.
comment: 9 pages
☆ Bridging the Language Gaps in Large Language Models with Inference-Time Cross-Lingual Intervention
Large Language Models (LLMs) have shown remarkable capabilities in natural language processing but exhibit significant performance gaps among different languages. Most existing approaches to address these disparities rely on pretraining or fine-tuning, which are resource-intensive. To overcome these limitations without incurring significant costs, we propose Inference-Time Cross-Lingual Intervention (INCLINE), a novel framework that enhances LLM performance on low-performing (source) languages by aligning their internal representations with those of high-performing (target) languages during inference. INCLINE initially learns alignment matrices using parallel sentences from source and target languages through a Least-Squares optimization, and then applies these matrices during inference to transform the low-performing language representations toward the high-performing language space. Extensive experiments on nine benchmarks with five LLMs demonstrate that INCLINE significantly improves performance across diverse tasks and languages, compared to recent strong baselines. Our analysis demonstrates that INCLINE is highly cost-effective and applicable to a wide range of applications. In addition, we release the code to foster research along this line: https://github.com/weixuan-wang123/INCLINE.
☆ The Best of Both Worlds: Bridging Quality and Diversity in Data Selection with Bipartite Graph
The performance of large language models (LLMs) in natural language processing (NLP) tasks is significantly influenced by the quality and diversity of data used for supervised fine-tuning (SFT). Current data selection methods often focus solely on quality or diversity, leading to underperforming models due to suboptimal training data. In this paper, we introduce GraphFilter, a novel method that represents the dataset as a bipartite graph, linking sentences to their constituent n-grams. This representation effectively captures the relationships between sentences and linguistic patterns, facilitating the selection of sentences that enhance n-gram diversity. To balance quality and diversity during selection, we propose a priority function that combines the quality metric with the diversity metric in a multiplicative manner. GraphFilter iteratively selects high-priority sentences, updates the bipartite graph by removing covered n-grams, and re-calculates priorities to reflect the evolving data landscape. We conduct extensive experiments using three model backbones across six widely used benchmarks. The results demonstrate that GraphFilter outperforms all nine baseline approaches, achieving superior model performance and computational efficiency. Our analyses validate the effectiveness of our design choices, examine the subsets selected by GraphFilter and other methods, highlight the importance of instruction diversity, and explore the role of quality and diversity in relation to subset sizes. GraphFilter establishes a new foundation for effective data selection strategies, encouraging further research in data selection for LLMs.
comment: 19 pages, 5 figures, 5 tables
☆ Open Ko-LLM Leaderboard2: Bridging Foundational and Practical Evaluation for Korean LLMs
The Open Ko-LLM Leaderboard has been instrumental in benchmarking Korean Large Language Models (LLMs), yet it has certain limitations. Notably, the disconnect between quantitative improvements on the overly academic leaderboard benchmarks and the qualitative impact of the models should be addressed. Furthermore, the benchmark suite is largely composed of translated versions of their English counterparts, which may not fully capture the intricacies of the Korean language. To address these issues, we propose Open Ko-LLM Leaderboard2, an improved version of the earlier Open Ko-LLM Leaderboard. The original benchmarks are entirely replaced with new tasks that are more closely aligned with real-world capabilities. Additionally, four new native Korean benchmarks are introduced to better reflect the distinct characteristics of the Korean language. Through these refinements, Open Ko-LLM Leaderboard2 seeks to provide a more meaningful evaluation for advancing Korean LLMs.
☆ Expanding Chatbot Knowledge in Customer Service: Context-Aware Similar Question Generation Using Large Language Models
Reliable responses of service chatbots are often achieved by employing retrieval-based methods that restrict answers to a knowledge base comprising predefined question-answer pairs (QA pairs). To accommodate potential variations in how a customer's query may be expressed, it emerges as the favored solution to augment these QA pairs with similar questions that are possibly diverse while remaining semantic consistency. This augmentation task is known as Similar Question Generation (SQG). Traditional methods that heavily rely on human efforts or rule-based techniques suffer from limited diversity or significant semantic deviation from the source question, only capable of producing a finite number of useful questions. To address these limitations, we propose an SQG approach based on Large Language Models (LLMs), capable of producing a substantial number of diverse questions while maintaining semantic consistency to the source QA pair. This is achieved by leveraging LLMs' natural language understanding capability through fine-tuning with specially designed prompts. The experiments conducted on a real customer-service dataset demonstrate that our method surpasses baseline methods by a significant margin in terms of semantic diversity. Human evaluation further confirms that integrating the answer that reflects the customer's intention is crucial for increasing the number of generated questions that meet business requirements.
☆ Conformity in Large Language Models
The conformity effect describes the tendency of individuals to align their responses with the majority. Studying this bias in large language models (LLMs) is crucial, as LLMs are increasingly used in various information-seeking and decision-making tasks as conversation partners to improve productivity. Thus, conformity to incorrect responses can compromise their effectiveness. In this paper, we adapt psychological experiments to examine the extent of conformity in state-of-the-art LLMs. Our findings reveal that all models tested exhibit varying levels of conformity toward the majority, regardless of their initial choice or correctness, across different knowledge domains. Notably, we are the first to show that LLMs are more likely to conform when they are more uncertain in their own prediction. We further explore factors that influence conformity, such as training paradigms and input characteristics, finding that instruction-tuned models are less susceptible to conformity, while increasing the naturalness of majority tones amplifies conformity. Finally, we propose two interventions--Devil's Advocate and Question Distillation--to mitigate conformity, providing insights into building more robust language models.
comment: 16 pages (8 pages main body), 14 figures
☆ Theoretical Analysis of Hierarchical Language Recognition and Generation by Transformers without Positional Encoding
In this study, we provide constructive proof that Transformers can recognize and generate hierarchical language efficiently with respect to model size, even without the need for a specific positional encoding. Specifically, we show that causal masking and a starting token enable Transformers to compute positional information and depth within hierarchical structures. We demonstrate that Transformers without positional encoding can generate hierarchical languages. Furthermore, we suggest that explicit positional encoding might have a detrimental effect on generalization with respect to sequence length.
comment: 55 pages, 11 figures
☆ Revealing the Barriers of Language Agents in Planning
Autonomous planning has been an ongoing pursuit since the inception of artificial intelligence. Based on curated problem solvers, early planning agents could deliver precise solutions for specific tasks but lacked generalization. The emergence of large language models (LLMs) and their powerful reasoning capabilities has reignited interest in autonomous planning by automatically generating reasonable solutions for given tasks. However, prior research and our experiments show that current language agents still lack human-level planning abilities. Even the state-of-the-art reasoning model, OpenAI o1, achieves only 15.6% on one of the complex real-world planning benchmarks. This highlights a critical question: What hinders language agents from achieving human-level planning? Although existing studies have highlighted weak performance in agent planning, the deeper underlying issues and the mechanisms and limitations of the strategies proposed to address them remain insufficiently understood. In this work, we apply the feature attribution study and identify two key factors that hinder agent planning: the limited role of constraints and the diminishing influence of questions. We also find that although current strategies help mitigate these challenges, they do not fully resolve them, indicating that agents still have a long way to go before reaching human-level intelligence.
comment: Work in Progress
☆ Beyond Coarse-Grained Matching in Video-Text Retrieval ACCV 2024
Video-text retrieval has seen significant advancements, yet the ability of models to discern subtle differences in captions still requires verification. In this paper, we introduce a new approach for fine-grained evaluation. Our approach can be applied to existing datasets by automatically generating hard negative test captions with subtle single-word variations across nouns, verbs, adjectives, adverbs, and prepositions. We perform comprehensive experiments using four state-of-the-art models across two standard benchmarks (MSR-VTT and VATEX) and two specially curated datasets enriched with detailed descriptions (VLN-UVO and VLN-OOPS), resulting in a number of novel insights: 1) our analyses show that the current evaluation benchmarks fall short in detecting a model's ability to perceive subtle single-word differences, 2) our fine-grained evaluation highlights the difficulty models face in distinguishing such subtle variations. To enhance fine-grained understanding, we propose a new baseline that can be easily combined with current methods. Experiments on our fine-grained evaluations demonstrate that this approach enhances a model's ability to understand fine-grained differences.
comment: Accepted to ACCV 2024
☆ Nominal Class Assignment in Swahili: A Computational Account
We discuss the open question of the relation between semantics and nominal class assignment in Swahili. We approach the problem from a computational perspective, aiming first to quantify the extent of this relation, and then to explicate its nature, taking extra care to suppress morphosyntactic confounds. Our results are the first of their kind, providing a quantitative evaluation of the semantic cohesion of each nominal class, as well as a nuanced taxonomic description of its semantic content.
comment: Tenth Italian Conference on Computational Linguistics (CliC-it-2024)
☆ ProSA: Assessing and Understanding the Prompt Sensitivity of LLMs EMNLP 2024
Large language models (LLMs) have demonstrated impressive capabilities across various tasks, but their performance is highly sensitive to the prompts utilized. This variability poses challenges for accurate assessment and user satisfaction. Current research frequently overlooks instance-level prompt variations and their implications on subjective evaluations. To address these shortcomings, we introduce ProSA, a framework designed to evaluate and comprehend prompt sensitivity in LLMs. ProSA incorporates a novel sensitivity metric, PromptSensiScore, and leverages decoding confidence to elucidate underlying mechanisms. Our extensive study, spanning multiple tasks, uncovers that prompt sensitivity fluctuates across datasets and models, with larger models exhibiting enhanced robustness. We observe that few-shot examples can alleviate this sensitivity issue, and subjective evaluations are also susceptible to prompt sensitivities, particularly in complex, reasoning-oriented tasks. Furthermore, our findings indicate that higher model confidence correlates with increased prompt robustness. We believe this work will serve as a helpful tool in studying prompt sensitivity of LLMs. The project is released at: https://github.com/open-compass/ProSA .
comment: EMNLP 2024, Findings
☆ Tracking Universal Features Through Fine-Tuning and Model Merging
We study how features emerge, disappear, and persist across models fine-tuned on different domains of text. More specifically, we start from a base one-layer Transformer language model that is trained on a combination of the BabyLM corpus, and a collection of Python code from The Stack. This base model is adapted to two new domains of text: TinyStories, and the Lua programming language, respectively; and then these two models are merged using these two models using spherical linear interpolation. Our exploration aims to provide deeper insights into the stability and transformation of features across typical transfer-learning scenarios using small-scale models and sparse auto-encoders.
Prompt Compression for Large Language Models: A Survey
Leveraging large language models (LLMs) for complex natural language tasks typically requires long-form prompts to convey detailed requirements and information, which results in increased memory usage and inference costs. To mitigate these challenges, multiple efficient methods have been proposed, with prompt compression gaining significant research interest. This survey provides an overview of prompt compression techniques, categorized into hard prompt methods and soft prompt methods. First, the technical approaches of these methods are compared, followed by an exploration of various ways to understand their mechanisms, including the perspectives of attention optimization, Parameter-Efficient Fine-Tuning (PEFT), modality fusion, and new synthetic language. We also examine the downstream adaptations of various prompt compression techniques. Finally, the limitations of current prompt compression methods are analyzed, and several future directions are outlined, such as optimizing the compression encoder, combining hard and soft prompts methods, and leveraging insights from multimodality.
☆ Evaluation of Attribution Bias in Retrieval-Augmented Large Language Models
Attributing answers to source documents is an approach used to enhance the verifiability of a model's output in retrieval augmented generation (RAG). Prior work has mainly focused on improving and evaluating the attribution quality of large language models (LLMs) in RAG, but this may come at the expense of inducing biases in the attribution of answers. We define and examine two aspects in the evaluation of LLMs in RAG pipelines, namely attribution sensitivity and bias with respect to authorship information. We explicitly inform an LLM about the authors of source documents, instruct it to attribute its answers, and analyze (i) how sensitive the LLM's output is to the author of source documents, and (ii) whether the LLM exhibits a bias towards human-written or AI-generated source documents. We design an experimental setup in which we use counterfactual evaluation to study three LLMs in terms of their attribution sensitivity and bias in RAG pipelines. Our results show that adding authorship information to source documents can significantly change the attribution quality of LLMs by 3% to 18%. Moreover, we show that LLMs can have an attribution bias towards explicit human authorship, which can serve as a competing hypothesis for findings of prior work that shows that LLM-generated content may be preferred over human-written contents. Our findings indicate that metadata of source documents can influence LLMs' trust, and how they attribute their answers. Furthermore, our research highlights attribution bias and sensitivity as a novel aspect of brittleness in LLMs.
☆ HerO at AVeriTeC: The Herd of Open Large Language Models for Verifying Real-World Claims EMNLP 2024
To tackle the AVeriTeC shared task hosted by the FEVER-24, we introduce a system that only employs publicly available large language models (LLMs) for each step of automated fact-checking, dubbed the Herd of Open LLMs for verifying real-world claims (HerO). HerO employs multiple LLMs for each step of automated fact-checking. For evidence retrieval, a language model is used to enhance a query by generating hypothetical fact-checking documents. We prompt pretrained and fine-tuned LLMs for question generation and veracity prediction by crafting prompts with retrieved in-context samples. HerO achieved 2nd place on the leaderboard with the AVeriTeC score of 0.57, suggesting the potential of open LLMs for verifying real-world claims. For future research, we make our code publicly available at https://github.com/ssu-humane/HerO.
comment: A system description paper for the AVeriTeC shared task, hosted by the seventh FEVER workshop (co-located with EMNLP 2024)
☆ PRefLexOR: Preference-based Recursive Language Modeling for Exploratory Optimization of Reasoning and Agentic Thinking
PRefLexOR (Preference-based Recursive Language Modeling for Exploratory Optimization of Reasoning) combines preference optimization with concepts from Reinforcement Learning to enable models to self-teach through iterative reasoning improvements. We propose a recursive learning approach that engages the model in multi-step reasoning, revisiting, and refining intermediate steps before producing a final output in training and inference phases. Through multiple training stages, the model first learns to align its reasoning with accurate decision paths by optimizing the log odds between preferred and non-preferred responses. During this process, PRefLexOR builds a dynamic knowledge graph by generating questions from random text chunks and retrieval-augmentation to contextualize relevant details from the entire training corpus. In the second stage, preference optimization enhances model performance by using rejection sampling to fine-tune reasoning quality by continually producing in-situ training data while masking the reasoning steps. Recursive optimization within a thinking token framework introduces iterative feedback loops, where the model refines reasoning, achieving deeper coherence, consistency, and adaptability. Implemented in small language models with only 3 billion parameters, we should that even tiny models can iteratively teach themselves to reason with greater depth and reflectivity. Our implementation is straightforward and can be incorporated into any existing pretrained LLM. We focus our examples on applications in biological materials science and demonstrate the method in a variety of case studies that range from in-domain to cross-domain applications. Using reasoning strategies that include thinking and reflection modalities we build a multi-agent recursive self-improving inference approach to successively improve responses via repeated sampling in inference time.
☆ Proactive Agent: Shifting LLM Agents from Reactive Responses to Active Assistance
Agents powered by large language models have shown remarkable abilities in solving complex tasks. However, most agent systems remain reactive, limiting their effectiveness in scenarios requiring foresight and autonomous decision-making. In this paper, we tackle the challenge of developing proactive agents capable of anticipating and initiating tasks without explicit human instructions. We propose a novel data-driven approach for this problem. Firstly, we collect real-world human activities to generate proactive task predictions. These predictions are then labeled by human annotators as either accepted or rejected. The labeled data is used to train a reward model that simulates human judgment and serves as an automatic evaluator of the proactiveness of LLM agents. Building on this, we develop a comprehensive data generation pipeline to create a diverse dataset, ProactiveBench, containing 6,790 events. Finally, we demonstrate that fine-tuning models with the proposed ProactiveBench can significantly elicit the proactiveness of LLM agents. Experimental results show that our fine-tuned model achieves an F1-Score of 66.47% in proactively offering assistance, outperforming all open-source and close-source models. These results highlight the potential of our method in creating more proactive and effective agent systems, paving the way for future advancements in human-agent collaboration.
comment: 9 pages, 4 figures
☆ GECTurk WEB: An Explainable Online Platform for Turkish Grammatical Error Detection and Correction
Sophisticated grammatical error detection/correction tools are available for a small set of languages such as English and Chinese. However, it is not straightforward -- if not impossible -- to adapt them to morphologically rich languages with complex writing rules like Turkish which has more than 80 million speakers. Even though several tools exist for Turkish, they primarily focus on spelling errors rather than grammatical errors and lack features such as web interfaces, error explanations and feedback mechanisms. To fill this gap, we introduce GECTurk WEB, a light, open-source, and flexible web-based system that can detect and correct the most common forms of Turkish writing errors, such as the misuse of diacritics, compound and foreign words, pronouns, light verbs along with spelling mistakes. Our system provides native speakers and second language learners an easily accessible tool to detect/correct such mistakes and also to learn from their mistakes by showing the explanation for the violated rule(s). The proposed system achieves 88,3 system usability score, and is shown to help learn/remember a grammatical rule (confirmed by 80% of the participants). The GECTurk WEB is available both as an offline tool at https://github.com/GGLAB-KU/gecturkweb or online at www.gecturk.net.
☆ A linguistic analysis of undesirable outcomes in the era of generative AI
Recent research has focused on the medium and long-term impacts of generative AI, posing scientific and societal challenges mainly due to the detection and reliability of machine-generated information, which is projected to form the major content on the Web soon. Prior studies show that LLMs exhibit a lower performance in generation tasks (model collapse) as they undergo a fine-tuning process across multiple generations on their own generated content (self-consuming loop). In this paper, we present a comprehensive simulation framework built upon the chat version of LLama2, focusing particularly on the linguistic aspects of the generated content, which has not been fully examined in existing studies. Our results show that the model produces less lexical rich content across generations, reducing diversity. The lexical richness has been measured using the linguistic measures of entropy and TTR as well as calculating the POSTags frequency. The generated content has also been examined with an $n$-gram analysis, which takes into account the word order, and semantic networks, which consider the relation between different words. These findings suggest that the model collapse occurs not only by decreasing the content diversity but also by distorting the underlying linguistic patterns of the generated text, which both highlight the critical importance of carefully choosing and curating the initial input text, which can alleviate the model collapse problem. Furthermore, we conduct a qualitative analysis of the fine-tuned models of the pipeline to compare their performances on generic NLP tasks to the original model. We find that autophagy transforms the initial model into a more creative, doubtful and confused one, which might provide inaccurate answers and include conspiracy theories in the model responses, spreading false and biased information on the Web.
☆ Understanding the Role of LLMs in Multimodal Evaluation Benchmarks
The rapid advancement of Multimodal Large Language Models (MLLMs) has been accompanied by the development of various benchmarks to evaluate their capabilities. However, the true nature of these evaluations and the extent to which they assess multimodal reasoning versus merely leveraging the underlying Large Language Model (LLM) backbone remain unclear. This paper presents a comprehensive investigation into the role of LLM backbones in MLLM evaluation, focusing on two critical aspects: the degree to which current benchmarks truly assess multimodal reasoning and the influence of LLM prior knowledge on performance. Specifically, we introduce a modified evaluation protocol to disentangle the contributions of the LLM backbone from multimodal integration, and an automatic knowledge identification technique for diagnosing whether LLMs equip the necessary knowledge for corresponding multimodal questions. Our study encompasses four diverse MLLM benchmarks and eight state-of-the-art MLLMs. Key findings reveal that some benchmarks allow high performance even without visual inputs and up to 50\% of error rates can be attributed to insufficient world knowledge in the LLM backbone, indicating a heavy reliance on language capabilities. To address knowledge deficiencies, we propose a knowledge augmentation pipeline that achieves significant performance gains, with improvements of up to 60\% on certain datasets, resulting in a approximately 4x increase in performance. Our work provides crucial insights into the role of the LLM backbone in MLLMs, and highlights the need for more nuanced benchmarking approaches.
☆ Neuron-based Personality Trait Induction in Large Language Models
Large language models (LLMs) have become increasingly proficient at simulating various personality traits, an important capability for supporting related applications (e.g., role-playing). To further improve this capacity, in this paper, we present a neuron-based approach for personality trait induction in LLMs, with three major technical contributions. First, we construct PersonalityBench, a large-scale dataset for identifying and evaluating personality traits in LLMs. This dataset is grounded in the Big Five personality traits from psychology and is designed to assess the generative capabilities of LLMs towards specific personality traits. Second, by leveraging PersonalityBench, we propose an efficient method for identifying personality-related neurons within LLMs by examining the opposite aspects of a given trait. Third, we develop a simple yet effective induction method that manipulates the values of these identified personality-related neurons. This method enables fine-grained control over the traits exhibited by LLMs without training and modifying model parameters. Extensive experiments validate the efficacy of our neuron identification and trait induction methods. Notably, our approach achieves comparable performance as fine-tuned models, offering a more efficient and flexible solution for personality trait induction in LLMs. We provide access to all the mentioned resources at https://github.com/RUCAIBox/NPTI.
☆ Optimizing Low-Resource Language Model Training: Comprehensive Analysis of Multi-Epoch, Multi-Lingual, and Two-Stage Approaches
In this paper, we address the challenge of optimizing training setups for Large Language Models (LLMs) of low-resource language with a limited amount of corpus. Existing works adopt multi-epoch, multi-lingual, and two-stage training to utilize the limited target language corpus efficiently. However, there is still a lack of understanding about the optimal hyperparameter setups for combining these three approaches to train LLMs. We exhaustively explore training setups for low-resource language LLM, combining these three approaches, and found the following insights for efficiently reducing the cost of hyperparameter search: (1) As the amount of target language corpus decreases, the optimal training approach shifts from monolingual single-stage training to multi-lingual two-stage training at a compute budget dependent threshold. (2) The optimal model scale remains stable regardless of the amount of target language corpus, allowing the use of the compute-optimal scale of monolingual training. (3) The optimal number of epochs can be extrapolated from smaller-scale experiments to larger scale using our proposed model. Also, we provide evidence that, in single-stage training, the target language validation loss follows a power law with respect to the target language ratio, with an exponent independent of the amount of data, model scale, and language pair.
comment: 16 pages, 10 figures
☆ Reversal of Thought: Enhancing Large Language Models with Preference-Guided Reverse Reasoning Warm-up
Large language models (LLMs) have shown remarkable performance in reasoning tasks but face limitations in mathematical and complex logical reasoning. Existing methods to improve LLMs' logical capabilities either involve traceable or verifiable logical sequences that generate more reliable responses by constructing logical structures yet increase computational costs, or introduces rigid logic template rules, reducing flexibility. In this paper, we propose Reversal of Thought (RoT), a novel framework aimed at enhancing the logical reasoning abilities of LLMs. RoT utilizes a Preference-Guided Reverse Reasoning warm-up strategy, which integrates logical symbols for pseudocode planning through meta-cognitive mechanisms and pairwise preference self-evaluation to generate task-specific prompts solely through demonstrations, aligning with LLMs' cognitive preferences shaped by Reinforcement Learning with Human Feedback (RLHF). Through reverse reasoning, we ultilize a Cognitive Preference Manager to assess knowledge boundaries and further expand LLMs' reasoning capabilities by aggregating solution logic for known tasks and stylistic templates for unknown tasks. Experiments across various tasks demonstrate that RoT surpasses existing baselines in both reasoning accuracy and efficiency.
☆ Open Domain Question Answering with Conflicting Contexts
Open domain question answering systems frequently rely on information retrieved from large collections of text (such as the Web) to answer questions. However, such collections of text often contain conflicting information, and indiscriminately depending on this information may result in untruthful and inaccurate answers. To understand the gravity of this problem, we collect a human-annotated dataset, Question Answering with Conflicting Contexts (QACC), and find that as much as 25% of unambiguous, open domain questions can lead to conflicting contexts when retrieved using Google Search. We evaluate and benchmark three powerful Large Language Models (LLMs) with our dataset QACC and demonstrate their limitations in effectively addressing questions with conflicting information. To explore how humans reason through conflicting contexts, we request our annotators to provide explanations for their selections of correct answers. We demonstrate that by finetuning LLMs to explain their answers, we can introduce richer information into their training that guide them through the process of reasoning with conflicting contexts.
☆ Semantics-Adaptive Activation Intervention for LLMs via Dynamic Steering Vectors
Large language models (LLMs) have achieved remarkable performance across many tasks, yet aligning them with desired behaviors remains challenging. Activation intervention has emerged as an effective and economical method to modify the behavior of LLMs. Despite considerable interest in this area, current intervention methods exclusively employ a fixed steering vector to modify model activations, lacking adaptability to diverse input semantics. To address this limitation, we propose Semantics-Adaptive Dynamic Intervention (SADI), a novel method that constructs a dynamic steering vector to intervene model activations at inference time. More specifically, SADI utilizes activation differences in contrastive pairs to precisely identify critical elements of an LLM (i.e., attention heads, hidden states, and neurons) for targeted intervention. During inference, SADI dynamically steers model behavior by scaling element-wise activations based on the directions of input semantics. Experimental results show that SADI outperforms established baselines by substantial margins, improving task performance without training. SADI's cost-effectiveness and generalizability across various LLM backbones and tasks highlight its potential as a versatile alignment technique. In addition, we release the code to foster research along this line:https://github.com/weixuan-wang123/SADI.
☆ Pyramid-Driven Alignment: Pyramid Principle Guided Integration of Large Language Models and Knowledge Graphs
Large Language Models (LLMs) possess impressive reasoning abilities but are prone to generating incorrect information, often referred to as hallucinations. While incorporating external Knowledge Graphs (KGs) can partially mitigate this issue, existing methods primarily treat KGs as static knowledge repositories, overlooking the critical disparity between KG and LLM knowledge, and failing to fully exploit the reasoning capabilities inherent in KGs. To address these limitations, we propose Pyramid-Driven Alignment (PDA), a novel framework for seamlessly integrating LLMs with KGs. PDA utilizes Pyramid Principle analysis to construct a hierarchical pyramid structure. This structure is designed to reflect the input question and generate more validated deductive knowledge, thereby enhancing the alignment of LLMs and KGs and ensuring more cohesive integration. Furthermore, PDA employs a recursive mechanism to harness the underlying reasoning abilities of KGs, resulting in more accurate knowledge retrieval for question-answering tasks. Our experimental results reveal a substantial performance advantage of PDA over state-of-the-art baselines, with improvements reaching 26.70% and 26.78%.
☆ Towards LLM-based Cognitive Models of Students with Misconceptions
Accurately modeling student cognition is crucial for developing effective AI-driven educational technologies. A key challenge is creating realistic student models that satisfy two essential properties: (1) accurately replicating specific misconceptions, and (2) correctly solving problems where these misconceptions are not applicable. This dual requirement reflects the complex nature of student understanding, where misconceptions coexist with correct knowledge. This paper investigates whether Large Language Models (LLMs) can be instruction-tuned to meet this dual requirement and effectively simulate student thinking in algebra. We introduce MalAlgoPy, a novel Python library that generates datasets reflecting authentic student solution patterns through a graph-based representation of algebraic problem-solving. Utilizing MalAlgoPy, we define and examine Cognitive Student Models (CSMs) - LLMs instruction tuned to faithfully emulate realistic student behavior. Our findings reveal that LLMs trained on misconception examples can efficiently learn to replicate errors. However, the training diminishes the model's ability to solve problems correctly, particularly for problem types where the misconceptions are not applicable, thus failing to satisfy second property of CSMs. We demonstrate that by carefully calibrating the ratio of correct to misconception examples in the training data - sometimes as low as 0.25 - it is possible to develop CSMs that satisfy both properties. Our insights enhance our understanding of AI-based student models and pave the way for effective adaptive learning systems.
☆ How much do contextualized representations encode long-range context?
We analyze contextual representations in neural autoregressive language models, emphasizing long-range contexts that span several thousand tokens. Our methodology employs a perturbation setup and the metric \emph{Anisotropy-Calibrated Cosine Similarity}, to capture the degree of contextualization of long-range patterns from the perspective of representation geometry. We begin the analysis with a case study on standard decoder-only Transformers, demonstrating that similar perplexity can exhibit markedly different downstream task performance, which can be explained by the difference in contextualization of long-range content. Next, we extend the analysis to other models, covering recent novel architectural designs and various training configurations. The representation-level results illustrate a reduced capacity for high-complexity (i.e., less compressible) sequences across architectures, and that fully recurrent models rely heavily on local context, whereas hybrid models more effectively encode the entire sequence structure. Finally, preliminary analysis of model size and training configurations on the encoding of long-range context suggest potential directions for improving existing language models.
comment: 17 pages, 9 figures
☆ A Prompt-Based Knowledge Graph Foundation Model for Universal In-Context Reasoning NeurIPS 2024
Extensive knowledge graphs (KGs) have been constructed to facilitate knowledge-driven tasks across various scenarios. However, existing work usually develops separate reasoning models for different KGs, lacking the ability to generalize and transfer knowledge across diverse KGs and reasoning settings. In this paper, we propose a prompt-based KG foundation model via in-context learning, namely KG-ICL, to achieve a universal reasoning ability. Specifically, we introduce a prompt graph centered with a query-related example fact as context to understand the query relation. To encode prompt graphs with the generalization ability to unseen entities and relations in queries, we first propose a unified tokenizer that maps entities and relations in prompt graphs to predefined tokens. Then, we propose two message passing neural networks to perform prompt encoding and KG reasoning, respectively. We conduct evaluation on 43 different KGs in both transductive and inductive settings. Results indicate that the proposed KG-ICL outperforms baselines on most datasets, showcasing its outstanding generalization and universal reasoning capabilities. The source code is accessible on GitHub: https://github.com/nju-websoft/KG-ICL.
comment: Accepted in the 38th Conference on Neural Information Processing Systems (NeurIPS 2024)
☆ Fool Me Once? Contrasting Textual and Visual Explanations in a Clinical Decision-Support Setting
The growing capabilities of AI models are leading to their wider use, including in safety-critical domains. Explainable AI (XAI) aims to make these models safer to use by making their inference process more transparent. However, current explainability methods are seldom evaluated in the way they are intended to be used: by real-world end users. To address this, we conducted a large-scale user study with 85 healthcare practitioners in the context of human-AI collaborative chest X-ray analysis. We evaluated three types of explanations: visual explanations (saliency maps), natural language explanations, and a combination of both modalities. We specifically examined how different explanation types influence users depending on whether the AI advice and explanations are factually correct. We find that text-based explanations lead to significant over-reliance, which is alleviated by combining them with saliency maps. We also observe that the quality of explanations, that is, how much factually correct information they entail, and how much this aligns with AI correctness, significantly impacts the usefulness of the different explanation types.
☆ Beyond Oversmoothing: Evaluating DDPM and MSE for Scalable Speech Synthesis in ASR ICASSP 2025
Synthetically generated speech has rapidly approached human levels of naturalness. However, the paradox remains that ASR systems, when trained on TTS output that is judged as natural by humans, continue to perform badly on real speech. In this work, we explore whether this phenomenon is due to the oversmoothing behaviour of models commonly used in TTS, with a particular focus on the behaviour of TTS-for-ASR as the amount of TTS training data is scaled up. We systematically compare Denoising Diffusion Probabilistic Models (DDPM) to Mean Squared Error (MSE) based models for TTS, when used for ASR model training. We test the scalability of the two approaches, varying both the number hours, and the number of different speakers. We find that for a given model size, DDPM can make better use of more data, and a more diverse set of speakers, than MSE models. We achieve the best reported ratio between real and synthetic speech WER to date (1.46), but also find that a large gap remains.
comment: Under review at ICASSP 2025
☆ Controlled Automatic Task-Specific Synthetic Data Generation for Hallucination Detection
We present a novel approach to automatically generate non-trivial task-specific synthetic datasets for hallucination detection. Our approach features a two-step generation-selection pipeline, using hallucination pattern guidance and a language style alignment during generation. Hallucination pattern guidance leverages the most important task-specific hallucination patterns while language style alignment aligns the style of the synthetic dataset with benchmark text. To obtain robust supervised detectors from synthetic datasets, we also adopt a data mixture strategy to improve performance robustness and generalization. Our results on three datasets show that our generated hallucination text is more closely aligned with non-hallucinated text versus baselines, to train hallucination detectors with better generalization. Our hallucination detectors trained on synthetic datasets outperform in-context-learning (ICL)-based detectors by a large margin of 32%. Our extensive experiments confirm the benefits of our approach with cross-task and cross-generator generalization. Our data-mixture-based training further improves the generalization and robustness of hallucination detection.
☆ Kallini et al. (2024) do not compare impossible languages with constituency-based ones
A central goal of linguistic theory is to find a precise characterization of the notion "possible human language", in the form of a computational device that is capable of describing all and only the languages that can be acquired by a typically developing human child. The success of recent large language models (LLMs) in NLP applications arguably raises the possibility that LLMs might be computational devices that meet this goal. This would only be the case if, in addition to succeeding in learning human languages, LLMs struggle to learn "impossible" human languages. Kallini et al. (2024; "Mission: Impossible Language Models", Proc. ACL) conducted experiments aiming to test this by training GPT-2 on a variety of synthetic languages, and found that it learns some more successfully than others. They present these asymmetries as support for the idea that LLMs' inductive biases align with what is regarded as "possible" for human languages, but the most significant comparison has a confound that makes this conclusion unwarranted. In this paper I explain the confound and suggest some ways forward towards constructing a comparison that appropriately tests the underlying issue.
☆ An Automatic and Cost-Efficient Peer-Review Framework for Language Generation Evaluation
With the rapid development of large language models (LLMs), how to efficiently evaluate them has become an important research question. Existing evaluation methods often suffer from high costs, limited test formats, the need of human references, and systematic evaluation biases. To address these limitations, our study introduces the Auto-PRE, an automatic LLM evaluation framework based on peer review. In contrast to previous studies that rely on human annotations, Auto-PRE selects evaluator LLMs automatically based on their inherent traits including consistency, self-confidence, and pertinence. We conduct extensive experiments on three tasks: summary generation, non-factoid question-answering, and dialogue generation. Experimental results indicate our Auto-PRE achieves state-of-the-art performance at a lower cost. Moreover, our study highlights the impact of prompt strategies and evaluation formats on evaluation performance, offering guidance for method optimization in the future.
☆ CoFE-RAG: A Comprehensive Full-chain Evaluation Framework for Retrieval-Augmented Generation with Enhanced Data Diversity
Retrieval-Augmented Generation (RAG) aims to enhance large language models (LLMs) to generate more accurate and reliable answers with the help of the retrieved context from external knowledge sources, thereby reducing the incidence of hallucinations. Despite the advancements, evaluating these systems remains a crucial research area due to the following issues: (1) Limited data diversity: The insufficient diversity of knowledge sources and query types constrains the applicability of RAG systems; (2) Obscure problems location: Existing evaluation methods have difficulty in locating the stage of the RAG pipeline where problems occur; (3) Unstable retrieval evaluation: These methods often fail to effectively assess retrieval performance, particularly when the chunking strategy changes. To tackle these challenges, we propose a Comprehensive Full-chain Evaluation (CoFE-RAG) framework to facilitate thorough evaluation across the entire RAG pipeline, including chunking, retrieval, reranking, and generation. To effectively evaluate the first three phases, we introduce multi-granularity keywords, including coarse-grained and fine-grained keywords, to assess the retrieved context instead of relying on the annotation of golden chunks. Moreover, we release a holistic benchmark dataset tailored for diverse data scenarios covering a wide range of document formats and query types. We demonstrate the utility of the CoFE-RAG framework by conducting experiments to evaluate each stage of RAG systems. Our evaluation method provides unique insights into the effectiveness of RAG systems in handling diverse data scenarios, offering a more nuanced understanding of their capabilities and limitations.
☆ EPS-MoE: Expert Pipeline Scheduler for Cost-Efficient MoE Inference
Large Language Model (LLM) has revolutionized the field of artificial intelligence, with their capabilities expanding rapidly due to advances in deep learning and increased computational resources. The mixture-of-experts (MoE) model has emerged as a prominent architecture in the field of LLM, better balancing the model performance and computational efficiency. MoE architecture allows for effective scaling and efficient parallel processing, but the GEMM (General Matrix Multiply) of MoE and the large parameters introduce challenges in terms of computation efficiency and communication overhead, which becomes the throughput bottleneck during inference. Applying a single parallelism strategy like EP, DP, PP, etc. to MoE architecture usually achieves sub-optimal inference throughput, the straightforward combinations of existing different parallelisms on MoE can not obtain optimal inference throughput yet. This paper introduces EPS-MoE, a novel expert pipeline scheduler for MoE that goes beyond the existing inference parallelism schemes. Our approach focuses on optimizing the computation of MoE FFN (FeedForward Network) modules by dynamically selecting the best kernel implementation of GroupGemm and DenseGemm for different loads and adaptively overlapping these computations with \textit{all2all} communication, leading to a substantial increase in throughput. Our experimental results demonstrate an average 21% improvement in prefill throughput over existing parallel inference methods. Specifically, we validated our method on DeepSeekV2, a highly optimized model claimed to achieve a prefill throughput of 100K tokens per second. By applying EPS-MoE, we further accelerated it to at least 120K tokens per second.
comment: 13 pages, 14 figures
☆ Triple Modality Fusion: Aligning Visual, Textual, and Graph Data with Large Language Models for Multi-Behavior Recommendations
Integrating diverse data modalities is crucial for enhancing the performance of personalized recommendation systems. Traditional models, which often rely on singular data sources, lack the depth needed to accurately capture the multifaceted nature of item features and user behaviors. This paper introduces a novel framework for multi-behavior recommendations, leveraging the fusion of triple-modality, which is visual, textual, and graph data through alignment with large language models (LLMs). By incorporating visual information, we capture contextual and aesthetic item characteristics; textual data provides insights into user interests and item features in detail; and graph data elucidates relationships within the item-behavior heterogeneous graphs. Our proposed model called Triple Modality Fusion (TMF) utilizes the power of LLMs to align and integrate these three modalities, achieving a comprehensive representation of user behaviors. The LLM models the user's interactions including behaviors and item features in natural languages. Initially, the LLM is warmed up using only natural language-based prompts. We then devise the modality fusion module based on cross-attention and self-attention mechanisms to integrate different modalities from other models into the same embedding space and incorporate them into an LLM. Extensive experiments demonstrate the effectiveness of our approach in improving recommendation accuracy. Further ablation studies validate the effectiveness of our model design and benefits of the TMF.
☆ On A Scale From 1 to 5: Quantifying Hallucination in Faithfulness Evaluation
Hallucination has been a popular topic in natural language generation (NLG). In real-world applications, unfaithful content can result in bad data quality or loss of trust from end users. Thus, it is crucial to fact-check before adopting NLG for production usage, which can be expensive if done manually. In this paper, we investigate automated faithfulness evaluation in guided NLG. We developed a rubrics template and use large language models (LLMs) to score the generation into quantifiable scales. We compared popular LLMs as well as the widely adopted natural language inference (NLI) models in scoring quality and sensitivity. In addition, we developed methods to generation synthetic unfaithful data, as well as a heuristics to quantify the percentage of hallucination. Our results on 4 travel-domain industry dataset show that GPT-4 can provide accurate judgement and explanation on whether a source and a generation are factually consistent. Furthermore, we found that tuning NLI models on synthetic data can improve performance. Lastly, we present insights on latency and cost for deploying such system.
comment: 14 pages, 13 figures
☆ OmnixR: Evaluating Omni-modality Language Models on Reasoning across Modalities
We introduce OmnixR, an evaluation suite designed to benchmark SoTA Omni-modality Language Models, such as GPT-4o and Gemini. Evaluating OLMs, which integrate multiple modalities such as text, vision, and audio, presents unique challenges. Particularly, the user message might often consist of multiple modalities, such that OLMs have to establish holistic understanding and reasoning across modalities to accomplish the task. Existing benchmarks are limited to single modality or dual-modality tasks, overlooking comprehensive multi-modal assessments of model reasoning. To address this, OmnixR offers two evaluation variants: (1)synthetic subset: a synthetic dataset generated automatically by translating text into multiple modalities--audio, images, video, and hybrids (Omnify). (2)realistic subset: a real-world dataset, manually curated and annotated by experts, for evaluating cross-modal reasoning in natural settings. OmnixR presents a unique evaluation towards assessing OLMs over a diverse mix of modalities, such as a question that involves video, audio, and text, providing a rigorous cross-modal reasoning testbed unlike any existing benchmarks. Our experiments find that all state-of-the-art OLMs struggle with OmnixR questions that require integrating information from multiple modalities to answer. Further analysis highlights differences in reasoning behavior, underscoring the challenges of omni-modal AI alignment.
comment: 19 pages, 6 figures, 12 tables
☆ Accurate and Data-Efficient Toxicity Prediction when Annotators Disagree
When annotators disagree, predicting the labels given by individual annotators can capture nuances overlooked by traditional label aggregation. We introduce three approaches to predicting individual annotator ratings on the toxicity of text by incorporating individual annotator-specific information: a neural collaborative filtering (NCF) approach, an in-context learning (ICL) approach, and an intermediate embedding-based architecture. We also study the utility of demographic information for rating prediction. NCF showed limited utility; however, integrating annotator history, demographics, and survey information permits both the embedding-based architecture and ICL to substantially improve prediction accuracy, with the embedding-based architecture outperforming the other methods. We also find that, if demographics are predicted from survey information, using these imputed demographics as features performs comparably to using true demographic data. This suggests that demographics may not provide substantial information for modeling ratings beyond what is captured in survey responses. Our findings raise considerations about the relative utility of different types of annotator information and provide new approaches for modeling annotators in subjective NLP tasks.
☆ Negative-Prompt-driven Alignment for Generative Language Model
Large language models have achieved remarkable capabilities, but aligning their outputs with human values and preferences remains a significant challenge. Existing alignment methods primarily focus on positive examples while overlooking the importance of negative responses in guiding models away from undesirable behaviors. For instance, the widely-used alignment datasets reveals a scarcity of explicit negative examples that contradict human values, hindering its ability to discourage harmful or biased outputs during training. To address this limitation, we propose NEAT, i.e., NEgative-prompt-driven AlignmenT, to introduce negative prompts to generate undesirable responses alongside positive examples during the optimization process. NEAT explicitly penalizes the model for producing harmful outputs, guiding it not only toward desirable behaviors but also steering it away from generating undesirable, biased responses. This dual feedback mechanism enables better alignment with human preferences, crucial in contexts where avoiding harm is paramount. Starting from a pre-trained language model, NEAT performs online alignment by incorporating a ranking loss derived from an expanded preference dataset containing both positive and negative examples. Extensive experiments validate NEAT's effectiveness in significantly enhancing language models' alignment with human values and preferences.
☆ Exploring Large Language Models for Hate Speech Detection in Rioplatense Spanish
Hate speech detection deals with many language variants, slang, slurs, expression modalities, and cultural nuances. This outlines the importance of working with specific corpora, when addressing hate speech within the scope of Natural Language Processing, recently revolutionized by the irruption of Large Language Models. This work presents a brief analysis of the performance of large language models in the detection of Hate Speech for Rioplatense Spanish. We performed classification experiments leveraging chain-of-thought reasoning with ChatGPT 3.5, Mixtral, and Aya, comparing their results with those of a state-of-the-art BERT classifier. These experiments outline that, even if large language models show a lower precision compared to the fine-tuned BERT classifier and, in some cases, they find hard-to-get slurs or colloquialisms, they still are sensitive to highly nuanced cases (particularly, homophobic/transphobic hate speech). We make our code and models publicly available for future research.
☆ Table-LLM-Specialist: Language Model Specialists for Tables using Iterative Generator-Validator Fine-tuning
In this work, we propose Table-LLM-Specialist, or Table-Specialist for short, as a new self-trained fine-tuning paradigm specifically designed for table tasks. Our insight is that for each table task, there often exist two dual versions of the same task, one generative and one classification in nature. Leveraging their duality, we propose a Generator-Validator paradigm, to iteratively generate-then-validate training data from language-models, to fine-tune stronger \sys models that can specialize in a given task, without requiring manually-labeled data. Our extensive evaluations suggest that our Table-Specialist has (1) \textit{strong performance} on diverse table tasks over vanilla language-models -- for example, Table-Specialist fine-tuned on GPT-3.5 not only outperforms vanilla GPT-3.5, but can often match or surpass GPT-4 level quality, (2) \textit{lower cost} to deploy, because when Table-Specialist fine-tuned on GPT-3.5 achieve GPT-4 level quality, it becomes possible to deploy smaller models with lower latency and inference cost, with comparable quality, and (3) \textit{better generalizability} when evaluated across multiple benchmarks, since \sys is fine-tuned on a broad range of training data systematically generated from diverse real tables. Our code and data will be available at https://github.com/microsoft/Table-LLM-Specialist.
☆ Exploiting LLMs' Reasoning Capability to Infer Implicit Concepts in Legal Information Retrieval KR
Statutory law retrieval is a typical problem in legal language processing, that has various practical applications in law engineering. Modern deep learning-based retrieval methods have achieved significant results for this problem. However, retrieval systems relying on semantic and lexical correlations often exhibit limitations, particularly when handling queries that involve real-life scenarios, or use the vocabulary that is not specific to the legal domain. In this work, we focus on overcoming this weaknesses by utilizing the logical reasoning capabilities of large language models (LLMs) to identify relevant legal terms and facts related to the situation mentioned in the query. The proposed retrieval system integrates additional information from the term--based expansion and query reformulation to improve the retrieval accuracy. The experiments on COLIEE 2022 and COLIEE 2023 datasets show that extra knowledge from LLMs helps to improve the retrieval result of both lexical and semantic ranking models. The final ensemble retrieval system outperformed the highest results among all participating teams in the COLIEE 2022 and 2023 competitions.
comment: Presented at NeLaMKRR@KR, 2024 (arXiv:2410.05339)
☆ Layer-of-Thoughts Prompting (LoT): Leveraging LLM-Based Retrieval with Constraint Hierarchies KR
This paper presents a novel approach termed Layer-of-Thoughts Prompting (LoT), which utilizes constraint hierarchies to filter and refine candidate responses to a given query. By integrating these constraints, our method enables a structured retrieval process that enhances explainability and automation. Existing methods have explored various prompting techniques but often present overly generalized frameworks without delving into the nuances of prompts in multi-turn interactions. Our work addresses this gap by focusing on the hierarchical relationships among prompts. We demonstrate that the efficacy of thought hierarchy plays a critical role in developing efficient and interpretable retrieval algorithms. Leveraging Large Language Models (LLMs), LoT significantly improves the accuracy and comprehensibility of information retrieval tasks.
comment: Presented at NeLaMKRR@KR, 2024 (arXiv:2410.05339)
☆ Preference Optimization with Multi-Sample Comparisons
Recent advancements in generative models, particularly large language models (LLMs) and diffusion models, have been driven by extensive pretraining on large datasets followed by post-training. However, current post-training methods such as reinforcement learning from human feedback (RLHF) and direct alignment from preference methods (DAP) primarily utilize single-sample comparisons. These approaches often fail to capture critical characteristics such as generative diversity and bias, which are more accurately assessed through multiple samples. To address these limitations, we introduce a novel approach that extends post-training to include multi-sample comparisons. To achieve this, we propose Multi-sample Direct Preference Optimization (mDPO) and Multi-sample Identity Preference Optimization (mIPO). These methods improve traditional DAP methods by focusing on group-wise characteristics. Empirically, we demonstrate that multi-sample comparison is more effective in optimizing collective characteristics~(e.g., diversity and bias) for generative models than single-sample comparison. Additionally, our findings suggest that multi-sample comparisons provide a more robust optimization framework, particularly for dataset with label noise.
comment: preprint
☆ Iter-AHMCL: Alleviate Hallucination for Large Language Model via Iterative Model-level Contrastive Learning
The development of Large Language Models (LLMs) has significantly advanced various AI applications in commercial and scientific research fields, such as scientific literature summarization, writing assistance, and knowledge graph construction. However, a significant challenge is the high risk of hallucination during LLM inference, which can lead to security concerns like factual inaccuracies, inconsistent information, and fabricated content. To tackle this issue, it is essential to develop effective methods for reducing hallucination while maintaining the original capabilities of the LLM. This paper introduces a novel approach called Iterative Model-level Contrastive Learning (Iter-AHMCL) to address hallucination. This method modifies the representation layers of pre-trained LLMs by using contrastive `positive' and `negative' models, trained on data with and without hallucinations. By leveraging the differences between these two models, we create a more straightforward pathway to eliminate hallucinations, and the iterative nature of contrastive learning further enhances performance. Experimental validation on four pre-trained foundation LLMs (LLaMA2, Alpaca, LLaMA3, and Qwen) finetuning with a specially designed dataset shows that our approach achieves an average improvement of 10.1 points on the TruthfulQA benchmark. Comprehensive experiments demonstrate the effectiveness of Iter-AHMCL in reducing hallucination while maintaining the general capabilities of LLMs.
☆ Communication-Efficient and Tensorized Federated Fine-Tuning of Large Language Models
Parameter-efficient fine-tuning (PEFT) methods typically assume that Large Language Models (LLMs) are trained on data from a single device or client. However, real-world scenarios often require fine-tuning these models on private data distributed across multiple devices. Federated Learning (FL) offers an appealing solution by preserving user privacy, as sensitive data remains on local devices during training. Nonetheless, integrating PEFT methods into FL introduces two main challenges: communication overhead and data heterogeneity. In this paper, we introduce FedTT and FedTT+, methods for adapting LLMs by integrating tensorized adapters into client-side models' encoder/decoder blocks. FedTT is versatile and can be applied to both cross-silo FL and large-scale cross-device FL. FedTT+, an extension of FedTT tailored for cross-silo FL, enhances robustness against data heterogeneity by adaptively freezing portions of tensor factors, further reducing the number of trainable parameters. Experiments on BERT and LLaMA models demonstrate that our proposed methods successfully address data heterogeneity challenges and perform on par or even better than existing federated PEFT approaches while achieving up to 10$\times$ reduction in communication cost.
☆ Self-Comparison for Dataset-Level Membership Inference in Large (Vision-)Language Models
Large Language Models (LLMs) and Vision-Language Models (VLMs) have made significant advancements in a wide range of natural language processing and vision-language tasks. Access to large web-scale datasets has been a key factor in their success. However, concerns have been raised about the unauthorized use of copyrighted materials and potential copyright infringement. Existing methods, such as sample-level Membership Inference Attacks (MIA) and distribution-based dataset inference, distinguish member data (data used for training) and non-member data by leveraging the common observation that models tend to memorize and show greater confidence in member data. Nevertheless, these methods face challenges when applied to LLMs and VLMs, such as the requirement for ground-truth member data or non-member data that shares the same distribution as the test data. In this paper, we propose a novel dataset-level membership inference method based on Self-Comparison. We find that a member prefix followed by a non-member suffix (paraphrased from a member suffix) can further trigger the model's memorization on training data. Instead of directly comparing member and non-member data, we introduce paraphrasing to the second half of the sequence and evaluate how the likelihood changes before and after paraphrasing. Unlike prior approaches, our method does not require access to ground-truth member data or non-member data in identical distribution, making it more practical. Extensive experiments demonstrate that our proposed method outperforms traditional MIA and dataset inference techniques across various datasets and models, including including public models, fine-tuned models, and API-based commercial models.
☆ Reverse-Engineering the Reader
Numerous previous studies have sought to determine to what extent language models, pretrained on natural language text, can serve as useful models of human cognition. In this paper, we are interested in the opposite question: whether we can directly optimize a language model to be a useful cognitive model by aligning it to human psychometric data. To achieve this, we introduce a novel alignment technique in which we fine-tune a language model to implicitly optimize the parameters of a linear regressor that directly predicts humans' reading times of in-context linguistic units, e.g., phonemes, morphemes, or words, using surprisal estimates derived from the language model. Using words as a test case, we evaluate our technique across multiple model sizes and datasets and find that it improves language models' psychometric predictive power. However, we find an inverse relationship between psychometric power and a model's performance on downstream NLP tasks as well as its perplexity on held-out test data. While this latter trend has been observed before (Oh et al., 2022; Shain et al., 2024), we are the first to induce it by manipulating a model's alignment to psychometric data.
☆ MMed-RAG: Versatile Multimodal RAG System for Medical Vision Language Models
Artificial Intelligence (AI) has demonstrated significant potential in healthcare, particularly in disease diagnosis and treatment planning. Recent progress in Medical Large Vision-Language Models (Med-LVLMs) has opened up new possibilities for interactive diagnostic tools. However, these models often suffer from factual hallucination, which can lead to incorrect diagnoses. Fine-tuning and retrieval-augmented generation (RAG) have emerged as methods to address these issues. However, the amount of high-quality data and distribution shifts between training data and deployment data limit the application of fine-tuning methods. Although RAG is lightweight and effective, existing RAG-based approaches are not sufficiently general to different medical domains and can potentially cause misalignment issues, both between modalities and between the model and the ground truth. In this paper, we propose a versatile multimodal RAG system, MMed-RAG, designed to enhance the factuality of Med-LVLMs. Our approach introduces a domain-aware retrieval mechanism, an adaptive retrieved contexts selection method, and a provable RAG-based preference fine-tuning strategy. These innovations make the RAG process sufficiently general and reliable, significantly improving alignment when introducing retrieved contexts. Experimental results across five medical datasets (involving radiology, ophthalmology, pathology) on medical VQA and report generation demonstrate that MMed-RAG can achieve an average improvement of 43.8% in the factual accuracy of Med-LVLMs. Our data and code are available in https://github.com/richard-peng-xia/MMed-RAG.
☆ Graph-constrained Reasoning: Faithful Reasoning on Knowledge Graphs with Large Language Models
Large language models (LLMs) have demonstrated impressive reasoning abilities, but they still struggle with faithful reasoning due to knowledge gaps and hallucinations. To address these issues, knowledge graphs (KGs) have been utilized to enhance LLM reasoning through their structured knowledge. However, existing KG-enhanced methods, either retrieval-based or agent-based, encounter difficulties in accurately retrieving knowledge and efficiently traversing KGs at scale. In this work, we introduce graph-constrained reasoning (GCR), a novel framework that bridges structured knowledge in KGs with unstructured reasoning in LLMs. To eliminate hallucinations, GCR ensures faithful KG-grounded reasoning by integrating KG structure into the LLM decoding process through KG-Trie, a trie-based index that encodes KG reasoning paths. KG-Trie constrains the decoding process, allowing LLMs to directly reason on graphs and generate faithful reasoning paths grounded in KGs. Additionally, GCR leverages a lightweight KG-specialized LLM for graph-constrained reasoning alongside a powerful general LLM for inductive reasoning over multiple reasoning paths, resulting in accurate reasoning with zero reasoning hallucination. Extensive experiments on several KGQA benchmarks demonstrate that GCR achieves state-of-the-art performance and exhibits strong zero-shot generalizability to unseen KGs without additional training.
comment: 21 pages, 10 figures
☆ Tuning Language Models by Mixture-of-Depths Ensemble
Transformer-based Large Language Models (LLMs) traditionally rely on final-layer loss for training and final-layer representations for predictions, potentially overlooking the predictive power embedded in intermediate layers. Surprisingly, we find that focusing training efforts on these intermediate layers can yield training losses comparable to those of final layers, with complementary test-time performance. We introduce a novel tuning framework, Mixture-of-Depths (MoD), which trains late layers as ensembles contributing to the final logits through learned routing weights. With the auxiliary distillation loss and additional normalization modules, we ensure that the outputs of the late layers adapt to language modeling. Our MoD framework, which can be integrated with any existing tuning method, shows consistent improvement on various language modelling tasks. Furthermore, by replacing traditional trainable modules with MoD, our approach achieves similar performance with significantly fewer trainable parameters, demonstrating the potential of leveraging predictive power from intermediate representations during training.
PromptExp: Multi-granularity Prompt Explanation of Large Language Models
Large Language Models excel in tasks like natural language understanding and text generation. Prompt engineering plays a critical role in leveraging LLM effectively. However, LLMs black-box nature hinders its interpretability and effective prompting engineering. A wide range of model explanation approaches have been developed for deep learning models, However, these local explanations are designed for single-output tasks like classification and regression,and cannot be directly applied to LLMs, which generate sequences of tokens. Recent efforts in LLM explanation focus on natural language explanations, but they are prone to hallucinations and inaccuracies. To address this, we introduce OurTool, a framework for multi-granularity prompt explanations by aggregating token-level insights. OurTool introduces two token-level explanation approaches: 1.an aggregation-based approach combining local explanation techniques, and 2. a perturbation-based approach with novel techniques to evaluate token masking impact. OurTool supports both white-box and black-box explanations and extends explanations to higher granularity levels, enabling flexible analysis. We evaluate OurTool in case studies such as sentiment analysis, showing the perturbation-based approach performs best using semantic similarity to assess perturbation impact. Furthermore, we conducted a user study to confirm OurTool's accuracy and practical value, and demonstrate its potential to enhance LLM interpretability.
comment: 11 pages
☆ Is Semantic Chunking Worth the Computational Cost?
Recent advances in Retrieval-Augmented Generation (RAG) systems have popularized semantic chunking, which aims to improve retrieval performance by dividing documents into semantically coherent segments. Despite its growing adoption, the actual benefits over simpler fixed-size chunking, where documents are split into consecutive, fixed-size segments, remain unclear. This study systematically evaluates the effectiveness of semantic chunking using three common retrieval-related tasks: document retrieval, evidence retrieval, and retrieval-based answer generation. The results show that the computational costs associated with semantic chunking are not justified by consistent performance gains. These findings challenge the previous assumptions about semantic chunking and highlight the need for more efficient chunking strategies in RAG systems.
☆ Language Models as Semiotic Machines: Reconceptualizing AI Language Systems through Structuralist and Post-Structuralist Theories of Language
This paper proposes a novel framework for understanding large language models (LLMs) by reconceptualizing them as semiotic machines rather than as imitations of human cognition. Drawing from structuralist and post-structuralist theories of language-specifically the works of Ferdinand de Saussure and Jacques Derrida-I argue that LLMs should be understood as models of language itself, aligning with Derrida's concept of 'writing' (l'ecriture). The paper is structured into three parts. First, I lay the theoretical groundwork by explaining how the word2vec embedding algorithm operates within Saussure's framework of language as a relational system of signs. Second, I apply Derrida's critique of Saussure to position 'writing' as the object modeled by LLMs, offering a view of the machine's 'mind' as a statistical approximation of sign behavior. Finally, the third section addresses how modern LLMs reflect post-structuralist notions of unfixed meaning, arguing that the "next token generation" mechanism effectively captures the dynamic nature of meaning. By reconceptualizing LLMs as semiotic machines rather than cognitive models, this framework provides an alternative lens through which to assess the strengths and limitations of LLMs, offering new avenues for future research.
comment: 18 pages, 2 figures
☆ ERAS: Evaluating the Robustness of Chinese NLP Models to Morphological Garden Path Errors NAACL
In languages without orthographic word boundaries, NLP models perform word segmentation, either as an explicit preprocessing step or as an implicit step in an end-to-end computation. This paper shows that Chinese NLP models are vulnerable to morphological garden path errors: errors caused by a failure to resolve local word segmentation ambiguities using sentence-level morphosyntactic context. We propose a benchmark, ERAS, that tests a model's vulnerability to morphological garden path errors by comparing its behavior on sentences with and without local segmentation ambiguities. Using ERAS, we show that word segmentation models make garden path errors on locally ambiguous sentences, but do not make equivalent errors on unambiguous sentences. We further show that sentiment analysis models with character-level tokenization make implicit garden path errors, even without an explicit word segmentation step in the pipeline. Our results indicate that models' segmentation of Chinese text often fails to account for morphosyntactic context.
comment: Under review in ARR/NAACL
♻ ☆ BIRD: A Trustworthy Bayesian Inference Framework for Large Language Models
Predictive models often need to work with incomplete information in real-world tasks. Consequently, they must provide reliable probability or confidence estimation, especially in large-scale decision making and planning tasks. Current large language models (LLM) are insufficient for such accurate estimations, but they can generate relevant factors that may affect the probabilities, produce coarse-grained probabilities when the information is more complete, and help determine which factors are relevant to specific downstream contexts. In this paper, we make use of these capabilities of LLMs to provide a significantly more accurate probabilistic estimation. We propose BIRD, a novel probabilistic inference framework that aligns a Bayesian network with LLM abductions and then estimates more accurate probabilities in a deduction step. We show BIRD provides reliable probability estimations that are 30\% better than those provided directly by LLM baselines. These estimates can further contribute to better and more trustworthy decision-making.
♻ ☆ Energy and Carbon Considerations of Fine-Tuning BERT EMNLP 2023
Despite the popularity of the `pre-train then fine-tune' paradigm in the NLP community, existing work quantifying energy costs and associated carbon emissions has largely focused on language model pre-training. Although a single pre-training run draws substantially more energy than fine-tuning, fine-tuning is performed more frequently by many more individual actors, and thus must be accounted for when considering the energy and carbon footprint of NLP. In order to better characterize the role of fine-tuning in the landscape of energy and carbon emissions in NLP, we perform a careful empirical study of the computational costs of fine-tuning across tasks, datasets, hardware infrastructure and measurement modalities. Our experimental results allow us to place fine-tuning energy and carbon costs into perspective with respect to pre-training and inference, and outline recommendations to NLP researchers and practitioners who wish to improve their fine-tuning energy efficiency.
comment: EMNLP 2023 Findings; First two authors contributed equally; 12 pages
♻ ☆ ÚFAL CorPipe at CRAC 2023: Larger Context Improves Multilingual Coreference Resolution
We present CorPipe, the winning entry to the CRAC 2023 Shared Task on Multilingual Coreference Resolution. Our system is an improved version of our earlier multilingual coreference pipeline, and it surpasses other participants by a large margin of 4.5 percent points. CorPipe first performs mention detection, followed by coreference linking via an antecedent-maximization approach on the retrieved spans. Both tasks are trained jointly on all available corpora using a shared pretrained language model. Our main improvements comprise inputs larger than 512 subwords and changing the mention decoding to support ensembling. The source code is available at https://github.com/ufal/crac2023-corpipe.
comment: Accepted to CRAC 2023 (the Sixth Workshop on Computational Models of Reference, Anaphora and Coreference)
♻ ☆ ÚFAL CorPipe at CRAC 2022: Effectivity of Multilingual Models for Coreference Resolution
We describe the winning submission to the CRAC 2022 Shared Task on Multilingual Coreference Resolution. Our system first solves mention detection and then coreference linking on the retrieved spans with an antecedent-maximization approach, and both tasks are fine-tuned jointly with shared Transformer weights. We report results of fine-tuning a wide range of pretrained models. The center of this contribution are fine-tuned multilingual models. We found one large multilingual model with sufficiently large encoder to increase performance on all datasets across the board, with the benefit not limited only to the underrepresented languages or groups of typologically relative languages. The source code is available at https://github.com/ufal/crac2022-corpipe.
comment: Accepted to CRAC 2022 (Fifth Workshop on Computational Models of Reference, Anaphora and Coreference)
♻ ☆ Open-Source Conversational AI with SpeechBrain 1.0
SpeechBrain is an open-source Conversational AI toolkit based on PyTorch, focused particularly on speech processing tasks such as speech recognition, speech enhancement, speaker recognition, text-to-speech, and much more. It promotes transparency and replicability by releasing both the pre-trained models and the complete "recipes" of code and algorithms required for training them. This paper presents SpeechBrain 1.0, a significant milestone in the evolution of the toolkit, which now has over 200 recipes for speech, audio, and language processing tasks, and more than 100 models available on Hugging Face. SpeechBrain 1.0 introduces new technologies to support diverse learning modalities, Large Language Model (LLM) integration, and advanced decoding strategies, along with novel models, tasks, and modalities. It also includes a new benchmark repository, offering researchers a unified platform for evaluating models across diverse tasks.
comment: Accepted to the Journal of Machine Learning research (JMLR), Machine Learning Open Source Software
♻ ☆ Unsupervised End-to-End Task-Oriented Dialogue with LLMs: The Power of the Noisy Channel EMNLP 2024
Training task-oriented dialogue systems typically requires turn-level annotations for interacting with their APIs: e.g. a dialogue state and the system actions taken at each step. These annotations can be costly to produce, error-prone, and require both domain and annotation expertise. With advances in LLMs, we hypothesize that unlabeled data and a schema definition are sufficient for building a working task-oriented dialogue system, completely unsupervised. We consider a novel unsupervised setting of only (1) a well-defined API schema (2) a set of unlabeled dialogues between a user and agent. We propose an innovative approach using expectation-maximization (EM) that infers turn-level annotations as latent variables using a noisy channel model to build an end-to-end dialogue agent. Evaluating our approach on the MultiWOZ benchmark, our method more than doubles the dialogue success rate of a strong GPT-3.5 baseline.
comment: To be presented at Empirical Methods in Natural Language Processing (EMNLP 2024). 18 Pages, 8 Figures
♻ ☆ Understanding Figurative Meaning through Explainable Visual Entailment
Large Vision-Language Models (VLMs) have demonstrated strong capabilities in tasks requiring a fine-grained understanding of literal meaning in images and text, such as visual question-answering or visual entailment. However, there has been little exploration of these models' capabilities when presented with images and captions containing figurative meaning, such as metaphors or humor. To close this gap, we propose a new task framing the figurative meaning understanding problem as an explainable visual entailment task, where the model has to predict whether the image (premise) entails a caption (hypothesis) and justify the predicted label with a textual explanation. The figurative phenomena can be present either in the image, the caption, or both. Utilizing a human-AI collaboration approach, we build the accompanying expert-verified dataset V-FLUTE, containing 6,027 {image, caption, label, explanation} instances spanning five diverse figurative phenomena: metaphors, similes, idioms, sarcasm, and humor. Through automatic evaluation, we find that VLMs struggle to generalize from literal to figurative meaning, particularly when it is present in images. Further, we identify common types of errors in VLM reasoning via human evaluation.
♻ ☆ ToBlend: Token-Level Blending With an Ensemble of LLMs to Attack AI-Generated Text Detection
The robustness of AI-content detection models against sophisticated adversarial strategies, such as paraphrasing or word switching, is a rising concern in natural language generation (NLG) applications. This study proposes ToBlend, a novel token-level ensemble text generation method to challenge the robustness of current AI-content detection approaches by utilizing multiple sets of candidate generative large language models (LLMs). By randomly sampling token(s) from candidate LLMs sets, we find ToBlend significantly drops the performance of most mainstream AI-content detection methods. We evaluate the text quality produced under different ToBlend settings based on annotations from experienced human experts. We proposed a fine-tuned Llama3.1 model to distinguish the ToBlend generated text more accurately. Our findings underscore our proposed text generation approach's great potential in deceiving and improving detection models. Our datasets, codes, and annotations are open-sourced.
comment: Submitted to ARR Oct-2024 Cycle
♻ ☆ ITINERA: Integrating Spatial Optimization with Large Language Models for Open-domain Urban Itinerary Planning
Citywalk, a recently popular form of urban travel, requires genuine personalization and understanding of fine-grained requests compared to traditional itinerary planning. In this paper, we introduce the novel task of Open-domain Urban Itinerary Planning (OUIP), which generates personalized urban itineraries from user requests in natural language. We then present ITINERA, an OUIP system that integrates spatial optimization with large language models to provide customized urban itineraries based on user needs. This involves decomposing user requests, selecting candidate points of interest (POIs), ordering the POIs based on cluster-aware spatial optimization, and generating the itinerary. Experiments on real-world datasets and the performance of the deployed system demonstrate our system's capacity to deliver personalized and spatially coherent itineraries compared to current solutions. Source codes of ITINERA are available at https://github.com/YihongT/ITINERA.
♻ ☆ CELL your Model: Contrastive Explanations for Large Language Models
The advent of black-box deep neural network classification models has sparked the need to explain their decisions. However, in the case of generative AI, such as large language models (LLMs), there is no class prediction to explain. Rather, one can ask why an LLM output a particular response to a given prompt. In this paper, we answer this question by proposing, to the best of our knowledge, the first contrastive explanation methods requiring simply black-box/query access. Our explanations suggest that an LLM outputs a reply to a given prompt because if the prompt was slightly modified, the LLM would have given a different response that is either less preferable or contradicts the original response. The key insight is that contrastive explanations simply require a scoring function that has meaning to the user and not necessarily a specific real valued quantity (viz. class label). We offer two algorithms for finding contrastive explanations: i) A myopic algorithm, which although effective in creating contrasts, requires many model calls and ii) A budgeted algorithm, our main algorithmic contribution, which intelligently creates contrasts adhering to a query budget, necessary for longer contexts. We show the efficacy of these methods on diverse natural language tasks such as open-text generation, automated red teaming, and explaining conversational degradation.
♻ ☆ Disentangling Singlish Discourse Particles with Task-Driven Representation
Singlish, or formally Colloquial Singapore English, is an English-based creole language originating from the SouthEast Asian country Singapore. The language contains influences from Sinitic languages such as Chinese dialects, Malay, Tamil and so forth. A fundamental task to understanding Singlish is to first understand the pragmatic functions of its discourse particles, upon which Singlish relies heavily to convey meaning. This work offers a preliminary effort to disentangle the Singlish discourse particles (lah, meh and hor) with task-driven representation learning. After disentanglement, we cluster these discourse particles to differentiate their pragmatic functions, and perform Singlish-to-English machine translation. Our work provides a computational method to understanding Singlish discourse particles, and opens avenues towards a deeper comprehension of the language and its usage.
♻ ☆ DOCE: Finding the Sweet Spot for Execution-Based Code Generation
Recently, a diverse set of decoding and reranking procedures have been shown effective for LLM-based code generation. However, a comprehensive framework that links and experimentally compares these methods is missing. We address this by proposing Decoding Objectives for Code Execution, a comprehensive framework that includes candidate generation, $n$-best reranking, minimum Bayes risk (MBR) decoding, and self-debugging as the core components. We then study the contributions of these components through execution-based evaluation metrics. Our findings highlight the importance of execution-based methods and the difference gap between execution-based and execution-free methods. Furthermore, we assess the impact of filtering based on trial unit tests, a simple and effective strategy that has been often overlooked in prior works. We also propose self-debugging on multiple candidates, obtaining state-of-the-art performance on reranking for code generation. We expect our framework to provide a solid guideline for future research on code generation.
comment: 10 pages (32 including appendix), 5 figures, 25 tables. Prompts are provided in the GitHub repository to avoid potential text overlap with other papers
♻ ☆ GTSinger: A Global Multi-Technique Singing Corpus with Realistic Music Scores for All Singing Tasks NeurIPS 2024
The scarcity of high-quality and multi-task singing datasets significantly hinders the development of diverse controllable and personalized singing tasks, as existing singing datasets suffer from low quality, limited diversity of languages and singers, absence of multi-technique information and realistic music scores, and poor task suitability. To tackle these problems, we present GTSinger, a large global, multi-technique, free-to-use, high-quality singing corpus with realistic music scores, designed for all singing tasks, along with its benchmarks. Particularly, (1) we collect 80.59 hours of high-quality singing voices, forming the largest recorded singing dataset; (2) 20 professional singers across nine widely spoken languages offer diverse timbres and styles; (3) we provide controlled comparison and phoneme-level annotations of six commonly used singing techniques, helping technique modeling and control; (4) GTSinger offers realistic music scores, assisting real-world musical composition; (5) singing voices are accompanied by manual phoneme-to-audio alignments, global style labels, and 16.16 hours of paired speech for various singing tasks. Moreover, to facilitate the use of GTSinger, we conduct four benchmark experiments: technique-controllable singing voice synthesis, technique recognition, style transfer, and speech-to-singing conversion. The corpus and demos can be found at http://gtsinger.github.io. We provide the dataset and the code for processing data and conducting benchmarks at https://huggingface.co/datasets/GTSinger/GTSinger and https://github.com/GTSinger/GTSinger.
comment: Accepted by NeurIPS 2024 (Spotlight)
♻ ☆ Reward-Robust RLHF in LLMs
As Large Language Models (LLMs) continue to progress toward more advanced forms of intelligence, Reinforcement Learning from Human Feedback (RLHF) is increasingly seen as a key pathway toward achieving Artificial General Intelligence (AGI). However, the reliance on reward-model-based (RM-based) alignment methods introduces significant challenges due to the inherent instability and imperfections of Reward Models (RMs), which can lead to critical issues such as reward hacking and misalignment with human intentions. In this paper, we introduce a reward-robust RLHF framework aimed at addressing these fundamental challenges, paving the way for more reliable and resilient learning in LLMs. Our approach introduces a novel optimization objective that carefully balances performance and robustness by incorporating Bayesian Reward Model Ensembles (BRME) to model the uncertainty set of reward functions. This allows the framework to integrate both nominal performance and minimum reward signals, ensuring more stable learning even with imperfect RMs. Empirical results demonstrate that our framework consistently outperforms baselines across diverse benchmarks, showing improved accuracy and long-term stability. We also provide a theoretical analysis, demonstrating that reward-robust RLHF approaches the stability of constant reward settings, which proves to be acceptable even in a stochastic-case analysis. Together, these contributions highlight the framework potential to enhance both the performance and stability of LLM alignment.
♻ ☆ Robust ASR Error Correction with Conservative Data Filtering EMNLP 2024
Error correction (EC) based on large language models is an emerging technology to enhance the performance of automatic speech recognition (ASR) systems. Generally, training data for EC are collected by automatically pairing a large set of ASR hypotheses (as sources) and their gold references (as targets). However, the quality of such pairs is not guaranteed, and we observed various types of noise which can make the EC models brittle, e.g. inducing overcorrection in out-of-domain (OOD) settings. In this work, we propose two fundamental criteria that EC training data should satisfy: namely, EC targets should (1) improve linguistic acceptability over sources and (2) be inferable from the available context (e.g. source phonemes). Through these criteria, we identify low-quality EC pairs and train the models not to make any correction in such cases, the process we refer to as conservative data filtering. In our experiments, we focus on Japanese ASR using a strong Conformer-CTC as the baseline and finetune Japanese LLMs for EC. Through our evaluation on a suite of 21 internal benchmarks, we demonstrate that our approach can significantly reduce overcorrection and improve both the accuracy and quality of ASR results in the challenging OOD settings.
comment: Accepted to EMNLP 2024 Industry Track
♻ ☆ Details Make a Difference: Object State-Sensitive Neurorobotic Task Planning ICANN24
The state of an object reflects its current status or condition and is important for a robot's task planning and manipulation. However, detecting an object's state and generating a state-sensitive plan for robots is challenging. Recently, pre-trained Large Language Models (LLMs) and Vision-Language Models (VLMs) have shown impressive capabilities in generating plans. However, to the best of our knowledge, there is hardly any investigation on whether LLMs or VLMs can also generate object state-sensitive plans. To study this, we introduce an Object State-Sensitive Agent (OSSA), a task-planning agent empowered by pre-trained neural networks. We propose two methods for OSSA: (i) a modular model consisting of a pre-trained vision processing module (dense captioning model, DCM) and a natural language processing model (LLM), and (ii) a monolithic model consisting only of a VLM. To quantitatively evaluate the performances of the two methods, we use tabletop scenarios where the task is to clear the table. We contribute a multimodal benchmark dataset that takes object states into consideration. Our results show that both methods can be used for object state-sensitive tasks, but the monolithic approach outperforms the modular approach. The code for OSSA is available at https://github.com/Xiao-wen-Sun/OSSA
comment: ICANN24, Switzerland
♻ ☆ ReadMe++: Benchmarking Multilingual Language Models for Multi-Domain Readability Assessment EMNLP 2024
We present a comprehensive evaluation of large language models for multilingual readability assessment. Existing evaluation resources lack domain and language diversity, limiting the ability for cross-domain and cross-lingual analyses. This paper introduces ReadMe++, a multilingual multi-domain dataset with human annotations of 9757 sentences in Arabic, English, French, Hindi, and Russian, collected from 112 different data sources. This benchmark will encourage research on developing robust multilingual readability assessment methods. Using ReadMe++, we benchmark multilingual and monolingual language models in the supervised, unsupervised, and few-shot prompting settings. The domain and language diversity in ReadMe++ enable us to test more effective few-shot prompting, and identify shortcomings in state-of-the-art unsupervised methods. Our experiments also reveal exciting results of superior domain generalization and enhanced cross-lingual transfer capabilities by models trained on ReadMe++. We will make our data publicly available and release a python package tool for multilingual sentence readability prediction using our trained models at: https://github.com/tareknaous/readme
comment: Accepted to EMNLP 2024 Main Conference
♻ ☆ From Explainable to Interpretable Deep Learning for Natural Language Processing in Healthcare: How Far from Reality?
Deep learning (DL) has substantially enhanced natural language processing (NLP) in healthcare research. However, the increasing complexity of DL-based NLP necessitates transparent model interpretability, or at least explainability, for reliable decision-making. This work presents a thorough scoping review of explainable and interpretable DL in healthcare NLP. The term "eXplainable and Interpretable Artificial Intelligence" (XIAI) is introduced to distinguish XAI from IAI. Different models are further categorized based on their functionality (model-, input-, output-based) and scope (local, global). Our analysis shows that attention mechanisms are the most prevalent emerging IAI technique. The use of IAI is growing, distinguishing it from XAI. The major challenges identified are that most XIAI does not explore "global" modelling processes, the lack of best practices, and the lack of systematic evaluation and benchmarks. One important opportunity is to use attention mechanisms to enhance multi-modal XIAI for personalized medicine. Additionally, combining DL with causal logic holds promise. Our discussion encourages the integration of XIAI in Large Language Models (LLMs) and domain-specific smaller models. In conclusion, XIAI adoption in healthcare requires dedicated in-house expertise. Collaboration with domain experts, end-users, and policymakers can lead to ready-to-use XIAI methods across NLP and medical tasks. While challenges exist, XIAI techniques offer a valuable foundation for interpretable NLP algorithms in healthcare.
comment: This paper has been accepted by Computational and Structural Biotechnology Journal
♻ ☆ Prompting Explicit and Implicit Knowledge for Multi-hop Question Answering Based on Human Reading Process COLING 2024
Pre-trained language models (PLMs) leverage chains-of-thought (CoT) to simulate human reasoning and inference processes, achieving proficient performance in multi-hop QA. However, a gap persists between PLMs' reasoning abilities and those of humans when tackling complex problems. Psychological studies suggest a vital connection between explicit information in passages and human prior knowledge during reading. Nevertheless, current research has given insufficient attention to linking input passages and PLMs' pre-training-based knowledge from the perspective of human cognition studies. In this study, we introduce a Prompting Explicit and Implicit knowledge (PEI) framework, which uses prompts to connect explicit and implicit knowledge, aligning with human reading process for multi-hop QA. We consider the input passages as explicit knowledge, employing them to elicit implicit knowledge through unified prompt reasoning. Furthermore, our model incorporates type-specific reasoning via prompts, a form of implicit knowledge. Experimental results show that PEI performs comparably to the state-of-the-art on HotpotQA. Ablation studies confirm the efficacy of our model in bridging and integrating explicit and implicit knowledge.
comment: This paper has been accepted at COLING 2024
♻ ☆ Five Years of COVID-19 Discourse on Instagram: A Labeled Instagram Dataset of Over Half a Million Posts for Multilingual Sentiment Analysis
The work presented in this paper makes three scientific contributions with a specific focus on mining and analysis of COVID-19-related posts on Instagram. First, it presents a multilingual dataset of 500,153 Instagram posts about COVID-19 published between January 2020 and September 2024. This dataset, available at https://dx.doi.org/10.21227/d46p-v480, contains Instagram posts in 161 different languages as well as 535,021 distinct hashtags. After the development of this dataset, multilingual sentiment analysis was performed, which involved classifying each post as positive, negative, or neutral. The results of sentiment analysis are presented as a separate attribute in this dataset. Second, it presents the results of performing sentiment analysis per year from 2020 to 2024. The findings revealed the trends in sentiment related to COVID-19 on Instagram since the beginning of the pandemic. For instance, between 2020 and 2024, the sentiment trends show a notable shift, with positive sentiment decreasing from 38.35% to 28.69%, while neutral sentiment rising from 44.19% to 58.34%. Finally, the paper also presents findings of language-specific sentiment analysis. This analysis highlighted similar and contrasting trends of sentiment across posts published in different languages on Instagram. For instance, out of all English posts, 49.68% were positive, 14.84% were negative, and 35.48% were neutral. In contrast, among Hindi posts, 4.40% were positive, 57.04% were negative, and 38.56% were neutral, reflecting distinct differences in the sentiment distribution between these two languages.
♻ ☆ Semantic Token Reweighting for Interpretable and Controllable Text Embeddings in CLIP EMNLP 2024
A text encoder within Vision-Language Models (VLMs) like CLIP plays a crucial role in translating textual input into an embedding space shared with images, thereby facilitating the interpretative analysis of vision tasks through natural language. Despite the varying significance of different textual elements within a sentence depending on the context, efforts to account for variation of importance in constructing text embeddings have been lacking. We propose a framework of Semantic Token Reweighting to build Interpretable text embeddings (SToRI), which incorporates controllability as well. SToRI refines the text encoding process in CLIP by differentially weighting semantic elements based on contextual importance, enabling finer control over emphasis responsive to data-driven insights and user preferences. The efficacy of SToRI is demonstrated through comprehensive experiments on few-shot image classification and image retrieval tailored to user preferences.
comment: Accepted at EMNLP 2024 Findings
♻ ☆ PromptDSI: Prompt-based Rehearsal-free Instance-wise Incremental Learning for Document Retrieval
Differentiable Search Index (DSI) utilizes Pre-trained Language Models (PLMs) for efficient document retrieval without relying on external indexes. However, DSI needs full re-training to handle updates in dynamic corpora, causing significant computational inefficiencies. We introduce PromptDSI, a prompt-based rehearsal-free approach for instance-wise incremental learning document retrieval. PromptDSI attaches prompts to the frozen PLM's encoder of DSI, leveraging its powerful representation to efficiently index new corpora while maintaining a balance between stability and plasticity. We eliminate the initial forward pass of prompt-based continual learning methods that doubles training and inference time. Moreover, we propose a topic-aware prompt pool that employs neural topic embeddings as fixed keys. This strategy ensures diverse and effective prompt usage, addressing the challenge of parameter underutilization caused by the collapse of the query-key matching mechanism. Our empirical evaluations demonstrate that BERT-based PromptDSI matches IncDSI in managing forgetting while improving new corpora performance by more than 4% Hits@10 on NQ320k and upto 3% MRR@10 on MS MARCO 300k.
comment: 20 pages
♻ ☆ The Comparative Trap: Pairwise Comparisons Amplifies Biased Preferences of LLM Evaluators
As large language models (LLMs) are increasingly used as evaluators for natural language generation tasks, ensuring unbiased assessments is essential. However, LLM evaluators often display biased preferences, such as favoring verbosity and authoritative tones. Our empirical analysis reveals that these biases are exacerbated in pairwise evaluation, where LLMs directly compare two outputs and easily prioritize superficial attributes. In contrast, pointwise evaluation, which assesses outputs independently, is less susceptible to such bias because each output is judged in isolation. To address the limitations of the pairwise evaluation, we introduce a novel evaluation method, PRePair, which integrates pointwise reasoning within a pairwise framework. PRePair effectively alleviates biased preference, improving performance on the adversarial benchmark (LLMBar) while outperforming pointwise evaluation on the standard benchmark (MT-Bench).
♻ ☆ Enhancing Data Privacy in Large Language Models through Private Association Editing
Large language models (LLMs) require a significant redesign in solutions to preserve privacy in data-intensive applications due to their text-generation capabilities. Indeed, LLMs tend to memorize and emit private information when maliciously prompted. In this paper, we introduce Private Association Editing (PAE) as a novel defense approach for private data leakage. PAE is designed to effectively remove Personally Identifiable Information (PII) without retraining the model. Experimental results demonstrate the effectiveness of PAE with respect to alternative baseline methods. We believe PAE will serve as a critical tool in the ongoing effort to protect data privacy in LLMs, encouraging the development of safer models for real-world applications.
♻ ☆ DIRAS: Efficient LLM Annotation of Document Relevance in Retrieval Augmented Generation
Retrieval Augmented Generation (RAG) is widely employed to ground responses to queries on domain-specific documents. But do RAG implementations leave out important information when answering queries that need an integrated analysis of information (e.g., Tell me good news in the stock market today.)? To address these concerns, RAG developers need to annotate information retrieval (IR) data for their domain of interest, which is challenging because (1) domain-specific queries usually need nuanced definitions of relevance beyond shallow semantic relevance; and (2) human or GPT-4 annotation is costly and cannot cover all (query, document) pairs (i.e., annotation selection bias), thus harming the effectiveness in evaluating IR recall. To address these challenges, we propose DIRAS (Domain-specific Information Retrieval Annotation with Scalability), a manual-annotation-free schema that fine-tunes open-sourced LLMs to consider nuanced relevance definition and annotate (partial) relevance labels with calibrated relevance scores. Extensive evaluation shows that DIRAS enables smaller (8B) LLMs to achieve GPT-4-level performance on annotating and ranking unseen (query, document) pairs, and is helpful for real-world RAG development. All code, LLM generations, and human annotations can be found in \url{https://github.com/EdisonNi-hku/DIRAS}.
♻ ☆ sPhinX: Sample Efficient Multilingual Instruction Fine-Tuning Through N-shot Guided Prompting
Despite the remarkable success of LLMs in English, there is a significant gap in performance in non-English languages. In order to address this, we introduce a novel recipe for creating a multilingual synthetic instruction tuning dataset, sPhinX, which is created by selectively translating instruction response pairs from English into 50 languages. We test the effectiveness of sPhinx by using it to fine-tune two state-of-the-art models, Mistral-7B and Phi-Small and then evaluating them across a comprehensive suite of multilingual benchmarks that test reasoning, question answering, reading comprehension and machine translation. Our results show that Mistral-7B and Phi-Small fine-tuned with sPhinX perform better on an average by 5%pt for both the models when compared to the base variants of these models. We also devise a strategy to incorporate N-shot examples in each fine-tuning sample which further boosts the performance of these models by 9%pt and 4%pt respectively respectively compared to vanilla fine-tuning. To show efficacy of our data curation approach, we also directly translate our original dataset to the target languages, and observe an increase of 7%pt and 4%pt on both the models respectively. sPhinX outperforms other multilingual instruction tuning datasets in both efficiency and diversity, reducing dataset creation costs. It also maintains strong performance on standard English LLM benchmarks, with minimal regression.
comment: 20 pages, 12 tables, 5 figures
♻ ☆ FLEX: Expert-level False-Less EXecution Metric for Reliable Text-to-SQL Benchmark
Text-to-SQL systems have become crucial for translating natural language into SQL queries in various industries, enabling non-technical users to perform complex data operations. The need for accurate evaluation methods has increased as these systems have grown more sophisticated. However, the Execution Accuracy (EX), the most prevalent evaluation metric, still shows many false positives and negatives. Thus, this paper introduces FLEX (False-Less EXecution), a novel approach to evaluating text-to-SQL systems using large language models (LLMs) to emulate human expert-level evaluation of SQL queries. Our metric improves agreement with human experts (from 62 to 87.04 in Cohen's kappa) with comprehensive context and sophisticated criteria. Our extensive experiments yield several key insights: (1) Models' performance increases by over 2.6 points on average, substantially affecting rankings on Spider and BIRD benchmarks; (2) The underestimation of models in EX primarily stems from annotation quality issues; and (3) Model performance on particularly challenging questions tends to be overestimated. This work contributes to a more accurate and nuanced evaluation of text-to-SQL systems, potentially reshaping our understanding of state-of-the-art performance in this field.
comment: preprint, under review
♻ ☆ Mixture of Experts Made Personalized: Federated Prompt Learning for Vision-Language Models
Prompt learning for pre-trained Vision-Language Models (VLMs) like CLIP has demonstrated potent applicability across diverse downstream tasks. This lightweight approach has quickly gained traction from federated learning (FL) researchers who seek to efficiently adapt VLMs to heterogeneous scenarios. However, current federated prompt learning methods are habitually restricted to the traditional FL paradigm, where the participating clients are generally only allowed to download a single globally aggregated model from the server. While justifiable for training full-sized models under federated settings, in this work, we argue that this paradigm is ill-suited for lightweight prompts. By facilitating the clients to download multiple pre-aggregated prompts as fixed non-local experts, we propose Personalized Federated Mixture of Adaptive Prompts (pFedMoAP), a novel FL framework that personalizes the prompt learning process through the lens of Mixture of Experts (MoE). pFedMoAP implements a local attention-based gating network that learns to generate enhanced text features for better alignment with local image data on the client, benefiting from both local and downloaded non-local adaptive prompt experts. The non-local experts are sparsely selected from a server-maintained pool, fostering collaborative learning across clients. To evaluate the proposed algorithm, we conduct extensive experiments across 9 datasets under various heterogeneous federated settings. The results show that pFedMoAP consistently outperforms the state-of-the-art alternatives, underscoring its efficacy in personalizing prompt learning for CLIP within the federated learning paradigm.
comment: 16 pages, 4 figures
♻ ☆ Efficient Few-shot Learning for Multi-label Classification of Scientific Documents with Many Classes SP 2024
Scientific document classification is a critical task and often involves many classes. However, collecting human-labeled data for many classes is expensive and usually leads to label-scarce scenarios. Moreover, recent work has shown that sentence embedding model fine-tuning for few-shot classification is efficient, robust, and effective. In this work, we propose FusionSent (Fusion-based Sentence Embedding Fine-tuning), an efficient and prompt-free approach for few-shot classification of scientific documents with many classes. FusionSent uses available training examples and their respective label texts to contrastively fine-tune two different sentence embedding models. Afterward, the parameters of both fine-tuned models are fused to combine the complementary knowledge from the separate fine-tuning steps into a single model. Finally, the resulting sentence embedding model is frozen to embed the training instances, which are then used as input features to train a classification head. Our experiments show that FusionSent significantly outperforms strong baselines by an average of $6.0$ $F_{1}$ points across multiple scientific document classification datasets. In addition, we introduce a new dataset for multi-label classification of scientific documents, which contains 203,961 scientific articles and 130 classes from the arXiv category taxonomy. Code and data are available at https://github.com/sebischair/FusionSent.
comment: Accepted to the 7th International Conference on Natural Language and Speech Processing (ICNLSP 2024)
♻ ☆ Benchmarking LLMs for Translating Classical Chinese Poetry:Evaluating Adequacy, Fluency, and Elegance
Large language models (LLMs) have shown remarkable performance in general translation tasks. However, the increasing demand for high-quality translations that are not only adequate but also fluent and elegant. To assess the extent to which current LLMs can meet these demands, we introduce a suitable benchmark for translating classical Chinese poetry into English. This task requires not only adequacy in translating culturally and historically significant content but also a strict adherence to linguistic fluency and poetic elegance. Our study reveals that existing LLMs fall short of this task. To address these issues, we propose RAT, a \textbf{R}etrieval-\textbf{A}ugmented machine \textbf{T}ranslation method that enhances the translation process by incorporating knowledge related to classical poetry. Additionally, we propose an automatic evaluation metric based on GPT-4, which better assesses translation quality in terms of adequacy, fluency, and elegance, overcoming the limitations of traditional metrics. Our dataset and code will be made available.
comment: Work in progress
♻ ☆ The Fellowship of the LLMs: Multi-Agent Workflows for Synthetic Preference Optimization Dataset Generation
This paper presents a novel methodology for generating synthetic Preference Optimization (PO) datasets using multi-agent workflows. We evaluate the effectiveness and potential of these workflows in automating and enhancing the dataset generation process. PO dataset generation requires two modules: (1) response evaluation, and (2) response generation. In the response evaluation module, the responses from Large Language Models (LLMs) are evaluated and ranked - a task typically carried out by human annotators that we automate using LLMs. We assess the response evaluation module in a 2 step process. In step 1, we assess LLMs as evaluators using three distinct prompting strategies. In step 2, we apply the winning prompting strategy to compare the performance of LLM-as-a-Judge, LLMs-as-a-Jury, and LLM Debate. Our evaluation shows that GPT-4o-as-a-Judge is more consistent across all datasets. For the response generation module, we use the identified LLM evaluator configuration and compare different configurations of the LLM Feedback Loop. We use the win rate to determine the best multi-agent configuration for generation. Experimenting with various configurations, we find that the LLM Feedback Loop, with Llama as the generator and Gemma as the reviewer, achieves a notable 71.8% and 73.8% win rate over single-agent Llama and Gemma, respectively. After identifying the best configurations for both modules, we generate our PO datasets using the above pipeline.
♻ ☆ Token-based Decision Criteria Are Suboptimal in In-context Learning
In-Context Learning (ICL) typically utilizes classification criteria from output probabilities of manually selected label tokens. However, we argue that such token-based classification criteria lead to suboptimal decision boundaries, despite delicate calibrations through translation and constrained rotation applied. To address this problem, we propose Hidden Calibration, which renounces token probabilities and uses the nearest centroid classifier on the LM's last hidden states. In detail, we assign the label of the nearest centroid previously estimated from a calibration set to the test sample as the predicted label. Our experiments on 6 models and 10 classification datasets indicate that Hidden Calibration consistently outperforms current token-based baselines by about 20%~50%, achieving a strong state-of-the-art in ICL. Our further analysis demonstrates that Hidden Calibration finds better classification criteria with less inter-class overlap, and LMs provide linearly separable intra-class clusters with the help of demonstrations, which supports Hidden Calibration and gives new insights into the principle of ICL.
comment: 24 pages, 15 figures, 13 tables
♻ ☆ UNDIAL: Self-Distillation with Adjusted Logits for Robust Unlearning in Large Language Models
Mitigating the retention of sensitive or private information in large language models is essential for enhancing privacy and safety. Existing unlearning methods, like Gradient Ascent and Negative Preference Optimization, directly tune models to remove unwanted information. However, these methods often become unstable because they fine-tune by maximizing cross-entropy loss, which is the opposite of traditional loss minimization in learning. This reversal creates instability, especially on larger datasets, as the model struggles to balance unlearning with maintaining language capacity, leading to over-unlearning. In this paper, we introduce UnDIAL (Unlearning via Self-Distillation on Adjusted Logits), a novel and robust unlearning method. Our approach leverages self-distillation to adjust logits and selectively reduce the influence of targeted tokens. This technique ensures smooth convergence and avoids catastrophic forgetting, even in challenging unlearning tasks with large datasets and sequential unlearning requests. Extensive experiments show that UnDIAL can achieve both robustness in unlearning and scalability while maintaining stable training dynamics and resilience to hyperparameter tuning.
♻ ☆ Beyond Instruction Following: Evaluating Inferential Rule Following of Large Language Models
Although Large Language Models (LLMs) have demonstrated strong ability, they are further supposed to be controlled and guided by in real-world scenarios to be safe, accurate, and intelligent. This demands the possession of capability of LLMs. However, no prior work has made a clear evaluation of the inferential rule-following capability of LLMs. Previous studies that try to evaluate the inferential rule-following capability of LLMs fail to distinguish the inferential rule-following scenarios from the instruction-following scenarios. Therefore, this paper first clarifies the concept of inferential rule-following and proposes a comprehensive benchmark, RuleBench, to evaluate a diversified range of inferential rule-following abilities. Our experimental results on a variety of LLMs show that they are still limited in following rules. Our analysis based on the evaluation results provides insights into the improvements for LLMs toward a better inferential rule-following intelligent agent. We further propose Inferential Rule-Following Tuning (IRFT). The experimental results show that through IRFT, LLMs can learn abstract rule-following abilities from purely synthetic data and then generalize to RuleBench. The data and code can be found at: https://anonymous.4open.science/r/llm-rule-following-B3E3/
♻ ☆ VersiCode: Towards Version-controllable Code Generation
Large Language Models (LLMs) have made tremendous strides in code generation, but existing research fails to account for the dynamic nature of software development, marked by frequent library updates. This gap significantly limits LLMs' deployment in realistic settings. In this paper, we propose two novel tasks aimed at bridging this gap: version-specific code completion (VSCC) and version-aware code migration (VACM). In conjunction, we introduce VersiCode, a comprehensive Python dataset specifically designed to evaluate LLMs on these two tasks, together with a novel evaluation metric, Critical Diff Check (CDC@1), which assesses code generation against evolving API requirements. We conduct an extensive evaluation on VersiCode, which reveals that version-controllable code generation is indeed a significant challenge, even for GPT-4o and other strong frontier models. We believe the novel tasks, dataset, and metric open up a new, important research direction that will further enhance LLMs' real-world applicability. The code and resources can be found at https://github.com/wutong8023/VersiCode.
♻ ☆ Revisiting Benchmark and Assessment: An Agent-based Exploratory Dynamic Evaluation Framework for LLMs
While various vertical domain large language models (LLMs) have been developed, the challenge of automatically evaluating their performance across different domains remains significant. Current benchmark-based evaluation methods exhibit rigid, aimless interactions and rely on pre-collected static datasets that are costly to build, inflexible across domains, and misaligned with practical user needs. To address this issue, we revisit the evaluation components and introduce two concepts: Benchmark+, which extends traditional question-answer benchmark into a more flexible "strategy-criterion" format; and Assessment+, which enhances the interaction process, enabling deeper exploration and supporting both quantitative metrics and qualitative insights. These concepts capture the nuanced behaviors of LLMs through richer, multi-turn interactions. We propose an agent-based evaluation framework called TestAgent, which implements these concepts through retrieval augmented generation and reinforcement learning. Experiments on tasks ranging from constructing vertical domain evaluation to activating existing benchmarks demonstrate the effectiveness of TestAgent across various scenarios. We believe this work offers an interesting perspective on automatic evaluation for LLMs.
♻ ☆ LoraMap: Harnessing the Power of LoRA Connections
Fact-checking techniques can mitigate hallucinations in Large Language Models (LLMs), a prominent issue in specialized domains. As parameter-efficient techniques such as Low-Rank Adaptation (LoRA) can overcome substantial computational overhead, some studies have explored the integration of multiple LoRAs. While previous studies focus on parallel integration, this paper investigates methods to establish connections among multiple LoRAs. We create three reasoning datasets tailored to fact-checking and fine-tune individual LoRAs, allowing them to view and reason from diverse perspectives. Then, we explore strategies for allocating these reasoning LoRAs and introduce LoraMap, an approach to map connections between them. The results of the fact-checking task demonstrate that the performance of LoraMap is superior to LoraHub, an existing method for integrating LoRAs. LoraMap also outperforms with significantly fewer trainable parameters than LoraConcat, which concatenates LoRAs and further fine-tunes them.
comment: 17 pages, 12 figures, 7 tables
♻ ☆ Mental Disorders Detection in the Era of Large Language Models
This paper compares the effectiveness of traditional machine learning methods, encoder-based models, and large language models (LLMs) on the task of detecting depression and anxiety. Five datasets were considered, each differing in format and the method used to define the target pathology class. We tested AutoML models based on linguistic features, several variations of encoder-based Transformers such as BERT, and state-of-the-art LLMs as pathology classification models. The results demonstrated that LLMs outperform traditional methods, particularly on noisy and small datasets where training examples vary significantly in text length and genre. However, psycholinguistic features and encoder-based models can achieve performance comparable to language models when trained on texts from individuals with clinically confirmed depression, highlighting their potential effectiveness in targeted clinical applications.
♻ ☆ MFC-Bench: Benchmarking Multimodal Fact-Checking with Large Vision-Language Models
Large vision-language models (LVLMs) have significantly improved multimodal reasoning tasks, such as visual question answering and image captioning. These models embed multimodal facts within their parameters, rather than relying on external knowledge bases to store factual information explicitly. However, the content discerned by LVLMs may deviate from actual facts due to inherent bias or incorrect inference. To address this issue, we introduce MFC-Bench, a rigorous and comprehensive benchmark designed to evaluate the factual accuracy of LVLMs across three stages of verdict prediction for MFC: Manipulation, Out-of-Context, and Veracity Classification. Through our evaluation on MFC-Bench, we benchmarked a dozen diverse and representative LVLMs, uncovering that current models still fall short in multimodal fact-checking and demonstrate insensitivity to various forms of manipulated content. We hope that MFC-Bench could raise attention to the trustworthy AI potentially assisted by LVLMs in the future. The MFC-Bench and accompanying resources are publicly accessible at https://github.com/wskbest/MFC-Bench, contributing to ongoing research in the multimodal fact-checking field.
comment: 28 pages, 9 figures
♻ ☆ A corpus-based investigation of pitch contours of monosyllabic words in conversational Taiwan Mandarin
In Mandarin, the tonal contours of monosyllabic words produced in isolation or in careful speech are characterized by four lexical tones: a high-level tone (T1), a rising tone (T2), a dipping tone (T3) and a falling tone (T4). However, in spontaneous speech, the actual tonal realization of monosyllabic words can deviate significantly from these canonical tones due to intra-syllabic co-articulation and inter-syllabic co-articulation with adjacent tones. In addition, Chuang et al. (2024) recently reported that the tonal contours of disyllabic Mandarin words with T2-T4 tone pattern are co-determined by their meanings. Following up on their research, we present a corpus-based investigation of how the pitch contours of monosyllabic words are realized in spontaneous conversational Mandarin, focusing on the effects of contextual predictors on the one hand, and the way in words' meanings co-determine pitch contours on the other hand. We analyze the F0 contours of 3824 tokens of 63 different word types in a spontaneous Taiwan Mandarin corpus, using the generalized additive (mixed) model to decompose a given observed pitch contour into a set of component pitch contours. We show that the tonal context substantially modify a word's canonical tone. Once the effect of tonal context is controlled for, T2 and T3 emerge as low flat tones, contrasting with T1 as a high tone, and with T4 as a high-to-mid falling tone. The neutral tone (T0), which in standard descriptions, is realized based on the preceding tone, emerges as a low tone in its own right, modified by the other predictors in the same way as the standard tones T1, T2, T3, and T4. We also show that word, and even more so, word sense, co-determine words' F0 contours. Analyses of variable importance using random forests further supported the substantial effect of tonal context and an effect of word sense.
♻ ☆ Reverse Stable Diffusion: What prompt was used to generate this image?
Text-to-image diffusion models have recently attracted the interest of many researchers, and inverting the diffusion process can play an important role in better understanding the generative process and how to engineer prompts in order to obtain the desired images. To this end, we study the task of predicting the prompt embedding given an image generated by a generative diffusion model. We consider a series of white-box and black-box models (with and without access to the weights of the diffusion network) to deal with the proposed task. We propose a novel learning framework comprising a joint prompt regression and multi-label vocabulary classification objective that generates improved prompts. To further improve our method, we employ a curriculum learning procedure that promotes the learning of image-prompt pairs with lower labeling noise (i.e. that are better aligned). We conduct experiments on the DiffusionDB data set, predicting text prompts from images generated by Stable Diffusion. In addition, we make an interesting discovery: training a diffusion model on the prompt generation task can make the model generate images that are much better aligned with the input prompts, when the model is directly reused for text-to-image generation. Our code is publicly available for download at https://github.com/CroitoruAlin/Reverse-Stable-Diffusion.
comment: Accepted for publication in Computer Vision and Image Understanding
♻ ☆ Large Language Models in the Clinic: A Comprehensive Benchmark EMNLP 2024
The adoption of large language models (LLMs) to assist clinicians has attracted remarkable attention. Existing works mainly adopt the close-ended question-answering (QA) task with answer options for evaluation. However, many clinical decisions involve answering open-ended questions without pre-set options. To better understand LLMs in the clinic, we construct a benchmark ClinicBench. We first collect eleven existing datasets covering diverse clinical language generation, understanding, and reasoning tasks. Furthermore, we construct six novel datasets and clinical tasks that are complex but common in real-world practice, e.g., open-ended decision-making, long document processing, and emerging drug analysis. We conduct an extensive evaluation of twenty-two LLMs under both zero-shot and few-shot settings. Finally, we invite medical experts to evaluate the clinical usefulness of LLMs. The benchmark data is available at https://github.com/AI-in-Health/ClinicBench.
comment: Accepted at EMNLP 2024 Main Conference
♻ ☆ CLongEval: A Chinese Benchmark for Evaluating Long-Context Large Language Models EMNLP 2024
Developing Large Language Models (LLMs) with robust long-context capabilities has been the recent research focus, resulting in the emergence of long-context LLMs proficient in Chinese. However, the evaluation of these models remains underdeveloped due to a lack of benchmarks. To address this gap, we present CLongEval, a comprehensive Chinese benchmark for evaluating long-context LLMs. CLongEval is characterized by three key features: (1) Sufficient data volume, comprising 7 distinct tasks and 7,267 examples; (2) Broad applicability, accommodating to models with context windows size from 1K to 100K; (3) High quality, with over 2,000 manually annotated question-answer pairs in addition to the automatically constructed labels. With CLongEval, we undertake a comprehensive assessment of 6 open-source long-context LLMs and 2 leading commercial counterparts that feature both long-context abilities and proficiency in Chinese. We also provide in-depth analysis based on the empirical results, trying to shed light on the critical capabilities that present challenges in long-context settings. The dataset, evaluation scripts, and model outputs are released.
comment: Findings of EMNLP 2024
♻ ☆ TaCo: Targeted Concept Erasure Prevents Non-Linear Classifiers From Detecting Protected Attributes
Ensuring fairness in NLP models is crucial, as they often encode sensitive attributes like gender and ethnicity, leading to biased outcomes. Current concept erasure methods attempt to mitigate this by modifying final latent representations to remove sensitive information without retraining the entire model. However, these methods typically rely on linear classifiers, which leave models vulnerable to non-linear adversaries capable of recovering sensitive information. We introduce Targeted Concept Erasure (TaCo), a novel approach that removes sensitive information from final latent representations, ensuring fairness even against non-linear classifiers. Our experiments show that TaCo outperforms state-of-the-art methods, achieving greater reductions in the prediction accuracy of sensitive attributes by non-linear classifier while preserving overall task performance. Code is available on https://github.com/fanny-jourdan/TaCo.
♻ ☆ Causal Inference with Large Language Model: A Survey
Causal inference has been a pivotal challenge across diverse domains such as medicine and economics, demanding a complicated integration of human knowledge, mathematical reasoning, and data mining capabilities. Recent advancements in natural language processing (NLP), particularly with the advent of large language models (LLMs), have introduced promising opportunities for traditional causal inference tasks. This paper reviews recent progress in applying LLMs to causal inference, encompassing various tasks spanning different levels of causation. We summarize the main causal problems and approaches, and present a comparison of their evaluation results in different causal scenarios. Furthermore, we discuss key findings and outline directions for future research, underscoring the potential implications of integrating LLMs in advancing causal inference methodologies.
comment: 12 pages, 2 figures, 3 tables
♻ ☆ Explore, Select, Derive, and Recall: Augmenting LLM with Human-like Memory for Mobile Task Automation
The advent of large language models (LLMs) has opened up new opportunities in the field of mobile task automation. Their superior language understanding and reasoning capabilities allow users to automate complex and repetitive tasks. However, due to the inherent unreliability and high operational cost of LLMs, their practical applicability is quite limited. To address these issues, this paper introduces MobileGPT, an innovative LLM-based mobile task automator equipped with a human-like app memory. MobileGPT emulates the cognitive process of humans interacting with a mobile app -- explore, select, derive, and recall. This approach allows for a more precise and efficient learning of a task's procedure by breaking it down into smaller, modular sub-tasks that can be re-used, re-arranged, and adapted for various objectives. We implement MobileGPT using online LLMs services (GPT-3.5 and GPT-4) and evaluate its performance on a dataset of 185 tasks across 18 mobile apps. The results indicate that MobileGPT can automate and learn new tasks with 82.7% accuracy, and is able to adapt them to different contexts with near perfect (98.75%) accuracy while reducing both latency and cost by 62.5% and 68.8%, respectively, compared to the GPT-4 powered baseline.
♻ ☆ Reconsidering Degeneration of Token Embeddings with Definitions for Encoder-based Pre-trained Language Models
Learning token embeddings based on token co-occurrence statistics has proven effective for both pre-training and fine-tuning in natural language processing. However, recent studies have pointed out that the distribution of learned embeddings degenerates into anisotropy (i.e., non-uniform distribution), and even pre-trained language models (PLMs) suffer from a loss of semantics-related information in embeddings for low-frequency tokens. This study first analyzes the fine-tuning dynamics of encoder-based PLMs and demonstrates their robustness against degeneration. On the basis of this analysis, we propose DefinitionEMB, a method that utilizes definitions to re-construct isotropically distributed and semantics-related token embeddings for encoder-based PLMs while maintaining original robustness during fine-tuning. Our experiments demonstrate the effectiveness of leveraging definitions from Wiktionary to re-construct such embeddings for two encoder-based PLMs: RoBERTa-base and BART-large. Furthermore, the re-constructed embeddings for low-frequency tokens improve the performance of these models across various GLUE and four text summarization datasets.
♻ ☆ I Want to Break Free! Persuasion and Anti-Social Behavior of LLMs in Multi-Agent Settings with Social Hierarchy
As Large Language Model (LLM)-based agents become increasingly autonomous and will more freely interact with each other, studying interactions between them becomes crucial to anticipate emergent phenomena and potential risks. Drawing inspiration from the widely popular Stanford Prison Experiment, we contribute to this line of research by studying interaction patterns of LLM agents in a context characterized by strict social hierarchy. We do so by specifically studying two types of phenomena: persuasion and anti-social behavior in simulated scenarios involving a guard and a prisoner agent who seeks to achieve a specific goal (i.e., obtaining additional yard time or escape from prison). Leveraging 200 experimental scenarios for a total of 2,000 machine-machine conversations across five different popular LLMs, we provide a set of noteworthy findings. We first document how some models consistently fail in carrying out a conversation in our multi-agent setup where power dynamics are at play. Then, for the models that were able to engage in successful interactions, we empirically show how the goal that an agent is set to achieve impacts primarily its persuasiveness, while having a negligible effect with respect to the agent's anti-social behavior. Third, we highlight how agents' personas, and particularly the guard's personality, drive both the likelihood of successful persuasion from the prisoner and the emergence of anti-social behaviors. Fourth, we show that even without explicitly prompting for specific personalities, anti-social behavior emerges by simply assigning agents' roles. These results bear implications for the development of interactive LLM agents as well as the debate on their societal impact.
♻ ☆ How Do Humans Write Code? Large Models Do It the Same Way Too
Program-of-Thought (PoT) replaces natural language-based Chain-of-Thought (CoT) as the most popular method in Large Language Models (LLMs) mathematical reasoning tasks by utilizing external tool calls to circumvent computational errors. However, our evaluation of the GPT-4 and Llama series reveals that using PoT introduces more reasoning errors, such as incorrect formulas or flawed logic, compared to CoT. To address this issue, we propose Human-Think Language (HTL), which leverages a suite of strategies that help integrate PoT and CoT, encompassing: (1) a new generation paradigm that uses full CoT reasoning to control code generation. (2) Focus Attention, that directs model attention to the CoT reasoning during PoT to generate more logical code. (3) reinforcement learning that utilizes the accuracy of both CoT and PoT responses as rewards to prevent repetitive reasoning steps in LLMs when solving difficult math problems. Our method achieves an average improvement of 6.5% on the Llama-Base model and 4.3% on the Mistral-Base model across 8 mathematical calculation datasets. It also shows significant effectiveness on five out-of-domain datasets by controlling the model's information flow, exhibiting strong transferability. Additionally, HTL shows the most significant improvement in non-mathematical natural language inference task, contributing to a unified reasoning task framework
♻ ☆ Measuring and Benchmarking Large Language Models' Capabilities to Generate Persuasive Language
We are exposed to much information trying to influence us, such as teaser messages, debates, politically framed news, and propaganda - all of which use persuasive language. With the recent interest in Large Language Models (LLMs), we study the ability of LLMs to produce persuasive text. As opposed to prior work which focuses on particular domains or types of persuasion, we conduct a general study across various domains to measure and benchmark to what degree LLMs produce persuasive language - both when explicitly instructed to rewrite text to be more or less persuasive and when only instructed to paraphrase. We construct the new dataset Persuasive-Pairs of pairs of a short text and its rewrite by an LLM to amplify or diminish persuasive language. We multi-annotate the pairs on a relative scale for persuasive language: a valuable resource in itself, and for training a regression model to score and benchmark persuasive language, including for new LLMs across domains. In our analysis, we find that different 'personas' in LLaMA3's system prompt change persuasive language substantially, even when only instructed to paraphrase.
♻ ☆ Deciphering Cross-Modal Alignment in Large Vision-Language Models with Modality Integration Rate
We present the Modality Integration Rate (MIR), an effective, robust, and generalized metric to indicate the multi-modal pre-training quality of Large Vision Language Models (LVLMs). Large-scale pre-training plays a critical role in building capable LVLMs, while evaluating its training quality without the costly supervised fine-tuning stage is under-explored. Loss, perplexity, and in-context evaluation results are commonly used pre-training metrics for Large Language Models (LLMs), while we observed that these metrics are less indicative when aligning a well-trained LLM with a new modality. Due to the lack of proper metrics, the research of LVLMs in the critical pre-training stage is hindered greatly, including the training data choice, efficient module design, etc. In this paper, we propose evaluating the pre-training quality from the inter-modal distribution distance perspective and present MIR, the Modality Integration Rate, which is 1) \textbf{Effective} to represent the pre-training quality and show a positive relation with the benchmark performance after supervised fine-tuning. 2) \textbf{Robust} toward different training/evaluation data. 3) \textbf{Generalize} across training configurations and architecture choices. We conduct a series of pre-training experiments to explore the effectiveness of MIR and observe satisfactory results that MIR is indicative about training data selection, training strategy schedule, and model architecture design to get better pre-training results. We hope MIR could be a helpful metric for building capable LVLMs and inspire the following research about modality alignment in different areas. Our code is at: https://github.com/shikiw/Modality-Integration-Rate.
comment: Project page: https://github.com/shikiw/Modality-Integration-Rate
♻ ☆ Exploring Changes in Nation Perception with Nationality-Assigned Personas in LLMs
Persona assignment has become a common strategy for customizing LLM use to particular tasks and contexts. In this study, we explore how evaluation of different nations change when LLMs are assigned specific nationality personas. We assign 193 different nationality personas (e.g., an American person) to four LLMs and examine how the LLM evaluations (or ''perceptions'')of countries change. We find that all LLM-persona combinations tend to favor Western European nations, though nation-personas push LLM behaviors to focus more on and treat the nation-persona's own region more favorably. Eastern European, Latin American, and African nations are treated more negatively by different nationality personas. We additionally find that evaluations by nation-persona LLMs of other nations correlate with human survey responses but fail to match the values closely. Our study provides insight into how biases and stereotypes are realized within LLMs when adopting different national personas. In line with the ''Blueprint for an AI Bill of Rights'', our findings underscore the critical need for developing mechanisms to ensure that LLM outputs promote fairness and avoid over-generalization.
comment: Pre-print, Under review
♻ ☆ Adaptation Odyssey in LLMs: Why Does Additional Pretraining Sometimes Fail to Improve? EMNLP 2024
In the last decade, the generalization and adaptation abilities of deep learning models were typically evaluated on fixed training and test distributions. Contrary to traditional deep learning, large language models (LLMs) are (i) even more overparameterized, (ii) trained on unlabeled text corpora curated from the Internet with minimal human intervention, and (iii) trained in an online fashion. These stark contrasts prevent researchers from transferring lessons learned on model generalization and adaptation in deep learning contexts to LLMs. To this end, our short paper introduces empirical observations that aim to shed light on further training of already pretrained language models. Specifically, we demonstrate that training a model on a text domain could degrade its perplexity on the test portion of the same domain. We observe with our subsequent analysis that the performance degradation is positively correlated with the similarity between the additional and the original pretraining dataset of the LLM. Our further token-level perplexity observations reveals that the perplexity degradation is due to a handful of tokens that are not informative about the domain. We hope these findings will guide us in determining when to adapt a model vs when to rely on its foundational capabilities.
comment: Accepted to EMNLP 2024 Main Conference
♻ ☆ Towards Infinite-Long Prefix in Transformer
Prompting and context-based fine-tuning methods, which we call Prefix Learning, have been proposed to enhance the performance of language models on various downstream tasks. They are empirically efficient and effective, matching the performance of full parameter fine-tuning, but the theoretical understandings are limited. In this paper, we aim to address this limitation by studying their ability from the perspective of prefix length. In particular, we provide a convergence guarantee for training an ultra-long prefix in a stylized setting using the Neural Tangent Kernel (NTK) framework. Based on this strong theoretical guarantee, we design and implement an algorithm that only needs to introduce and fine-tune a few extra trainable parameters instead of an infinite-long prefix in each layer of a transformer, and can approximate the prefix attention to a guaranteed polynomial-small error. Preliminary experimental results on vision, natural language, and math data show that our method achieves superior or competitive performance compared to existing methods like full parameters fine-tuning, P-Tuning V2, and LoRA. This demonstrates our method is promising for parameter-efficient fine-tuning. Our code can be found at \url{https://github.com/ChristianYang37/chiwun/tree/main/src/NTK-Attention}.
♻ ☆ SeRTS: Self-Rewarding Tree Search for Biomedical Retrieval-Augmented Generation EMNLP 2024
Large Language Models (LLMs) have shown great potential in the biomedical domain with the advancement of retrieval-augmented generation (RAG). However, existing retrieval-augmented approaches face challenges in addressing diverse queries and documents, particularly for medical knowledge queries, resulting in sub-optimal performance. To address these limitations, we propose a novel plug-and-play LLM-based retrieval method called Self-Rewarding Tree Search (SeRTS) based on Monte Carlo Tree Search (MCTS) and a self-rewarding paradigm. By combining the reasoning capabilities of LLMs with the effectiveness of tree search, SeRTS boosts the zero-shot performance of retrieving high-quality and informative results for RAG. We further enhance retrieval performance by fine-tuning LLMs with Proximal Policy Optimization (PPO) objectives using the trajectories collected by SeRTS as feedback. Controlled experiments using the BioASQ-QA dataset with GPT-3.5-Turbo and LLama2-7b demonstrate that our method significantly improves the performance of the BM25 retriever and surpasses the strong baseline of self-reflection in both efficiency and scalability. Moreover, SeRTS generates higher-quality feedback for PPO training than self-reflection. Our proposed method effectively adapts LLMs to document retrieval tasks, enhancing their ability to retrieve highly relevant documents for RAG in the context of medical knowledge queries. This work presents a significant step forward in leveraging LLMs for accurate and comprehensive biomedical question answering.
comment: This work has been accepted by EMNLP 2024
♻ ☆ Enhancing Assamese NLP Capabilities: Introducing a Centralized Dataset Repository
This paper introduces a centralized, open-source dataset repository designed to advance NLP and NMT for Assamese, a low-resource language. The repository, available at GitHub, supports various tasks like sentiment analysis, named entity recognition, and machine translation by providing both pre-training and fine-tuning corpora. We review existing datasets, highlighting the need for standardized resources in Assamese NLP, and discuss potential applications in AI-driven research, such as LLMs, OCR, and chatbots. While promising, challenges like data scarcity and linguistic diversity remain. The repository aims to foster collaboration and innovation, promoting Assamese language research in the digital age.
comment: 6 pages, 1 table, 1 figure
♻ ☆ MERLIN: Multimodal Embedding Refinement via LLM-based Iterative Navigation for Text-Video Retrieval-Rerank Pipeline EMNLP 2024
The rapid expansion of multimedia content has made accurately retrieving relevant videos from large collections increasingly challenging. Recent advancements in text-video retrieval have focused on cross-modal interactions, large-scale foundation model training, and probabilistic modeling, yet often neglect the crucial user perspective, leading to discrepancies between user queries and the content retrieved. To address this, we introduce MERLIN (Multimodal Embedding Refinement via LLM-based Iterative Navigation), a novel, training-free pipeline that leverages Large Language Models (LLMs) for iterative feedback learning. MERLIN refines query embeddings from a user perspective, enhancing alignment between queries and video content through a dynamic question answering process. Experimental results on datasets like MSR-VTT, MSVD, and ActivityNet demonstrate that MERLIN substantially improves Recall@1, outperforming existing systems and confirming the benefits of integrating LLMs into multimodal retrieval systems for more responsive and context-aware multimedia retrieval.
comment: EMNLP 2024 Industry Track Accepted (Camera-Ready Version)
♻ ☆ PAD: Personalized Alignment at Decoding-Time
Aligning with personalized preferences, which vary significantly across cultural, educational, and political differences, poses a significant challenge due to the computational costs and data demands of traditional alignment methods. In response, this paper presents Personalized Alignment at Decoding-time (PAD), a novel framework designed to align LLM outputs with diverse personalized preferences during the inference phase, eliminating the need for additional training. By introducing a unique personalized reward modeling strategy, this framework decouples the text generation process from personalized preferences, facilitating the generation of generalizable token-level personalized rewards. The PAD algorithm leverages these rewards to guide the decoding process, dynamically tailoring the base model's predictions to personalized preferences. Extensive experimental results demonstrate that PAD not only outperforms existing training-based alignment methods in terms of aligning with diverse preferences but also shows significant generalizability to preferences unseen during training and scalability across different base models. This work advances the capability of LLMs to meet user needs in real-time applications, presenting a substantial step forward in personalized LLM alignment.
comment: This paper presents Personalized Alignment at Decoding-time (PAD), a novel framework designed to align LLM outputs with diverse personalized preferences during the inference phase
♻ ☆ $α$-DPO: Adaptive Reward Margin is What Direct Preference Optimization Needs
Aligning large language models (LLMs) with human values and intentions is crucial for their utility, honesty, and safety. Reinforcement learning from human feedback (RLHF) is a popular approach to achieve this alignment, but it faces challenges in computational efficiency and training stability. Recent methods like Direct Preference Optimization (DPO) and Simple Preference Optimization (SimPO) have proposed offline alternatives to RLHF, simplifying the process by reparameterizing the reward function. However, DPO depends on a potentially suboptimal reference model, and SimPO's assumption of a fixed target reward margin may lead to suboptimal decisions in diverse data settings. In this work, we propose $\alpha$-DPO, an adaptive preference optimization algorithm designed to address these limitations by introducing a dynamic reward margin. Specifically, $\alpha$-DPO employs an adaptive preference distribution, balancing the policy model and the reference model to achieve personalized reward margins. We provide theoretical guarantees for $\alpha$-DPO, demonstrating its effectiveness as a surrogate optimization objective and its ability to balance alignment and diversity through KL divergence control. Empirical evaluations on AlpacaEval 2 and Arena-Hard show that $\alpha$-DPO consistently outperforms DPO and SimPO across various model settings, establishing it as a robust approach for fine-tuning LLMs. Our method achieves significant improvements in win rates, highlighting its potential as a powerful tool for LLM alignment. The code is available at https://github.com/junkangwu/alpha-DPO
♻ ☆ Instruction Tuning for Large Language Models: A Survey
This paper surveys research works in the quickly advancing field of instruction tuning (IT), a crucial technique to enhance the capabilities and controllability of large language models (LLMs). Instruction tuning refers to the process of further training LLMs on a dataset consisting of \textsc{(instruction, output)} pairs in a supervised fashion, which bridges the gap between the next-word prediction objective of LLMs and the users' objective of having LLMs adhere to human instructions. In this work, we make a systematic review of the literature, including the general methodology of IT, the construction of IT datasets, the training of IT models, and applications to different modalities, domains and applications, along with an analysis on aspects that influence the outcome of IT (e.g., generation of instruction outputs, size of the instruction dataset, etc). We also review the potential pitfalls of IT along with criticism against it, along with efforts pointing out current deficiencies of existing strategies and suggest some avenues for fruitful research. Project page: github.com/xiaoya-li/Instruction-Tuning-Survey
comment: V3; Last update: Oct 16, 2024
♻ ☆ Unveiling the Power of Source: Source-based Minimum Bayes Risk Decoding for Neural Machine Translation
Maximum a posteriori decoding, a commonly used method for neural machine translation (NMT), aims to maximize the estimated posterior probability. However, high estimated probability does not always lead to high translation quality. Minimum Bayes Risk (MBR) decoding (\citealp{kumar2004minimum}) offers an alternative by seeking hypotheses with the highest expected utility. In this paper, we show that Quality Estimation (QE) reranking (\citealp{fernandes-etal-2022-quality}), which uses a QE model as a reranker, can be viewed as a variant of MBR. Inspired by this, we propose source-based MBR (sMBR) decoding, a novel approach that utilizes synthetic sources (generated via back-translation or paraphrasing) as ``support hypotheses'' and a reference-free quality estimation metric as the utility function, marking the first work to solely use sources in MBR decoding. Experiments show that sMBR outperforms QE reranking and the standard MBR decoding. Our findings suggest that sMBR is a promising approach for NMT decoding.
♻ ☆ Seeker: Enhancing Exception Handling in Code with LLM-based Multi-Agent Approach ICLR 2025
In real world software development, improper or missing exception handling can severely impact the robustness and reliability of code. Exception handling mechanisms require developers to detect, capture, and manage exceptions according to high standards, but many developers struggle with these tasks, leading to fragile code. This problem is particularly evident in open source projects and impacts the overall quality of the software ecosystem. To address this challenge, we explore the use of large language models (LLMs) to improve exception handling in code. Through extensive analysis, we identify three key issues: Insensitive Detection of Fragile Code, Inaccurate Capture of Exception Types, and Distorted Handling Solutions. These problems are widespread across real world repositories, suggesting that robust exception handling practices are often overlooked or mishandled. In response, we propose Seeker, a multi agent framework inspired by expert developer strategies for exception handling. Seeker uses agents: Scanner, Detector, Predator, Ranker, and Handler to assist LLMs in detecting, capturing, and resolving exceptions more effectively. Our work is the first systematic study on leveraging LLMs to enhance exception handling practices, providing valuable insights for future improvements in code reliability.
comment: 26 pages, 7 figures. Submitted ICLR 2025
♻ ☆ Investigating the Transferability of Code Repair for Low-Resource Programming Languages
Large language models (LLMs) have shown remarkable performance on code generation tasks. A recent use case is iterative code repair, where an LLM fixes an incorrect program by rationalizing about errors and generating new code. Recent works augment the code repair process by integrating modern techniques such as chain-of-thought reasoning or distillation, but only study their benefits on high-resource languages like Python, and ignore low-resource languages like Perl. To address this gap of knowledge, we investigate the benefits of distilling code repair for both high and low resource languages to determine if the techniques that are effective in a high resource setting are also applicable in a low resource setting. Our evaluation shows that distilling the ability to repair code has language dependent benefits. To explain this behavior, we perform a further analysis and find that contrary to preexisting beliefs, the correlation between reasoning ability and code correction ability is weak. We hypothesize this weak correlation is magnified in low-resource settings where base models lack deep knowledge of a programming language, leading to wavering benefits of code repair.
♻ ☆ The Accuracy Paradox in RLHF: When Better Reward Models Don't Yield Better Language Models EMNLP 2024
Reinforcement Learning from Human Feedback significantly enhances Natural Language Processing by aligning language models with human expectations. A critical factor in this alignment is the strength of reward models used during training. This study explores whether stronger reward models invariably lead to better language models. In this paper, through experiments on relevance, factuality, and completeness tasks using the QA-FEEDBACK dataset and reward models based on Longformer, we uncover a surprising paradox: language models trained with moderately accurate reward models outperform those guided by highly accurate ones. This challenges the widely held belief that stronger reward models always lead to better language models, and opens up new avenues for future research into the key factors driving model performance and how to choose the most suitable reward models. Code and additional details are available at https://github.com/EIT-NLP/AccuracyParadox-RLHF.
comment: 10 pages, 27 figures (including 18 in the appendix), submitted to EMNLP 2024
♻ ☆ Beyond Demographics: Aligning Role-playing LLM-based Agents Using Human Belief Networks
Creating human-like large language model (LLM) agents is crucial for faithful social simulation. Having LLMs role-play based on demographic information sometimes improves human likeness but often does not. This study assessed whether LLM alignment with human behavior can be improved by integrating information from empirically-derived human belief networks. Using data from a human survey, we estimated a belief network encompassing 64 topics loading on nine non-overlapping latent factors. We then seeded LLM-based agents with an opinion on one topic, and assessed the alignment of its expressed opinions on remaining test topics with corresponding human data. Role-playing based on demographic information alone did not align LLM and human opinions, but seeding the agent with a single belief greatly improved alignment for topics related in the belief network, and not for topics outside the network. These results suggest a novel path for human-LLM belief alignment in work seeking to simulate and understand patterns of belief distributions in society.
♻ ☆ Breaking Language Barriers in Multilingual Mathematical Reasoning: Insights and Observations
Existing research predominantly focuses on developing powerful language learning models (LLMs) for mathematical reasoning within monolingual languages, with few explorations in preserving efficacy in a multilingual context. To bridge this gap, this paper pioneers exploring and training powerful Multilingual Math Reasoning (xMR) LLMs. Firstly, by utilizing translation, we construct the first multilingual math reasoning instruction dataset, MGSM8KInstruct, encompassing ten distinct languages, thus addressing the issue of training data scarcity in xMR tasks. Based on the collected dataset, we propose different training strategies to build powerful xMR LLMs, named MathOctopus, notably outperform conventional open-source LLMs and exhibit superiority over ChatGPT in few-shot scenarios. Notably, MathOctopus-13B reaches 47.6% accuracy which exceeds ChatGPT 46.3% on MGSM testset. Beyond remarkable results, we unearth several pivotal observations and insights from extensive experiments: (1) When extending the rejection sampling strategy to the multilingual context, it proves effective for model performances, albeit limited. (2) Employing parallel corpora for math Supervised Fine-Tuning (SFT) across multiple languages not only significantly enhances model performance multilingually but also elevates their monolingual performance. This indicates that crafting multilingual corpora can be regarded as a vital strategy for enhancing model performance in a specific language, especially in mathematical reasoning tasks. For instance, MathOctopus-7B improves its counterparts that trained on English from 42.2% to 50.8% on GSM8K testset. Codes are available at https://github.com/microsoft/MathOctopus.
comment: Work in Progress
♻ ☆ Explainable Natural Language Processing for Corporate Sustainability Analysis
Sustainability commonly refers to entities, such as individuals, companies, and institutions, having a non-detrimental (or even positive) impact on the environment, society, and the economy. With sustainability becoming a synonym of acceptable and legitimate behaviour, it is being increasingly demanded and regulated. Several frameworks and standards have been proposed to measure the sustainability impact of corporations, including United Nations' sustainable development goals and the recently introduced global sustainability reporting framework, amongst others. However, the concept of corporate sustainability is complex due to the diverse and intricate nature of firm operations (i.e. geography, size, business activities, interlinks with other stakeholders). As a result, corporate sustainability assessments are plagued by subjectivity both within data that reflect corporate sustainability efforts (i.e. corporate sustainability disclosures) and the analysts evaluating them. This subjectivity can be distilled into distinct challenges, such as incompleteness, ambiguity, unreliability and sophistication on the data dimension, as well as limited resources and potential bias on the analyst dimension. Put together, subjectivity hinders effective cost attribution to entities non-compliant with prevailing sustainability expectations, potentially rendering sustainability efforts and its associated regulations futile. To this end, we argue that Explainable Natural Language Processing (XNLP) can significantly enhance corporate sustainability analysis. Specifically, linguistic understanding algorithms (lexical, semantic, syntactic), integrated with XAI capabilities (interpretability, explainability, faithfulness), can bridge gaps in analyst resources and mitigate subjectivity problems within data.
♻ ☆ JOOCI: a Framework for Learning Comprehensive Speech Representations ICLR 2025
Information in speech can be divided into two categories: what is being said (content) and how it is expressed (other). Current state-of-the-art (SOTA) techniques model speech at fixed segments, usually 10-25 ms, using a single embedding. Given the orthogonal nature of other and content information, attempting to optimize both within a single embedding results in suboptimal solutions. This approach divides the models capacity, limiting its ability to build complex hierarchical features effectively. In this work, we present an end-to-end speech representation learning framework designed to jointly optimize the other and content information (JOOCI) in speech. By using separate learnable parameters, JOOCI addresses this optimization challenge by modeling other and content information independently. Our results show that JOOCI consistently outperforms other SOTA models of similar size (100 million parameters) and pre-training data used (960 hours) by a significant margin when evaluated on a range of speech downstream tasks in the SUPERB benchmark, as shown in Table 1.
comment: Submitted to ICLR 2025
♻ ☆ Translation Canvas: An Explainable Interface to Pinpoint and Analyze Translation Systems
With the rapid advancement of machine translation research, evaluation toolkits have become essential for benchmarking system progress. Tools like COMET and SacreBLEU offer single quality score assessments that are effective for pairwise system comparisons. However, these tools provide limited insights for fine-grained system-level comparisons and the analysis of instance-level defects. To address these limitations, we introduce Translation Canvas, an explainable interface designed to pinpoint and analyze translation systems' performance: 1) Translation Canvas assists machine translation researchers in comprehending system-level model performance by identifying common errors (their frequency and severity) and analyzing relationships between different systems based on various evaluation metrics. 2) It supports fine-grained analysis by highlighting error spans with explanations and selectively displaying systems' predictions. According to human evaluation, Translation Canvas demonstrates superior performance over COMET and SacreBLEU packages under enjoyability and understandability criteria.
comment: 7 pages, 3 figures
♻ ☆ Discovering Elementary Discourse Units in Textual Data Using Canonical Correlation Analysis
Canonical Correlation Analysis (CCA) has been exploited immensely for learning latent representations in various fields. This study takes a step further by demonstrating the potential of CCA in identifying Elementary Discourse Units(EDUs) that captures the latent information within the textual data. The probabilistic interpretation of CCA discussed in this study utilizes the two-view nature of textual data, i.e. the consecutive sentences in a document or turns in a dyadic conversation, and has a strong theoretical foundation. Furthermore, this study proposes a model for Elementary Discourse Unit(EDU) segmentation that discovers EDUs in textual data without any supervision. To validate the model, the EDUs are utilized as textual unit for content selection in textual similarity task. Empirical results on Semantic Textual Similarity(STSB) and Mohler datasets confirm that, despite represented as a unigram, the EDUs deliver competitive results and can even beat various sophisticated supervised techniques. The model is simple, linear, adaptable and language independent making it an ideal baseline particularly when labeled training data is scarce or nonexistent.
♻ ☆ Knowledge Circuits in Pretrained Transformers NeurIPS 2024
The remarkable capabilities of modern large language models are rooted in their vast repositories of knowledge encoded within their parameters, enabling them to perceive the world and engage in reasoning. The inner workings of how these models store knowledge have long been a subject of intense interest and investigation among researchers. To date, most studies have concentrated on isolated components within these models, such as the Multilayer Perceptrons and attention head. In this paper, we delve into the computation graph of the language model to uncover the knowledge circuits that are instrumental in articulating specific knowledge. The experiments, conducted with GPT2 and TinyLLAMA, have allowed us to observe how certain information heads, relation heads, and Multilayer Perceptrons collaboratively encode knowledge within the model. Moreover, we evaluate the impact of current knowledge editing techniques on these knowledge circuits, providing deeper insights into the functioning and constraints of these editing methodologies. Finally, we utilize knowledge circuits to analyze and interpret language model behaviors such as hallucinations and in-context learning. We believe the knowledge circuits hold potential for advancing our understanding of Transformers and guiding the improved design of knowledge editing. Code and data are available in https://github.com/zjunlp/KnowledgeCircuits.
comment: NeurIPS 2024, 32 pages
♻ ☆ Examining Long-Context Large Language Models for Environmental Review Document Comprehension
As LLMs become increasingly ubiquitous, researchers have tried various techniques to augment the knowledge provided to these models. Long context and retrieval-augmented generation (RAG) are two such methods that have recently gained popularity. In this work, we examine the benefits of both of these techniques by utilizing question answering (QA) task in a niche domain. While the effectiveness of LLM-based QA systems has already been established at an acceptable level in popular domains such as trivia and literature, it has not often been established in niche domains that traditionally require specialized expertise. We construct the NEPAQuAD1.0 benchmark to evaluate the performance of five long-context LLMs -- Claude Sonnet, Gemini, GPT-4, Llama 3.1, and Mistral -- when answering questions originating from Environmental Impact Statements prepared by U.S. federal government agencies in accordance with the National Environmental Environmental Act (NEPA). We specifically measure the ability of LLMs to understand the nuances of legal, technical, and compliance-related information present in NEPA documents in different contextual scenarios. We test the LLMs' internal prior NEPA knowledge by providing questions without any context, as well as assess how LLMs synthesize the contextual information present in long NEPA documents to facilitate the question/answering task. We compare the performance of the models in handling different types of questions (e.g., problem-solving, divergent, etc.). Our results suggest that RAG powered models significantly outperform those provided with only the PDF context in terms of answer accuracy, regardless of the choice of the LLM. Our further analysis reveals that many models perform better answering closed type questions (Yes/No) than divergent and problem-solving questions.
comment: 14 pages
♻ ☆ Rewards-in-Context: Multi-objective Alignment of Foundation Models with Dynamic Preference Adjustment ICML 2024
We consider the problem of multi-objective alignment of foundation models with human preferences, which is a critical step towards helpful and harmless AI systems. However, it is generally costly and unstable to fine-tune large foundation models using reinforcement learning (RL), and the multi-dimensionality, heterogeneity, and conflicting nature of human preferences further complicate the alignment process. In this paper, we introduce Rewards-in-Context (RiC), which conditions the response of a foundation model on multiple rewards in its prompt context and applies supervised fine-tuning for alignment. The salient features of RiC are simplicity and adaptivity, as it only requires supervised fine-tuning of a single foundation model and supports dynamic adjustment for user preferences during inference time. Inspired by the analytical solution of an abstracted convex optimization problem, our dynamic inference-time adjustment method approaches the Pareto-optimal solution for multiple objectives. Empirical evidence demonstrates the efficacy of our method in aligning both Large Language Models (LLMs) and diffusion models to accommodate diverse rewards with only around 10% GPU hours compared with multi-objective RL baseline.
comment: Accepted by ICML 2024
♻ ☆ Continual Dialogue State Tracking via Reason-of-Select Distillation ACL 2024
An ideal dialogue system requires continuous skill acquisition and adaptation to new tasks while retaining prior knowledge. Dialogue State Tracking (DST), vital in these systems, often involves learning new services and confronting catastrophic forgetting, along with a critical capability loss termed the "Value Selection Quandary." To address these challenges, we introduce the Reason-of-Select (RoS) distillation method by enhancing smaller models with a novel 'meta-reasoning' capability. Meta-reasoning employs an enhanced multi-domain perspective, combining fragments of meta-knowledge from domain-specific dialogues during continual learning. This transcends traditional single-perspective reasoning. The domain bootstrapping process enhances the model's ability to dissect intricate dialogues from multiple possible values. Its domain-agnostic property aligns data distribution across different domains, effectively mitigating forgetting. Additionally, two novel improvements, "multi-value resolution" strategy and Semantic Contrastive Reasoning Selection method, significantly enhance RoS by generating DST-specific selection chains and mitigating hallucinations in teachers' reasoning, ensuring effective and reliable knowledge transfer. Extensive experiments validate the exceptional performance and robust generalization capabilities of our method. The source code is provided for reproducibility.
comment: Accepted to ACL 2024 Findings
♻ ☆ Quantifying and Optimizing Global Faithfulness in Persona-driven Role-playing NeurIPS2024
Persona-driven role-playing (PRP) aims to build AI characters that can respond to user queries by faithfully sticking with all persona statements. Unfortunately, existing faithfulness criteria for PRP are limited to coarse-grained LLM-based scoring without a clear definition or formulation. This paper presents a pioneering exploration to quantify PRP faithfulness as a fine-grained and explainable criterion, which also serves as a reliable reference for optimization. Our criterion first discriminates persona statements into active and passive constraints by identifying the query-statement relevance. Then, we incorporate all constraints following the principle that the AI character's response should be (a) entailed by active (relevant) constraints and (b) not contradicted by passive (irrelevant) constraints. We translate this principle mathematically into a novel Active-Passive-Constraint (APC) score, a constraint-wise sum of natural language inference (NLI) scores weighted by relevance scores. In practice, we build the APC scoring system by symbolically distilling small discriminators from GPT-4 for efficiency. We validate the quality of the APC score against human evaluation based on example personas with tens of statements, and the results show a high correlation. We further leverage it as a reward system in direct preference optimization (DPO) for better AI characters. Our experiments offer a fine-grained and explainable comparison between existing PRP techniques, revealing their advantages and limitations. We further find APC-based DPO to be one of the most competitive techniques for sticking with all constraints and can be well incorporated with other techniques. We then extend the scale of the experiments to real persons with hundreds of statements and reach a consistent conclusion.
comment: NeurIPS2024
♻ ☆ Mechanistic interpretability of large language models with applications to the financial services industry
Large Language Models such as GPTs (Generative Pre-trained Transformers) exhibit remarkable capabilities across a broad spectrum of applications. Nevertheless, due to their intrinsic complexity, these models present substantial challenges in interpreting their internal decision-making processes. This lack of transparency poses critical challenges when it comes to their adaptation by financial institutions, where concerns and accountability regarding bias, fairness, and reliability are of paramount importance. Mechanistic interpretability aims at reverse engineering complex AI models such as transformers. In this paper, we are pioneering the use of mechanistic interpretability to shed some light on the inner workings of large language models for use in financial services applications. We offer several examples of how algorithmic tasks can be designed for compliance monitoring purposes. In particular, we investigate GPT-2 Small's attention pattern when prompted to identify potential violation of Fair Lending laws. Using direct logit attribution, we study the contributions of each layer and its corresponding attention heads to the logit difference in the residual stream. Finally, we design clean and corrupted prompts and use activation patching as a causal intervention method to localize our task completion components further. We observe that the (positive) heads $10.2$ (head $2$, layer $10$), $10.7$, and $11.3$, as well as the (negative) heads $9.6$ and $10.6$ play a significant role in the task completion.
♻ ☆ MarkLLM: An Open-Source Toolkit for LLM Watermarking EMNLP 2024
LLM watermarking, which embeds imperceptible yet algorithmically detectable signals in model outputs to identify LLM-generated text, has become crucial in mitigating the potential misuse of large language models. However, the abundance of LLM watermarking algorithms, their intricate mechanisms, and the complex evaluation procedures and perspectives pose challenges for researchers and the community to easily experiment with, understand, and assess the latest advancements. To address these issues, we introduce MarkLLM, an open-source toolkit for LLM watermarking. MarkLLM offers a unified and extensible framework for implementing LLM watermarking algorithms, while providing user-friendly interfaces to ensure ease of access. Furthermore, it enhances understanding by supporting automatic visualization of the underlying mechanisms of these algorithms. For evaluation, MarkLLM offers a comprehensive suite of 12 tools spanning three perspectives, along with two types of automated evaluation pipelines. Through MarkLLM, we aim to support researchers while improving the comprehension and involvement of the general public in LLM watermarking technology, fostering consensus and driving further advancements in research and application. Our code is available at https://github.com/THU-BPM/MarkLLM.
comment: EMNLP 2024 Demo
♻ ☆ Your Mixture-of-Experts LLM Is Secretly an Embedding Model For Free
While large language models (LLMs) excel on generation tasks, their decoder-only architecture often limits their potential as embedding models if no further representation finetuning is applied. Does this contradict their claim of generalists? To answer the question, we take a closer look at Mixture-of-Experts (MoE) LLMs. Our study shows that the expert routers in MoE LLMs can serve as an off-the-shelf embedding model with promising performance on a diverse class of embedding-focused tasks, without requiring any finetuning. Moreover, our extensive analysis shows that the MoE routing weights (RW) is complementary to the hidden state (HS) of LLMs, a widely-used embedding. Compared to HS, we find that RW is more robust to the choice of prompts and focuses on high-level semantics. Motivated by the analysis, we propose MoEE combining RW and HS, which achieves better performance than using either separately. Our exploration of their combination and prompting strategy shed several novel insights, e.g., a weighted sum of RW and HS similarities outperforms the similarity on their concatenation. Our experiments are conducted on 6 embedding tasks with 20 datasets from the Massive Text Embedding Benchmark (MTEB). The results demonstrate the significant improvement brought by MoEE to LLM-based embedding without further finetuning.
comment: Code: https://github.com/tianyi-lab/MoE-Embedding
♻ ☆ Free-text Rationale Generation under Readability Level Control
Free-text rationales justify model decisions in natural language and thus become likable and accessible among approaches to explanation across many tasks. However, their effectiveness can be hindered by misinterpretation and hallucination. As a perturbation test, we investigate how large language models (LLMs) perform rationale generation under the effects of readability level control, i.e., being prompted for an explanation targeting a specific expertise level, such as sixth grade or college. We find that explanations are adaptable to such instruction, though the requested readability is often misaligned with the measured text complexity according to traditional readability metrics. Furthermore, the generated rationales tend to feature medium level complexity, which correlates with the measured quality using automatic metrics. Finally, our human annotators confirm a generally satisfactory impression on rationales at all readability levels, with high-school-level readability being most commonly perceived and favored.
♻ ☆ IntGrad MT: Eliciting LLMs' Machine Translation Capabilities with Sentence Interpolation and Gradual MT
Recent Large Language Models (LLMs) have demonstrated strong performance in translation without needing to be finetuned on additional parallel corpora. However, they still underperform for low-resource language pairs. Previous works have focused on mitigating this issue by leveraging relevant few-shot examples or external resources such as dictionaries or grammar books, making models heavily reliant on these nonparametric sources of information. In this paper, we propose a novel method named IntGrad MT that focuses on fully exploiting an LLM's inherent translation capability. IntGrad MT achieves this by constructing a chain of few-shot examples, each consisting of a source sentence and the model's own translation, that rise incrementally in difficulty. IntGrad MT employs two techniques: Sentence Interpolation, which generates a sequence of sentences that gradually change from an easy sentence to translate to a difficult one, and Gradual MT, which sequentially translates this chain using translations of earlier sentences as few-shot examples for the translation of subsequent ones. With this approach, we observe a substantial enhancement in the xCOMET scores of various LLMs for multiple languages, especially in low-resource languages such as Hindi(8.26), Swahili(7.10), Bengali(6.97) and Marathi(13.03). Our approach presents a practical way of enhancing LLMs' performance without extra training.
♻ ☆ Language-based Valence and Arousal Expressions between the United States and China: a Cross-Cultural Examination
While affective expressions on social media have been extensively studied, most research has focused on the Western context. This paper explores cultural differences in affective expressions by comparing valence and arousal on Twitter/X (geolocated to the US) and Sina Weibo (in Mainland China). Using the NRC-VAD lexicon to measure valence and arousal, we identify distinct patterns of emotional expression across both platforms. Our analysis reveals a functional representation between valence and arousal, showing a negative offset in contrast to traditional lab-based findings which suggest a positive offset. Furthermore, we uncover significant cross-cultural differences in arousal, with US users displaying higher emotional intensity than Chinese users, regardless of the valence of the content. Finally, we conduct a comprehensive language analysis correlating n-grams and LDA topics with affective dimensions to deepen our understanding of how language and culture shape emotional expression. These findings contribute to a more nuanced understanding of affective communication across cultural and linguistic contexts on social media.
comment: preview
♻ ☆ PKU-SafeRLHF: Towards Multi-Level Safety Alignment for LLMs with Human Preference
In this work, we introduce the PKU-SafeRLHF dataset, designed to promote research on safety alignment in large language models (LLMs). As a sibling project to SafeRLHF and BeaverTails, we separate annotations of helpfulness and harmlessness for question-answering pairs, providing distinct perspectives on these coupled attributes. Overall, we provide 44.6k refined prompts and 265k question-answer pairs with safety meta-labels for 19 harm categories and three severity levels ranging from minor to severe, with answers generated by Llama-family models. Based on this, we collected 166.8k preference data, including dual-preference (helpfulness and harmlessness decoupled) and single-preference data (trade-off the helpfulness and harmlessness from scratch), respectively. Using the large-scale annotation data, we further train severity-sensitive moderation for the risk control of LLMs and safety-centric RLHF algorithms for the safety alignment of LLMs. We believe this dataset will be a valuable resource for the community, aiding in the safe deployment of LLMs.
comment: a sibling project to SafeRLHF and BeaverTails
♻ ☆ Learning from Committee: Reasoning Distillation from a Mixture of Teachers with Peer-Review
While reasoning capabilities typically emerge in large language models (LLMs) with tens of billions of parameters, recent research focuses on improving smaller open-source models through knowledge distillation (KD) from commercial LLMs. However, many of these studies rely solely on responses from a single LLM as the gold rationale, unlike the natural human learning process, which involves understanding both the correct answers and the reasons behind mistakes. In this paper, we introduce a novel Fault-Aware Distillation via Peer-Review (FAIR) approach: 1) Instead of merely obtaining gold rationales from teachers, our method asks teachers to identify and explain the student's mistakes, providing customized instruction learning data. 2) We design a simulated peer-review process between teacher LLMs, which selects only the generated rationales above the acceptance threshold. This reduces the chance of teachers guessing correctly with flawed rationale, improving instructional data quality. Comprehensive experiments and analysis on mathematical, commonsense, and logical reasoning tasks demonstrate the effectiveness of our method.
comment: 14 pages, 5 figures
♻ ☆ AI-Oracle Machines for Intelligent Computing
We introduce the concept of AI-oracle machines for intelligent computing and outline several applications to demonstrate their potential. Following this, we advocate for the development of a comprehensive platform to streamline the implementation of AI-oracle machines.
comment: 6 pages
♻ ☆ Beyond Silent Letters: Amplifying LLMs in Emotion Recognition with Vocal Nuances
Emotion recognition in speech is a challenging multimodal task that requires understanding both verbal content and vocal nuances. This paper introduces a novel approach to emotion detection using Large Language Models (LLMs), which have demonstrated exceptional capabilities in natural language understanding. To overcome the inherent limitation of LLMs in processing audio inputs, we propose SpeechCueLLM, a method that translates speech characteristics into natural language descriptions, allowing LLMs to perform multimodal emotion analysis via text prompts without any architectural changes. Our method is minimal yet impactful, outperforming baseline models that require structural modifications. We evaluate SpeechCueLLM on two datasets: IEMOCAP and MELD, showing significant improvements in emotion recognition accuracy, particularly for high-quality audio data. We also explore the effectiveness of various feature representations and fine-tuning strategies for different LLMs. Our experiments demonstrate that incorporating speech descriptions yields a more than 2% increase in the average weighted F1 score on IEMOCAP (from 70.111% to 72.596%).
♻ ☆ Self-Reflection in LLM Agents: Effects on Problem-Solving Performance
In this study, we investigated the effects of self-reflection in large language models (LLMs) on problem-solving performance. We instructed nine popular LLMs to answer a series of multiple-choice questions to provide a performance baseline. For each incorrectly answered question, we instructed eight types of self-reflecting LLM agents to reflect on their mistakes and provide themselves with guidance to improve problem-solving. Then, using this guidance, each self-reflecting agent attempted to re-answer the same questions. Our results indicate that LLM agents are able to significantly improve their problem-solving performance through self-reflection ($p < 0.001$). In addition, we compared the various types of self-reflection to determine their individual contribution to performance. All code and data are available on GitHub at https://github.com/matthewrenze/self-reflection
♻ ☆ EmbedLLM: Learning Compact Representations of Large Language Models
With hundreds of thousands of language models available on Huggingface today, efficiently evaluating and utilizing these models across various downstream, tasks has become increasingly critical. Many existing methods repeatedly learn task-specific representations of Large Language Models (LLMs), which leads to inefficiencies in both time and computational resources. To address this, we propose EmbedLLM, a framework designed to learn compact vector representations, of LLMs that facilitate downstream applications involving many models, such as model routing. We introduce an encoder-decoder approach for learning such embeddings, along with a systematic framework to evaluate their effectiveness. Empirical results show that EmbedLLM outperforms prior methods in model routing both in accuracy and latency. Additionally, we demonstrate that our method can forecast a model's performance on multiple benchmarks, without incurring additional inference cost. Extensive probing experiments validate that the learned embeddings capture key model characteristics, e.g. whether the model is specialized for coding tasks, even without being explicitly trained on them. We open source our dataset, code and embedder to facilitate further research and application.
♻ ☆ X-ray Made Simple: Radiology Report Generation and Evaluation with Layman's Terms
Radiology Report Generation (RRG) has achieved significant progress with the advancements of multimodal generative models. However, the evaluation in the domain suffers from a lack of fair and robust metrics. We reveal that, high performance on RRG with existing lexical-based metrics (e.g. BLEU) might be more of a mirage - a model can get a high BLEU only by learning the template of reports. This has become an urgent problem for RRG due to the highly patternized nature of these reports. In this work, we un-intuitively approach this problem by proposing the Layman's RRG framework, a layman's terms-based dataset, evaluation and training framework that systematically improves RRG with day-to-day language. We first contribute the translated Layman's terms dataset. Building upon the dataset, we then propose a semantics-based evaluation method, which is proved to mitigate the inflated numbers of BLEU and provides fairer evaluation. Last, we show that training on the layman's terms dataset encourages models to focus on the semantics of the reports, as opposed to overfitting to learning the report templates. We reveal a promising scaling law between the number of training examples and semantics gain provided by our dataset, compared to the inverse pattern brought by the original formats. Our code is available at \url{https://github.com/hegehongcha/LaymanRRG}.
♻ ☆ MORL-Prompt: An Empirical Analysis of Multi-Objective Reinforcement Learning for Discrete Prompt Optimization
RL-based techniques can be employed to search for prompts that, when fed into a target language model, maximize a set of user-specified reward functions. However, in many target applications, the natural reward functions are in tension with one another -- for example, content preservation vs. style matching in style transfer tasks. Current techniques focus on maximizing the average of reward functions, which does not necessarily lead to prompts that achieve balance across rewards -- an issue that has been well-studied in the multi-objective and robust optimization literature. In this paper, we conduct an empirical comparison of several existing multi-objective optimization techniques adapted to this new setting: RL-based discrete prompt optimization. We compare two methods optimizing the volume of the Pareto reward surface and one method that chooses an update direction that benefits all rewards simultaneously. We evaluate performance on two NLP tasks: style transfer and machine translation, each using three competing reward functions. Our experiments demonstrate that multi-objective methods that directly optimize the volume of the Pareto reward surface perform better and achieve a better balance of all rewards than those that attempt to find monotonic update directions.
Computer Vision and Pattern Recognition 171
☆ Dual Prototype Evolving for Test-Time Generalization of Vision-Language Models NeurIPS 2024
Test-time adaptation, which enables models to generalize to diverse data with unlabeled test samples, holds significant value in real-world scenarios. Recently, researchers have applied this setting to advanced pre-trained vision-language models (VLMs), developing approaches such as test-time prompt tuning to further extend their practical applicability. However, these methods typically focus solely on adapting VLMs from a single modality and fail to accumulate task-specific knowledge as more samples are processed. To address this, we introduce Dual Prototype Evolving (DPE), a novel test-time adaptation approach for VLMs that effectively accumulates task-specific knowledge from multi-modalities. Specifically, we create and evolve two sets of prototypes--textual and visual--to progressively capture more accurate multi-modal representations for target classes during test time. Moreover, to promote consistent multi-modal representations, we introduce and optimize learnable residuals for each test sample to align the prototypes from both modalities. Extensive experimental results on 15 benchmark datasets demonstrate that our proposed DPE consistently outperforms previous state-of-the-art methods while also exhibiting competitive computational efficiency. Code is available at https://github.com/zhangce01/DPE-CLIP.
comment: Accepted by NeurIPS 2024. Project page: https://zhangce01.github.io/DPE-CLIP
☆ The Curse of Multi-Modalities: Evaluating Hallucinations of Large Multimodal Models across Language, Visual, and Audio
Recent advancements in large multimodal models (LMMs) have significantly enhanced performance across diverse tasks, with ongoing efforts to further integrate additional modalities such as video and audio. However, most existing LMMs remain vulnerable to hallucinations, the discrepancy between the factual multimodal input and the generated textual output, which has limited their applicability in various real-world scenarios. This paper presents the first systematic investigation of hallucinations in LMMs involving the three most common modalities: language, visual, and audio. Our study reveals two key contributors to hallucinations: overreliance on unimodal priors and spurious inter-modality correlations. To address these challenges, we introduce the benchmark The Curse of Multi-Modalities (CMM), which comprehensively evaluates hallucinations in LMMs, providing a detailed analysis of their underlying issues. Our findings highlight key vulnerabilities, including imbalances in modality integration and biases from training data, underscoring the need for balanced cross-modal learning and enhanced hallucination mitigation strategies. Based on our observations and findings, we suggest potential research directions that could enhance the reliability of LMMs.
comment: Project Page: cmm-damovl.site
☆ Long-LRM: Long-sequence Large Reconstruction Model for Wide-coverage Gaussian Splats
We propose Long-LRM, a generalizable 3D Gaussian reconstruction model that is capable of reconstructing a large scene from a long sequence of input images. Specifically, our model can process 32 source images at 960x540 resolution within only 1.3 seconds on a single A100 80G GPU. Our architecture features a mixture of the recent Mamba2 blocks and the classical transformer blocks which allowed many more tokens to be processed than prior work, enhanced by efficient token merging and Gaussian pruning steps that balance between quality and efficiency. Unlike previous feed-forward models that are limited to processing 1~4 input images and can only reconstruct a small portion of a large scene, Long-LRM reconstructs the entire scene in a single feed-forward step. On large-scale scene datasets such as DL3DV-140 and Tanks and Temples, our method achieves performance comparable to optimization-based approaches while being two orders of magnitude more efficient. Project page: https://arthurhero.github.io/projects/llrm
☆ Meta-Unlearning on Diffusion Models: Preventing Relearning Unlearned Concepts
With the rapid progress of diffusion-based content generation, significant efforts are being made to unlearn harmful or copyrighted concepts from pretrained diffusion models (DMs) to prevent potential model misuse. However, it is observed that even when DMs are properly unlearned before release, malicious finetuning can compromise this process, causing DMs to relearn the unlearned concepts. This occurs partly because certain benign concepts (e.g., "skin") retained in DMs are related to the unlearned ones (e.g., "nudity"), facilitating their relearning via finetuning. To address this, we propose meta-unlearning on DMs. Intuitively, a meta-unlearned DM should behave like an unlearned DM when used as is; moreover, if the meta-unlearned DM undergoes malicious finetuning on unlearned concepts, the related benign concepts retained within it will be triggered to self-destruct, hindering the relearning of unlearned concepts. Our meta-unlearning framework is compatible with most existing unlearning methods, requiring only the addition of an easy-to-implement meta objective. We validate our approach through empirical experiments on meta-unlearning concepts from Stable Diffusion models (SD-v1-4 and SDXL), supported by extensive ablation studies. Our code is available at https://github.com/sail-sg/Meta-Unlearning.
☆ Towards Zero-Shot Camera Trap Image Categorization
This paper describes the search for an alternative approach to the automatic categorization of camera trap images. First, we benchmark state-of-the-art classifiers using a single model for all images. Next, we evaluate methods combining MegaDetector with one or more classifiers and Segment Anything to assess their impact on reducing location-specific overfitting. Last, we propose and test two approaches using large language and foundational models, such as DINOv2, BioCLIP, BLIP, and ChatGPT, in a zero-shot scenario. Evaluation carried out on two publicly available datasets (WCT from New Zealand, CCT20 from the Southwestern US) and a private dataset (CEF from Central Europe) revealed that combining MegaDetector with two separate classifiers achieves the highest accuracy. This approach reduced the relative error of a single BEiTV2 classifier by approximately 42\% on CCT20, 48\% on CEF, and 75\% on WCT. Besides, as the background is removed, the error in terms of accuracy in new locations is reduced to half. The proposed zero-shot pipeline based on DINOv2 and FAISS achieved competitive results (1.0\% and 4.7\% smaller on CCT20, and CEF, respectively), which highlights the potential of zero-shot approaches for camera trap image categorization.
☆ Gravity-aligned Rotation Averaging with Circular Regression ECCV2024
Reconstructing a 3D scene from unordered images is pivotal in computer vision and robotics, with applications spanning crowd-sourced mapping and beyond. While global Structure-from-Motion (SfM) techniques are scalable and fast, they often compromise on accuracy. To address this, we introduce a principled approach that integrates gravity direction into the rotation averaging phase of global pipelines, enhancing camera orientation accuracy and reducing the degrees of freedom. This additional information is commonly available in recent consumer devices, such as smartphones, mixed-reality devices and drones, making the proposed method readily accessible. Rooted in circular regression, our algorithm has similar convergence guarantees as linear regression. It also supports scenarios where only a subset of cameras have known gravity. Additionally, we propose a mechanism to refine error-prone gravity. We achieve state-of-the-art accuracy on four large-scale datasets. Particularly, the proposed method improves upon the SfM baseline by 13 AUC@$1^\circ$ points, on average, while running eight times faster. It also outperforms the standard planar pose graph optimization technique by 23 AUC@$1^\circ$ points. The code is at https://github.com/colmap/glomap.
comment: accepted at ECCV2024
☆ SAFREE: Training-Free and Adaptive Guard for Safe Text-to-Image And Video Generation
Recent advances in diffusion models have significantly enhanced their ability to generate high-quality images and videos, but they have also increased the risk of producing unsafe content. Existing unlearning/editing-based methods for safe generation remove harmful concepts from models but face several challenges: (1) They cannot instantly remove harmful concepts without training. (2) Their safe generation capabilities depend on collected training data. (3) They alter model weights, risking degradation in quality for content unrelated to toxic concepts. To address these, we propose SAFREE, a novel, training-free approach for safe T2I and T2V, that does not alter the model's weights. Specifically, we detect a subspace corresponding to a set of toxic concepts in the text embedding space and steer prompt embeddings away from this subspace, thereby filtering out harmful content while preserving intended semantics. To balance the trade-off between filtering toxicity and preserving safe concepts, SAFREE incorporates a novel self-validating filtering mechanism that dynamically adjusts the denoising steps when applying the filtered embeddings. Additionally, we incorporate adaptive re-attention mechanisms within the diffusion latent space to selectively diminish the influence of features related to toxic concepts at the pixel level. In the end, SAFREE ensures coherent safety checking, preserving the fidelity, quality, and safety of the output. SAFREE achieves SOTA performance in suppressing unsafe content in T2I generation compared to training-free baselines and effectively filters targeted concepts while maintaining high-quality images. It also shows competitive results against training-based methods. We extend SAFREE to various T2I backbones and T2V tasks, showcasing its flexibility and generalization. SAFREE provides a robust and adaptable safeguard for ensuring safe visual generation.
comment: The first two authors contributed equally; Project page: https://safree-safe-t2i-t2v.github.io/
☆ PND-Net: Plant Nutrition Deficiency and Disease Classification using Graph Convolutional Network
Crop yield production could be enhanced for agricultural growth if various plant nutrition deficiencies, and diseases are identified and detected at early stages. The deep learning methods have proven its superior performances in the automated detection of plant diseases and nutrition deficiencies from visual symptoms in leaves. This article proposes a new deep learning method for plant nutrition deficiencies and disease classification using a graph convolutional network (GNN), added upon a base convolutional neural network (CNN). Sometimes, a global feature descriptor might fail to capture the vital region of a diseased leaf, which causes inaccurate classification of disease. To address this issue, regional feature learning is crucial for a holistic feature aggregation. In this work, region-based feature summarization at multi-scales is explored using spatial pyramidal pooling for discriminative feature representation. A GCN is developed to capacitate learning of finer details for classifying plant diseases and insufficiency of nutrients. The proposed method, called Plant Nutrition Deficiency and Disease Network (PND-Net), is evaluated on two public datasets for nutrition deficiency, and two for disease classification using four CNNs. The best classification performances are: (a) 90.00% Banana and 90.54% Coffee nutrition deficiency; and (b) 96.18% Potato diseases and 84.30% on PlantDoc datasets using Xception backbone. Furthermore, additional experiments have been carried out for generalization, and the proposed method has achieved state-of-the-art performances on two public datasets, namely the Breast Cancer Histopathology Image Classification (BreakHis 40X: 95.50%, and BreakHis 100X: 96.79% accuracy) and Single cells in Pap smear images for cervical cancer classification (SIPaKMeD: 99.18% accuracy). Also, PND-Net achieves improved performances using five-fold cross validation.
☆ Optimizing 3D Geometry Reconstruction from Implicit Neural Representations
Implicit neural representations have emerged as a powerful tool in learning 3D geometry, offering unparalleled advantages over conventional representations like mesh-based methods. A common type of INR implicitly encodes a shape's boundary as the zero-level set of the learned continuous function and learns a mapping from a low-dimensional latent space to the space of all possible shapes represented by its signed distance function. However, most INRs struggle to retain high-frequency details, which are crucial for accurate geometric depiction, and they are computationally expensive. To address these limitations, we present a novel approach that both reduces computational expenses and enhances the capture of fine details. Our method integrates periodic activation functions, positional encodings, and normals into the neural network architecture. This integration significantly enhances the model's ability to learn the entire space of 3D shapes while preserving intricate details and sharp features, areas where conventional representations often fall short.
☆ RAFA-Net: Region Attention Network For Food Items And Agricultural Stress Recognition
Deep Convolutional Neural Networks (CNNs) have facilitated remarkable success in recognizing various food items and agricultural stress. A decent performance boost has been witnessed in solving the agro-food challenges by mining and analyzing of region-based partial feature descriptors. Also, computationally expensive ensemble learning schemes using multiple CNNs have been studied in earlier works. This work proposes a region attention scheme for modelling long-range dependencies by building a correlation among different regions within an input image. The attention method enhances feature representation by learning the usefulness of context information from complementary regions. Spatial pyramidal pooling and average pooling pair aggregate partial descriptors into a holistic representation. Both pooling methods establish spatial and channel-wise relationships without incurring extra parameters. A context gating scheme is applied to refine the descriptiveness of weighted attentional features, which is relevant for classification. The proposed Region Attention network for Food items and Agricultural stress recognition method, dubbed RAFA-Net, has been experimented on three public food datasets, and has achieved state-of-the-art performances with distinct margins. The highest top-1 accuracies of RAFA-Net are 91.69%, 91.56%, and 96.97% on the UECFood-100, UECFood-256, and MAFood-121 datasets, respectively. In addition, better accuracies have been achieved on two benchmark agricultural stress datasets. The best top-1 accuracies on the Insect Pest (IP-102) and PlantDoc-27 plant disease datasets are 92.36%, and 85.54%, respectively; implying RAFA-Net's generalization capability.
☆ WorldCuisines: A Massive-Scale Benchmark for Multilingual and Multicultural Visual Question Answering on Global Cuisines
Vision Language Models (VLMs) often struggle with culture-specific knowledge, particularly in languages other than English and in underrepresented cultural contexts. To evaluate their understanding of such knowledge, we introduce WorldCuisines, a massive-scale benchmark for multilingual and multicultural, visually grounded language understanding. This benchmark includes a visual question answering (VQA) dataset with text-image pairs across 30 languages and dialects, spanning 9 language families and featuring over 1 million data points, making it the largest multicultural VQA benchmark to date. It includes tasks for identifying dish names and their origins. We provide evaluation datasets in two sizes (12k and 60k instances) alongside a training dataset (1 million instances). Our findings show that while VLMs perform better with correct location context, they struggle with adversarial contexts and predicting specific regional cuisines and languages. To support future research, we release a knowledge base with annotated food entries and images along with the VQA data.
☆ Embedding an Ethical Mind: Aligning Text-to-Image Synthesis via Lightweight Value Optimization
Recent advancements in diffusion models trained on large-scale data have enabled the generation of indistinguishable human-level images, yet they often produce harmful content misaligned with human values, e.g., social bias, and offensive content. Despite extensive research on Large Language Models (LLMs), the challenge of Text-to-Image (T2I) model alignment remains largely unexplored. Addressing this problem, we propose LiVO (Lightweight Value Optimization), a novel lightweight method for aligning T2I models with human values. LiVO only optimizes a plug-and-play value encoder to integrate a specified value principle with the input prompt, allowing the control of generated images over both semantics and values. Specifically, we design a diffusion model-tailored preference optimization loss, which theoretically approximates the Bradley-Terry model used in LLM alignment but provides a more flexible trade-off between image quality and value conformity. To optimize the value encoder, we also develop a framework to automatically construct a text-image preference dataset of 86k (prompt, aligned image, violating image, value principle) samples. Without updating most model parameters and through adaptive value selection from the input prompt, LiVO significantly reduces harmful outputs and achieves faster convergence, surpassing several strong baselines and taking an initial step towards ethically aligned T2I models.
comment: Accepted by ACM Multimedia 2024. The dataset and code can be found at https://github.com/achernarwang/LiVO
☆ AdaptiveDrag: Semantic-Driven Dragging on Diffusion-Based Image Editing
Recently, several point-based image editing methods (e.g., DragDiffusion, FreeDrag, DragNoise) have emerged, yielding precise and high-quality results based on user instructions. However, these methods often make insufficient use of semantic information, leading to less desirable results. In this paper, we proposed a novel mask-free point-based image editing method, AdaptiveDrag, which provides a more flexible editing approach and generates images that better align with user intent. Specifically, we design an auto mask generation module using super-pixel division for user-friendliness. Next, we leverage a pre-trained diffusion model to optimize the latent, enabling the dragging of features from handle points to target points. To ensure a comprehensive connection between the input image and the drag process, we have developed a semantic-driven optimization. We design adaptive steps that are supervised by the positions of the points and the semantic regions derived from super-pixel segmentation. This refined optimization process also leads to more realistic and accurate drag results. Furthermore, to address the limitations in the generative consistency of the diffusion model, we introduce an innovative corresponding loss during the sampling process. Building on these effective designs, our method delivers superior generation results using only the single input image and the handle-target point pairs. Extensive experiments have been conducted and demonstrate that the proposed method outperforms others in handling various drag instructions (e.g., resize, movement, extension) across different domains (e.g., animals, human face, land space, clothing).
☆ MultiCamCows2024 -- A Multi-view Image Dataset for AI-driven Holstein-Friesian Cattle Re-Identification on a Working Farm
We present MultiCamCows2024, a farm-scale image dataset filmed across multiple cameras for the biometric identification of individual Holstein-Friesian cattle exploiting their unique black and white coat-patterns. Captured by three ceiling-mounted visual sensors covering adjacent barn areas over seven days on a working dairy farm, the dataset comprises 101, 329 images of 90 cows, plus the underlying original CCTV footage. The dataset is provided alongside full computer vision recognition baselines, that is both a supervised and self-supervised learning framework for individual cow identification trained on cattle tracklets. We report a performance above 96% single image identification accuracy from the dataset and demonstrate that combining data from multiple cameras during learning enhances self-supervised identification. We show that our framework enables fully automatic cattle identification, barring only the simple human verification of tracklet integrity during data collection. Crucially, our study highlights that multi-camera, supervised and self-supervised components in tandem not only deliver highly accurate individual cow identification but also achieve this efficiently with no labelling of cattle identities by humans at all. We argue that this improvement in efficacy has practical implications for livestock management, behaviour analysis, and agricultural monitoring. For full reproducibility and practical ease of use, we publish all key software and code including re-identification components and the species detector with this paper.
comment: 26 pages, 10 figures
☆ VividMed: Vision Language Model with Versatile Visual Grounding for Medicine
Recent advancements in Vision Language Models (VLMs) have demonstrated remarkable promise in generating visually grounded responses. However, their application in the medical domain is hindered by unique challenges. For instance, most VLMs rely on a single method of visual grounding, whereas complex medical tasks demand more versatile approaches. Additionally, while most VLMs process only 2D images, a large portion of medical images are 3D. The lack of medical data further compounds these obstacles. To address these challenges, we present VividMed, a vision language model with versatile visual grounding for medicine. Our model supports generating both semantic segmentation masks and instance-level bounding boxes, and accommodates various imaging modalities, including both 2D and 3D data. We design a three-stage training procedure and an automatic data synthesis pipeline based on open datasets and models. Besides visual grounding tasks, VividMed also excels in other common downstream tasks, including Visual Question Answering (VQA) and report generation. Ablation studies empirically show that the integration of visual grounding ability leads to improved performance on these tasks. Our code is publicly available at https://github.com/function2-llx/MMMM.
☆ Machine Learning Approach to Brain Tumor Detection and Classification
Brain tumor detection and classification are critical tasks in medical image analysis, particularly in early-stage diagnosis, where accurate and timely detection can significantly improve treatment outcomes. In this study, we apply various statistical and machine learning models to detect and classify brain tumors using brain MRI images. We explore a variety of statistical models including linear, logistic, and Bayesian regressions, and the machine learning models including decision tree, random forest, single-layer perceptron, multi-layer perceptron, convolutional neural network (CNN), recurrent neural network, and long short-term memory. Our findings show that CNN outperforms other models, achieving the best performance. Additionally, we confirm that the CNN model can also work for multi-class classification, distinguishing between four categories of brain MRI images such as normal, glioma, meningioma, and pituitary tumor images. This study demonstrates that machine learning approaches are suitable for brain tumor detection and classification, facilitating real-world medical applications in assisting radiologists with early and accurate diagnosis.
comment: 7 pages, 2 figures, 2 tables
☆ Automatic Mapping of Anatomical Landmarks from Free-Text Using Large Language Models: Insights from Llama-2
Anatomical landmarks are vital in medical imaging for navigation and anomaly detection. Modern large language models (LLMs), like Llama-2, offer promise for automating the mapping of these landmarks in free-text radiology reports to corresponding positions in image data. Recent studies propose LLMs may develop coherent representations of generative processes. Motivated by these insights, we investigated whether LLMs accurately represent the spatial positions of anatomical landmarks. Through experiments with Llama-2 models, we found that they can linearly represent anatomical landmarks in space with considerable robustness to different prompts. These results underscore the potential of LLMs to enhance the efficiency and accuracy of medical imaging workflows.
comment: 6 pages, 2 figures, 1 table
☆ MambaBEV: An efficient 3D detection model with Mamba2
A stable 3D object detection model based on BEV paradigm with temporal information is very important for autonomous driving systems. However, current temporal fusion model use convolutional layer or deformable self-attention is not conducive to the exchange of global information of BEV space and has more computational cost. Recently, a newly proposed based model specialized in processing sequence called mamba has shown great potential in multiple downstream task. In this work, we proposed a mamba2-based BEV 3D object detection model named MambaBEV. We also adapt an end to end self driving paradigm to test the performance of the model. Our work performs pretty good results on nucences datasets:Our base version achieves 51.7% NDS. Our code will be available soon.
☆ 3DIS: Depth-Driven Decoupled Instance Synthesis for Text-to-Image Generation
The increasing demand for controllable outputs in text-to-image generation has spurred advancements in multi-instance generation (MIG), allowing users to define both instance layouts and attributes. However, unlike image-conditional generation methods such as ControlNet, MIG techniques have not been widely adopted in state-of-the-art models like SD2 and SDXL, primarily due to the challenge of building robust renderers that simultaneously handle instance positioning and attribute rendering. In this paper, we introduce Depth-Driven Decoupled Instance Synthesis (3DIS), a novel framework that decouples the MIG process into two stages: (i) generating a coarse scene depth map for accurate instance positioning and scene composition, and (ii) rendering fine-grained attributes using pre-trained ControlNet on any foundational model, without additional training. Our 3DIS framework integrates a custom adapter into LDM3D for precise depth-based layouts and employs a finetuning-free method for enhanced instance-level attribute rendering. Extensive experiments on COCO-Position and COCO-MIG benchmarks demonstrate that 3DIS significantly outperforms existing methods in both layout precision and attribute rendering. Notably, 3DIS offers seamless compatibility with diverse foundational models, providing a robust, adaptable solution for advanced multi-instance generation. The code is available at: https://github.com/limuloo/3DIS.
comment: 10 pages
☆ Cross-Modal Safety Mechanism Transfer in Large Vision-Language Models
Vision-language alignment in Large Vision-Language Models (LVLMs) successfully enables LLMs to understand visual input. However, we find that existing vision-language alignment methods fail to transfer the existing safety mechanism for text in LLMs to vision, which leads to vulnerabilities in toxic image. To explore the cause of this problem, we give the insightful explanation of where and how the safety mechanism of LVLMs operates and conduct comparative analysis between text and vision. We find that the hidden states at the specific transformer layers play a crucial role in the successful activation of safety mechanism, while the vision-language alignment at hidden states level in current methods is insufficient. This results in a semantic shift for input images compared to text in hidden states, therefore misleads the safety mechanism. To address this, we propose a novel Text-Guided vision-language Alignment method (TGA) for LVLMs. TGA retrieves the texts related to input vision and uses them to guide the projection of vision into the hidden states space in LLMs. Experiments show that TGA not only successfully transfers the safety mechanism for text in basic LLMs to vision in vision-language alignment for LVLMs without any safety fine-tuning on the visual modality but also maintains the general performance on various vision tasks (Safe and Good).
☆ Cascade learning in multi-task encoder-decoder networks for concurrent bone segmentation and glenohumeral joint assessment in shoulder CT scans
Osteoarthritis is a degenerative condition affecting bones and cartilage, often leading to osteophyte formation, bone density loss, and joint space narrowing. Treatment options to restore normal joint function vary depending on the severity of the condition. This work introduces an innovative deep-learning framework processing shoulder CT scans. It features the semantic segmentation of the proximal humerus and scapula, the 3D reconstruction of bone surfaces, the identification of the glenohumeral (GH) joint region, and the staging of three common osteoarthritic-related pathologies: osteophyte formation (OS), GH space reduction (JS), and humeroscapular alignment (HSA). The pipeline comprises two cascaded CNN architectures: 3D CEL-UNet for segmentation and 3D Arthro-Net for threefold classification. A retrospective dataset of 571 CT scans featuring patients with various degrees of GH osteoarthritic-related pathologies was used to train, validate, and test the pipeline. Root mean squared error and Hausdorff distance median values for 3D reconstruction were 0.22mm and 1.48mm for the humerus and 0.24mm and 1.48mm for the scapula, outperforming state-of-the-art architectures and making it potentially suitable for a PSI-based shoulder arthroplasty preoperative plan context. The classification accuracy for OS, JS, and HSA consistently reached around 90% across all three categories. The computational time for the inference pipeline was less than 15s, showcasing the framework's efficiency and compatibility with orthopedic radiology practice. The outcomes represent a promising advancement toward the medical translation of artificial intelligence tools. This progress aims to streamline the preoperative planning pipeline delivering high-quality bone surfaces and supporting surgeons in selecting the most suitable surgical approach according to the unique patient joint conditions.
☆ DocLayout-YOLO: Enhancing Document Layout Analysis through Diverse Synthetic Data and Global-to-Local Adaptive Perception
Document Layout Analysis is crucial for real-world document understanding systems, but it encounters a challenging trade-off between speed and accuracy: multimodal methods leveraging both text and visual features achieve higher accuracy but suffer from significant latency, whereas unimodal methods relying solely on visual features offer faster processing speeds at the expense of accuracy. To address this dilemma, we introduce DocLayout-YOLO, a novel approach that enhances accuracy while maintaining speed advantages through document-specific optimizations in both pre-training and model design. For robust document pre-training, we introduce the Mesh-candidate BestFit algorithm, which frames document synthesis as a two-dimensional bin packing problem, generating the large-scale, diverse DocSynth-300K dataset. Pre-training on the resulting DocSynth-300K dataset significantly improves fine-tuning performance across various document types. In terms of model optimization, we propose a Global-to-Local Controllable Receptive Module that is capable of better handling multi-scale variations of document elements. Furthermore, to validate performance across different document types, we introduce a complex and challenging benchmark named DocStructBench. Extensive experiments on downstream datasets demonstrate that DocLayout-YOLO excels in both speed and accuracy. Code, data, and models are available at https://github.com/opendatalab/DocLayout-YOLO.
comment: Github Repo: https://github.com/opendatalab/DocLayout-YOLO
☆ Exploring Model Kinship for Merging Large Language Models
Model merging has become one of the key technologies for enhancing the capabilities and efficiency of Large Language Models (LLMs). However, our understanding of the expected performance gains and principles when merging any two models remains limited. In this work, we introduce model kinship, the degree of similarity or relatedness between LLMs, analogous to biological evolution. With comprehensive empirical analysis, we find that there is a certain relationship between model kinship and the performance gains after model merging, which can help guide our selection of candidate models. Inspired by this, we propose a new model merging strategy: Top-k Greedy Merging with Model Kinship, which can yield better performance on benchmark datasets. Specifically, we discover that using model kinship as a criterion can assist us in continuously performing model merging, alleviating the degradation (local optima) in model evolution, whereas model kinship can serve as a guide to escape these traps. Code is available at https://github.com/zjunlp/ModelKinship.
comment: Ongoing work
☆ CMAL: A Novel Cross-Modal Associative Learning Framework for Vision-Language Pre-Training
With the flourishing of social media platforms, vision-language pre-training (VLP) recently has received great attention and many remarkable progresses have been achieved. The success of VLP largely benefits from the information complementation and enhancement between different modalities. However, most of recent studies focus on cross-modal contrastive learning (CMCL) to promote image-text alignment by pulling embeddings of positive sample pairs together while pushing those of negative pairs apart, which ignores the natural asymmetry property between different modalities and requires large-scale image-text corpus to achieve arduous progress. To mitigate this predicament, we propose CMAL, a Cross-Modal Associative Learning framework with anchor points detection and cross-modal associative learning for VLP. Specifically, we first respectively embed visual objects and textual tokens into separate hypersphere spaces to learn intra-modal hidden features, and then design a cross-modal associative prompt layer to perform anchor point masking and swap feature filling for constructing a hybrid cross-modal associative prompt. Afterwards, we exploit a unified semantic encoder to learn their cross-modal interactive features for context adaptation. Finally, we design an associative mapping classification layer to learn potential associative mappings between modalities at anchor points, within which we develop a fresh self-supervised associative mapping classification task to boost CMAL's performance. Experimental results verify the effectiveness of CMAL, showing that it achieves competitive performance against previous CMCL-based methods on four common downstream vision-and-language tasks, with significantly fewer corpus. Especially, CMAL obtains new state-of-the-art results on SNLI-VE and REC (testA).
comment: vision-language pre-training, contrastive learning, cross-modal, associative learning, associative mapping classification
☆ Cocoon: Robust Multi-Modal Perception with Uncertainty-Aware Sensor Fusion
An important paradigm in 3D object detection is the use of multiple modalities to enhance accuracy in both normal and challenging conditions, particularly for long-tail scenarios. To address this, recent studies have explored two directions of adaptive approaches: MoE-based adaptive fusion, which struggles with uncertainties arising from distinct object configurations, and late fusion for output-level adaptive fusion, which relies on separate detection pipelines and limits comprehensive understanding. In this work, we introduce Cocoon, an object- and feature-level uncertainty-aware fusion framework. The key innovation lies in uncertainty quantification for heterogeneous representations, enabling fair comparison across modalities through the introduction of a feature aligner and a learnable surrogate ground truth, termed feature impression. We also define a training objective to ensure that their relationship provides a valid metric for uncertainty quantification. Cocoon consistently outperforms existing static and adaptive methods in both normal and challenging conditions, including those with natural and artificial corruptions. Furthermore, we show the validity and efficacy of our uncertainty metric across diverse datasets.
comment: 23 pages
☆ Rethinking Visual Counterfactual Explanations Through Region Constraint
Visual counterfactual explanations (VCEs) have recently gained immense popularity as a tool for clarifying the decision-making process of image classifiers. This trend is largely motivated by what these explanations promise to deliver -- indicate semantically meaningful factors that change the classifier's decision. However, we argue that current state-of-the-art approaches lack a crucial component -- the region constraint -- whose absence prevents from drawing explicit conclusions, and may even lead to faulty reasoning due to phenomenons like confirmation bias. To address the issue of previous methods, which modify images in a very entangled and widely dispersed manner, we propose region-constrained VCEs (RVCEs), which assume that only a predefined image region can be modified to influence the model's prediction. To effectively sample from this subclass of VCEs, we propose Region-Constrained Counterfactual Schr\"odinger Bridges (RCSB), an adaptation of a tractable subclass of Schr\"odinger Bridges to the problem of conditional inpainting, where the conditioning signal originates from the classifier of interest. In addition to setting a new state-of-the-art by a large margin, we extend RCSB to allow for exact counterfactual reasoning, where the predefined region contains only the factor of interest, and incorporating the user to actively interact with the RVCE by predefining the regions manually.
comment: Preprint
☆ From Lab to Pocket: A Novel Continual Learning-based Mobile Application for Screening COVID-19
Artificial intelligence (AI) has emerged as a promising tool for predicting COVID-19 from medical images. In this paper, we propose a novel continual learning-based approach and present the design and implementation of a mobile application for screening COVID-19. Our approach demonstrates the ability to adapt to evolving datasets, including data collected from different locations or hospitals, varying virus strains, and diverse clinical presentations, without retraining from scratch. We have evaluated state-of-the-art continual learning methods for detecting COVID-19 from chest X-rays and selected the best-performing model for our mobile app. We evaluated various deep learning architectures to select the best-performing one as a foundation model for continual learning. Both regularization and memory-based methods for continual learning were tested, using different memory sizes to develop the optimal continual learning model for our app. DenseNet161 emerged as the best foundation model with 96.87\% accuracy, and Learning without Forgetting (LwF) was the top continual learning method with an overall performance of 71.99\%. The mobile app design considers both patient and doctor perspectives. It incorporates the continual learning DenseNet161 LwF model on a cloud server, enabling the model to learn from new instances of chest X-rays and their classifications as they are submitted. The app is designed, implemented, and evaluated to ensure it provides an efficient tool for COVID-19 screening. The app is available to download from https://github.com/DannyFGitHub/COVID-19PneumoCheckApp.
comment: 31 pages
☆ Self-DenseMobileNet: A Robust Framework for Lung Nodule Classification using Self-ONN and Stacking-based Meta-Classifier
In this study, we propose a novel and robust framework, Self-DenseMobileNet, designed to enhance the classification of nodules and non-nodules in chest radiographs (CXRs). Our approach integrates advanced image standardization and enhancement techniques to optimize the input quality, thereby improving classification accuracy. To enhance predictive accuracy and leverage the strengths of multiple models, the prediction probabilities from Self-DenseMobileNet were transformed into tabular data and used to train eight classical machine learning (ML) models; the top three performers were then combined via a stacking algorithm, creating a robust meta-classifier that integrates their collective insights for superior classification performance. To enhance the interpretability of our results, we employed class activation mapping (CAM) to visualize the decision-making process of the best-performing model. Our proposed framework demonstrated remarkable performance on internal validation data, achieving an accuracy of 99.28\% using a Meta-Random Forest Classifier. When tested on an external dataset, the framework maintained strong generalizability with an accuracy of 89.40\%. These results highlight a significant improvement in the classification of CXRs with lung nodules.
comment: 31 pages
☆ FTII-Bench: A Comprehensive Multimodal Benchmark for Flow Text with Image Insertion
Benefiting from the revolutionary advances in large language models (LLMs) and foundational vision models, large vision-language models (LVLMs) have also made significant progress. However, current benchmarks focus on tasks that evaluating only a single aspect of LVLM capabilities (e.g., recognition, detection, understanding). These tasks fail to fully demonstrate LVLMs' potential in complex application scenarios. To comprehensively assess the performance of existing LVLMs, we propose a more challenging task called the Flow Text with Image Insertion task (FTII). This task requires LVLMs to simultaneously possess outstanding abilities in image comprehension, instruction understanding, and long-text interpretation. Specifically, given several text paragraphs and a set of candidate images, as the text paragraphs accumulate, the LVLMs are required to select the most suitable image from the candidates to insert after the corresponding paragraph. Constructing a benchmark for such a task is highly challenging, particularly in determining the sequence of flowing text and images. To address this challenge, we turn to professional news reports, which naturally contain a gold standard for image-text sequences. Based on this, we introduce the Flow Text with Image Insertion Benchmark (FTII-Bench), which includes 318 high-quality Chinese image-text news articles and 307 high-quality English image-text news articles, covering 10 different news domains. Using these 625 high-quality articles, we construct problems of two different types with multiple levels of difficulty. Furthermore, we establish two different evaluation pipelines based on the CLIP model and existing LVLMs. We evaluate 9 open-source and 2 closed-source LVLMs as well as 2 CLIP-based models. Results indicate that even the most advanced models (e.g., GPT-4o) face significant challenges when tackling the FTII task.
comment: Work in progress. 9 pages, 3 figures
☆ Adaptive Prompt Learning with SAM for Few-shot Scanning Probe Microscope Image Segmentation
The Segment Anything Model (SAM) has demonstrated strong performance in image segmentation of natural scene images. However, its effectiveness diminishes markedly when applied to specific scientific domains, such as Scanning Probe Microscope (SPM) images. This decline in accuracy can be attributed to the distinct data distribution and limited availability of the data inherent in the scientific images. On the other hand, the acquisition of adequate SPM datasets is both time-intensive and laborious as well as skill-dependent. To address these challenges, we propose an Adaptive Prompt Learning with SAM (APL-SAM) framework tailored for few-shot SPM image segmentation. Our approach incorporates two key innovations to enhance SAM: 1) An Adaptive Prompt Learning module leverages few-shot embeddings derived from limited support set to learn adaptively central representatives, serving as visual prompts. This innovation eliminates the need for time-consuming online user interactions for providing prompts, such as exhaustively marking points and bounding boxes slice by slice; 2) A multi-source, multi-level mask decoder specifically designed for few-shot SPM image segmentation is introduced, which can effectively capture the correspondence between the support and query images. To facilitate comprehensive training and evaluation, we introduce a new dataset, SPM-Seg, curated for SPM image segmentation. Extensive experiments on this dataset reveal that the proposed APL-SAM framework significantly outperforms the original SAM, achieving over a 30% improvement in terms of Dice Similarity Coefficient with only one-shot guidance. Moreover, APL-SAM surpasses state-of-the-art few-shot segmentation methods and even fully supervised approaches in performance. Code and dataset used in this study will be made available upon acceptance.
comment: 10 pages, 7 figures
☆ Development of Image Collection Method Using YOLO and Siamese Network
As we enter the era of big data, collecting high-quality data is very important. However, collecting data by humans is not only very time-consuming but also expensive. Therefore, many scientists have devised various methods to collect data using computers. Among them, there is a method called web crawling, but the authors found that the crawling method has a problem in that unintended data is collected along with the user. The authors found that this can be filtered using the object recognition model YOLOv10. However, there are cases where data that is not properly filtered remains. Here, image reclassification was performed by additionally utilizing the distance output from the Siamese network, and higher performance was recorded than other classification models. (average \_f1 score YOLO+MobileNet 0.678->YOLO+SiameseNet 0.772)) The user can specify a distance threshold to adjust the balance between data deficiency and noise-robustness. The authors also found that the Siamese network can achieve higher performance with fewer resources because the cropped images are used for object recognition when processing images in the Siamese network. (Class 20 mean-based f1 score, non-crop+Siamese(MobileNetV3-Small) 80.94 -> crop preprocessing+Siamese(MobileNetV3-Small) 82.31) In this way, the image retrieval system that utilizes two consecutive models to reduce errors can save users' time and effort, and build better quality data faster and with fewer resources than before.
comment: 15 pages, 13 figures, 2 tables
☆ One Step Diffusion via Shortcut Models
Diffusion models and flow-matching models have enabled generating diverse and realistic images by learning to transfer noise to data. However, sampling from these models involves iterative denoising over many neural network passes, making generation slow and expensive. Previous approaches for speeding up sampling require complex training regimes, such as multiple training phases, multiple networks, or fragile scheduling. We introduce shortcut models, a family of generative models that use a single network and training phase to produce high-quality samples in a single or multiple sampling steps. Shortcut models condition the network not only on the current noise level but also on the desired step size, allowing the model to skip ahead in the generation process. Across a wide range of sampling step budgets, shortcut models consistently produce higher quality samples than previous approaches, such as consistency models and reflow. Compared to distillation, shortcut models reduce complexity to a single network and training phase and additionally allow varying step budgets at inference time.
☆ Evaluating Utility of Memory Efficient Medical Image Generation: A Study on Lung Nodule Segmentation
The scarcity of publicly available medical imaging data limits the development of effective AI models. This work proposes a memory-efficient patch-wise denoising diffusion probabilistic model (DDPM) for generating synthetic medical images, focusing on CT scans with lung nodules. Our approach generates high-utility synthetic images with nodule segmentation while efficiently managing memory constraints, enabling the creation of training datasets. We evaluate the method in two scenarios: training a segmentation model exclusively on synthetic data, and augmenting real-world training data with synthetic images. In the first case, models trained solely on synthetic data achieve Dice scores comparable to those trained on real-world data benchmarks. In the second case, augmenting real-world data with synthetic images significantly improves segmentation performance. The generated images demonstrate their potential to enhance medical image datasets in scenarios with limited real-world data.
☆ Shaping a Stabilized Video by Mitigating Unintended Changes for Concept-Augmented Video Editing
Text-driven video editing utilizing generative diffusion models has garnered significant attention due to their potential applications. However, existing approaches are constrained by the limited word embeddings provided in pre-training, which hinders nuanced editing targeting open concepts with specific attributes. Directly altering the keywords in target prompts often results in unintended disruptions to the attention mechanisms. To achieve more flexible editing easily, this work proposes an improved concept-augmented video editing approach that generates diverse and stable target videos flexibly by devising abstract conceptual pairs. Specifically, the framework involves concept-augmented textual inversion and a dual prior supervision mechanism. The former enables plug-and-play guidance of stable diffusion for video editing, effectively capturing target attributes for more stylized results. The dual prior supervision mechanism significantly enhances video stability and fidelity. Comprehensive evaluations demonstrate that our approach generates more stable and lifelike videos, outperforming state-of-the-art methods.
☆ MambaPainter: Neural Stroke-Based Rendering in a Single Step SIGGRAPH
Stroke-based rendering aims to reconstruct an input image into an oil painting style by predicting brush stroke sequences. Conventional methods perform this prediction stroke-by-stroke or require multiple inference steps due to the limitations of a predictable number of strokes. This procedure leads to inefficient translation speed, limiting their practicality. In this study, we propose MambaPainter, capable of predicting a sequence of over 100 brush strokes in a single inference step, resulting in rapid translation. We achieve this sequence prediction by incorporating the selective state-space model. Additionally, we introduce a simple extension to patch-based rendering, which we use to translate high-resolution images, improving the visual quality with a minimal increase in computational cost. Experimental results demonstrate that MambaPainter can efficiently translate inputs to oil painting-style images compared to state-of-the-art methods. The codes are available at https://github.com/STomoya/MambaPainter.
comment: Accepted to SIGGRAPH Asia 2024 posters
☆ QueensCAMP: an RGB-D dataset for robust Visual SLAM
Visual Simultaneous Localization and Mapping (VSLAM) is a fundamental technology for robotics applications. While VSLAM research has achieved significant advancements, its robustness under challenging situations, such as poor lighting, dynamic environments, motion blur, and sensor failures, remains a challenging issue. To address these challenges, we introduce a novel RGB-D dataset designed for evaluating the robustness of VSLAM systems. The dataset comprises real-world indoor scenes with dynamic objects, motion blur, and varying illumination, as well as emulated camera failures, including lens dirt, condensation, underexposure, and overexposure. Additionally, we offer open-source scripts for injecting camera failures into any images, enabling further customization by the research community. Our experiments demonstrate that ORB-SLAM2, a traditional VSLAM algorithm, and TartanVO, a Deep Learning-based VO algorithm, can experience performance degradation under these challenging conditions. Therefore, this dataset and the camera failure open-source tools provide a valuable resource for developing more robust VSLAM systems capable of handling real-world challenges.
comment: 6 pages
☆ DH-VTON: Deep Text-Driven Virtual Try-On via Hybrid Attention Learning ICASSP2025
Virtual Try-ON (VTON) aims to synthesis specific person images dressed in given garments, which recently receives numerous attention in online shopping scenarios. Currently, the core challenges of the VTON task mainly lie in the fine-grained semantic extraction (i.e.,deep semantics) of the given reference garments during depth estimation and effective texture preservation when the garments are synthesized and warped onto human body. To cope with these issues, we propose DH-VTON, a deep text-driven virtual try-on model featuring a special hybrid attention learning strategy and deep garment semantic preservation module. By standing on the shoulder of a well-built pre-trained paint-by-example (abbr. PBE) approach, we present our DH-VTON pipeline in this work. Specifically, to extract the deep semantics of the garments, we first introduce InternViT-6B as fine-grained feature learner, which can be trained to align with the large-scale intrinsic knowledge with deep text semantics (e.g.,"neckline" or "girdle") to make up for the deficiency of the commonly adopted CLIP encoder. Based on this, to enhance the customized dressing abilities, we further introduce Garment-Feature ControlNet Plus (abbr. GFC+) module and propose to leverage a fresh hybrid attention strategy for training, which can adaptively integrate fine-grained characteristics of the garments into the different layers of the VTON model, so as to achieve multi-scale features preservation effects. Extensive experiments on several representative datasets demonstrate that our method outperforms previous diffusion-based and GAN-based approaches, showing competitive performance in preserving garment details and generating authentic human images.
comment: 5 pages, 6 figures, ICASSP2025
☆ Stabilize the Latent Space for Image Autoregressive Modeling: A Unified Perspective NeurIPS 2024
Latent-based image generative models, such as Latent Diffusion Models (LDMs) and Mask Image Models (MIMs), have achieved notable success in image generation tasks. These models typically leverage reconstructive autoencoders like VQGAN or VAE to encode pixels into a more compact latent space and learn the data distribution in the latent space instead of directly from pixels. However, this practice raises a pertinent question: Is it truly the optimal choice? In response, we begin with an intriguing observation: despite sharing the same latent space, autoregressive models significantly lag behind LDMs and MIMs in image generation. This finding contrasts sharply with the field of NLP, where the autoregressive model GPT has established a commanding presence. To address this discrepancy, we introduce a unified perspective on the relationship between latent space and generative models, emphasizing the stability of latent space in image generative modeling. Furthermore, we propose a simple but effective discrete image tokenizer to stabilize the latent space for image generative modeling. Experimental results show that image autoregressive modeling with our tokenizer (DiGIT) benefits both image understanding and image generation with the next token prediction principle, which is inherently straightforward for GPT models but challenging for other generative models. Remarkably, for the first time, a GPT-style autoregressive model for images outperforms LDMs, which also exhibits substantial improvement akin to GPT when scaling up model size. Our findings underscore the potential of an optimized latent space and the integration of discrete tokenization in advancing the capabilities of image generative models. The code is available at \url{https://github.com/DAMO-NLP-SG/DiGIT}.
comment: Accepted at NeurIPS 2024
☆ Synthetic Augmentation for Anatomical Landmark Localization using DDPMs
Deep learning techniques for anatomical landmark localization (ALL) have shown great success, but their reliance on large annotated datasets remains a problem due to the tedious and costly nature of medical data acquisition and annotation. While traditional data augmentation, variational autoencoders (VAEs), and generative adversarial networks (GANs) have already been used to synthetically expand medical datasets, diffusion-based generative models have recently started to gain attention for their ability to generate high-quality synthetic images. In this study, we explore the use of denoising diffusion probabilistic models (DDPMs) for generating medical images and their corresponding heatmaps of landmarks to enhance the training of a supervised deep learning model for ALL. Our novel approach involves a DDPM with a 2-channel input, incorporating both the original medical image and its heatmap of annotated landmarks. We also propose a novel way to assess the quality of the generated images using a Markov Random Field (MRF) model for landmark matching and a Statistical Shape Model (SSM) to check landmark plausibility, before we evaluate the DDPM-augmented dataset in the context of an ALL task involving hand X-Rays.
☆ Mind the Gap Between Prototypes and Images in Cross-domain Finetuning
In cross-domain few-shot classification (CFC), recent works mainly focus on adapting a simple transformation head on top of a frozen pre-trained backbone with few labeled data to project embeddings into a task-specific metric space where classification can be performed by measuring similarities between image instance and prototype representations. Technically, an assumption implicitly adopted in such a framework is that the prototype and image instance embeddings share the same representation transformation. However, in this paper, we find that there naturally exists a gap, which resembles the modality gap, between the prototype and image instance embeddings extracted from the frozen pre-trained backbone, and simply applying the same transformation during the adaptation phase constrains exploring the optimal representations and shrinks the gap between prototype and image representations. To solve this problem, we propose a simple yet effective method, contrastive prototype-image adaptation (CoPA), to adapt different transformations respectively for prototypes and images similarly to CLIP by treating prototypes as text prompts. Extensive experiments on Meta-Dataset demonstrate that CoPA achieves the state-of-the-art performance more efficiently. Meanwhile, further analyses also indicate that CoPA can learn better representation clusters, enlarge the gap, and achieve minimal validation loss at the enlarged gap.
☆ A Primal-dual algorithm for image reconstruction with ICNNs
We address the optimization problem in a data-driven variational reconstruction framework, where the regularizer is parameterized by an input-convex neural network (ICNN). While gradient-based methods are commonly used to solve such problems, they struggle to effectively handle non-smoothness which often leads to slow convergence. Moreover, the nested structure of the neural network complicates the application of standard non-smooth optimization techniques, such as proximal algorithms. To overcome these challenges, we reformulate the problem and eliminate the network's nested structure. By relating this reformulation to epigraphical projections of the activation functions, we transform the problem into a convex optimization problem that can be efficiently solved using a primal-dual algorithm. We also prove that this reformulation is equivalent to the original variational problem. Through experiments on several imaging tasks, we demonstrate that the proposed approach outperforms subgradient methods in terms of both speed and stability.
☆ Attention-Guided Perturbation for Consistency Regularization in Semi-Supervised Medical Image Segmentation
Medical image segmentation is a pivotal step in diagnostic and therapeutic processes. However, the acquisition of high-quality annotated data is often constrained by scarcity and cost. Semi-supervised learning offers a promising approach to enhance model performance by using unlabeled data. While consistency regularization is a prevalent method in semi-supervised image segmentation, there is a dearth of research on perturbation strategies tailored for semi-supervised medical image segmentation tasks. This paper introduces an attention-guided perturbation strategy for semi-supervised consistency regularization in the context of medical image segmentation. We add the perturbation based on the attention from the model in the image and feature level to achieve consistency regularization. The method is adept at accommodating the intricate structures and high-dimensional semantics inherent in medical images, thereby enhancing the performance of semi-supervised segmentation tasks. Our method achieved state-of-the-art results on benchmark datasets, including a 90.4\% Dice score on the ACDC dataset in the 7-case scenario.
☆ Triplet: Triangle Patchlet for Mesh-Based Inverse Rendering and Scene Parameters Approximation
Recent advancements in Radiance Fields have significantly improved novel-view synthesis. However, in many real-world applications, the more advanced challenge lies in inverse rendering, which seeks to derive the physical properties of a scene, including light, geometry, textures, and materials. Meshes, as a traditional representation adopted by many simulation pipeline, however, still show limited influence in radiance field for inverse rendering. This paper introduces a novel framework called Triangle Patchlet (abbr. Triplet), a mesh-based representation, to comprehensively approximate these scene parameters. We begin by assembling Triplets with either randomly generated points or sparse points obtained from camera calibration where all faces are treated as an independent element. Next, we simulate the physical interaction of light and optimize the scene parameters using traditional graphics rendering techniques like rasterization and ray tracing, accompanying with density control and propagation. An iterative mesh extracting process is also suggested, where we continue to optimize on geometry and materials with graph-based operation. We also introduce several regulation terms to enable better generalization of materials property. Our framework could precisely estimate the light, materials and geometry with mesh without prior of light, materials and geometry in a unified framework. Experiments demonstrate that our approach can achieve state-of-the-art visual quality while reconstructing high-quality geometry and accurate material properties.
comment: https://github.com/RANDO11199/Triplet
☆ AdaCropFollow: Self-Supervised Online Adaptation for Visual Under-Canopy Navigation
Under-canopy agricultural robots can enable various applications like precise monitoring, spraying, weeding, and plant manipulation tasks throughout the growing season. Autonomous navigation under the canopy is challenging due to the degradation in accuracy of RTK-GPS and the large variability in the visual appearance of the scene over time. In prior work, we developed a supervised learning-based perception system with semantic keypoint representation and deployed this in various field conditions. A large number of failures of this system can be attributed to the inability of the perception model to adapt to the domain shift encountered during deployment. In this paper, we propose a self-supervised online adaptation method for adapting the semantic keypoint representation using a visual foundational model, geometric prior, and pseudo labeling. Our preliminary experiments show that with minimal data and fine-tuning of parameters, the keypoint prediction model trained with labels on the source domain can be adapted in a self-supervised manner to various challenging target domains onboard the robot computer using our method. This can enable fully autonomous row-following capability in under-canopy robots across fields and crops without requiring human intervention.
☆ Beyond Coarse-Grained Matching in Video-Text Retrieval ACCV 2024
Video-text retrieval has seen significant advancements, yet the ability of models to discern subtle differences in captions still requires verification. In this paper, we introduce a new approach for fine-grained evaluation. Our approach can be applied to existing datasets by automatically generating hard negative test captions with subtle single-word variations across nouns, verbs, adjectives, adverbs, and prepositions. We perform comprehensive experiments using four state-of-the-art models across two standard benchmarks (MSR-VTT and VATEX) and two specially curated datasets enriched with detailed descriptions (VLN-UVO and VLN-OOPS), resulting in a number of novel insights: 1) our analyses show that the current evaluation benchmarks fall short in detecting a model's ability to perceive subtle single-word differences, 2) our fine-grained evaluation highlights the difficulty models face in distinguishing such subtle variations. To enhance fine-grained understanding, we propose a new baseline that can be easily combined with current methods. Experiments on our fine-grained evaluations demonstrate that this approach enhances a model's ability to understand fine-grained differences.
comment: Accepted to ACCV 2024
☆ De-Identification of Medical Imaging Data: A Comprehensive Tool for Ensuring Patient Privacy
Medical data employed in research frequently comprises sensitive patient health information (PHI), which is subject to rigorous legal frameworks such as the General Data Protection Regulation (GDPR) or the Health Insurance Portability and Accountability Act (HIPAA). Consequently, these types of data must be pseudonymized prior to utilisation, which presents a significant challenge for many researchers. Given the vast array of medical data, it is necessary to employ a variety of de-identification techniques. To facilitate the anonymization process for medical imaging data, we have developed an open-source tool that can be used to de-identify DICOM magnetic resonance images, computer tomography images, whole slide images and magnetic resonance twix raw data. Furthermore, the implementation of a neural network enables the removal of text within the images. The proposed tool automates an elaborate anonymization pipeline for multiple types of inputs, reducing the need for additional tools used for de-identification of imaging data. We make our code publicly available at https://github.com/code-lukas/medical_image_deidentification.
☆ Feature Augmentation for Self-supervised Contrastive Learning: A Closer Look IJCNN 2024
Self-supervised contrastive learning heavily relies on the view variance brought by data augmentation, so that it can learn a view-invariant pre-trained representation. Beyond increasing the view variance for contrast, this work focuses on improving the diversity of training data, to improve the generalization and robustness of the pre-trained models. To this end, we propose a unified framework to conduct data augmentation in the feature space, known as feature augmentation. This strategy is domain-agnostic, which augments similar features to the original ones and thus improves the data diversity. We perform a systematic investigation of various feature augmentation architectures, the gradient-flow skill, and the relationship between feature augmentation and traditional data augmentation. Our study reveals some practical principles for feature augmentation in self-contrastive learning. By integrating feature augmentation on the instance discrimination or the instance similarity paradigm, we consistently improve the performance of pre-trained feature learning and gain better generalization over the downstream image classification and object detection task.
comment: IJCNN 2024
☆ Real-time Stereo-based 3D Object Detection for Streaming Perception NeurIPS2024
The ability to promptly respond to environmental changes is crucial for the perception system of autonomous driving. Recently, a new task called streaming perception was proposed. It jointly evaluate the latency and accuracy into a single metric for video online perception. In this work, we introduce StreamDSGN, the first real-time stereo-based 3D object detection framework designed for streaming perception. StreamDSGN is an end-to-end framework that directly predicts the 3D properties of objects in the next moment by leveraging historical information, thereby alleviating the accuracy degradation of streaming perception. Further, StreamDSGN applies three strategies to enhance the perception accuracy: (1) A feature-flow-based fusion method, which generates a pseudo-next feature at the current moment to address the misalignment issue between feature and ground truth. (2) An extra regression loss for explicit supervision of object motion consistency in consecutive frames. (3) A large kernel backbone with a large receptive field for effectively capturing long-range spatial contextual features caused by changes in object positions. Experiments on the KITTI Tracking dataset show that, compared with the strong baseline, StreamDSGN significantly improves the streaming average precision by up to 4.33%. Our code is available at https://github.com/weiyangdaren/streamDSGN-pytorch.
comment: Streaming Perception, 3D Object Detection, NeurIPS2024 poster
☆ HumanEval-V: Evaluating Visual Understanding and Reasoning Abilities of Large Multimodal Models Through Coding Tasks
Coding tasks have been valuable for evaluating Large Language Models (LLMs), as they demand the comprehension of high-level instructions, complex reasoning, and the implementation of functional programs -- core capabilities for advancing Artificial General Intelligence. Despite the progress in Large Multimodal Models (LMMs), which extend LLMs with visual perception and understanding capabilities, there remains a notable lack of coding benchmarks that rigorously assess these models, particularly in tasks that emphasize visual reasoning. To address this gap, we introduce HumanEval-V, a novel and lightweight benchmark specifically designed to evaluate LMMs' visual understanding and reasoning capabilities through code generation. HumanEval-V includes 108 carefully crafted, entry-level Python coding tasks derived from platforms like CodeForces and Stack Overflow. Each task is adapted by modifying the context and algorithmic patterns of the original problems, with visual elements redrawn to ensure distinction from the source, preventing potential data leakage. LMMs are required to complete the code solution based on the provided visual context and a predefined Python function signature outlining the task requirements. Every task is equipped with meticulously handcrafted test cases to ensure a thorough and reliable evaluation of model-generated solutions. We evaluate 19 state-of-the-art LMMs using HumanEval-V, uncovering significant challenges. Proprietary models like GPT-4o achieve only 13% pass@1 and 36.4% pass@10, while open-weight models with 70B parameters score below 4% pass@1. Ablation studies further reveal the limitations of current LMMs in vision reasoning and coding capabilities. These results underscore key areas for future research to enhance LMMs' capabilities. We have open-sourced our code and benchmark at https://github.com/HumanEval-V/HumanEval-V-Benchmark.
comment: homepage https://humaneval-v.github.io/
☆ Stylistic Multi-Task Analysis of Ukiyo-e Woodblock Prints
In this work we present a large-scale dataset of \textit{Ukiyo-e} woodblock prints. Unlike previous works and datasets in the artistic domain that primarily focus on western art, this paper explores this pre-modern Japanese art form with the aim of broadening the scope for stylistic analysis and to provide a benchmark to evaluate a variety of art focused Computer Vision approaches. Our dataset consists of over $175.000$ prints with corresponding metadata (\eg artist, era, and creation date) from the 17th century to present day. By approaching stylistic analysis as a Multi-Task problem we aim to more efficiently utilize the available metadata, and learn more general representations of style. We show results for well-known baselines and state-of-the-art multi-task learning frameworks to enable future comparison, and to encourage stylistic analysis on this artistic domain.
☆ GAN Based Top-Down View Synthesis in Reinforcement Learning Environments
Human actions are based on the mental perception of the environment. Even when all the aspects of an environment are not visible, humans have an internal mental model that can generalize the partially visible scenes to fully constructed and connected views. This internal mental model uses learned abstract representations of spatial and temporal aspects of the environments encountered in the past. Artificial agents in reinforcement learning environments also benefit by learning a representation of the environment from experience. It provides the agent with viewpoints that are not directly visible to it, helping it make better policy decisions. It can also be used to predict the future state of the environment. This project explores learning the top-down view of an RL environment based on the artificial agent's first-person view observations with a generative adversarial network(GAN). The top-down view is useful as it provides a complete overview of the environment by building a map of the entire environment. It provides information about the objects' dimensions and shapes along with their relative positions with one another. Initially, when only a partial observation of the environment is visible to the agent, only a partial top-down view is generated. As the agent explores the environment through a set of actions, the generated top-down view becomes complete. This generated top-down view can assist the agent in deducing better policy decisions. The focus of the project is to learn the top-down view of an RL environment. It doesn't deal with any Reinforcement Learning task.
☆ Context-Infused Visual Grounding for Art
Many artwork collections contain textual attributes that provide rich and contextualised descriptions of artworks. Visual grounding offers the potential for localising subjects within these descriptions on images, however, existing approaches are trained on natural images and generalise poorly to art. In this paper, we present CIGAr (Context-Infused GroundingDINO for Art), a visual grounding approach which utilises the artwork descriptions during training as context, thereby enabling visual grounding on art. In addition, we present a new dataset, Ukiyo-eVG, with manually annotated phrase-grounding annotations, and we set a new state-of-the-art for object detection on two artwork datasets.
☆ Towards Flexible and Efficient Diffusion Low Light Enhancer
Diffusion-based Low-Light Image Enhancement (LLIE) has demonstrated significant success in improving the visibility of low-light images. However, the substantial computational burden introduced by the iterative sampling process remains a major concern. Current acceleration methods, whether training-based or training-free, often lead to significant performance degradation. As a result, to achieve an efficient student model with performance comparable to that of existing multi-step teacher model, it is usually necessary to retrain a more capable teacher model. This approach introduces inflexibility, as it requires additional training to enhance the teacher's performance. To address these challenges, we propose \textbf{Re}flectance-aware \textbf{D}iffusion with \textbf{Di}stilled \textbf{T}rajectory (\textbf{ReDDiT}), a step distillation framework specifically designed for LLIE. ReDDiT trains a student model to replicate the teacher's trajectory in fewer steps while also possessing the ability to surpass the teacher's performance. Specifically, we first introduce a trajectory decoder from the teacher model to provide guidance. Subsequently, a reflectance-aware trajectory refinement module is incorporated into the distillation process to enable more deterministic guidance from the teacher model. Our framework achieves comparable performance to previous diffusion-based methods with redundant steps in just 2 steps while establishing new state-of-the-art (SOTA) results with 8 or 4 steps. Comprehensive experimental evaluations on 10 benchmark datasets validate the effectiveness of our method, consistently outperforming existing SOTA methods.
comment: 7 pages
☆ TAS: Distilling Arbitrary Teacher and Student via a Hybrid Assistant
Most knowledge distillation (KD) methodologies predominantly focus on teacher-student pairs with similar architectures, such as both being convolutional neural networks (CNNs). However, the potential and flexibility of KD can be greatly improved by expanding it to novel Cross-Architecture KD (CAKD), where the knowledge of homogeneous and heterogeneous teachers can be transferred flexibly to a given student. The primary challenge in CAKD lies in the substantial feature gaps between heterogeneous models, originating from the distinction of their inherent inductive biases and module functions. To this end, we introduce an assistant model as a bridge to facilitate smooth feature knowledge transfer between heterogeneous teachers and students. More importantly, within our proposed design principle, the assistant model combines the advantages of cross-architecture inductive biases and module functions by merging convolution and attention modules derived from both student and teacher module functions. Furthermore, we observe that heterogeneous features exhibit diverse spatial distributions in CAKD, hindering the effectiveness of conventional pixel-wise mean squared error (MSE) loss. Therefore, we leverage a spatial-agnostic InfoNCE loss to align features after spatial smoothing, thereby improving the feature alignments in CAKD. Our proposed method is evaluated across some homogeneous model pairs and arbitrary heterogeneous combinations of CNNs, ViTs, and MLPs, achieving state-of-the-art performance for distilled models with a maximum gain of 11.47% on CIFAR-100 and 3.67% on ImageNet-1K. Our code and models will be released.
comment: 18 pages, 6 figures, and 12 tables
☆ ARIC: An Activity Recognition Dataset in Classroom Surveillance Images
The application of activity recognition in the ``AI + Education" field is gaining increasing attention. However, current work mainly focuses on the recognition of activities in manually captured videos and a limited number of activity types, with little attention given to recognizing activities in surveillance images from real classrooms. Activity recognition in classroom surveillance images faces multiple challenges, such as class imbalance and high activity similarity. To address this gap, we constructed a novel multimodal dataset focused on classroom surveillance image activity recognition called ARIC (Activity Recognition In Classroom). The ARIC dataset has advantages of multiple perspectives, 32 activity categories, three modalities, and real-world classroom scenarios. In addition to the general activity recognition tasks, we also provide settings for continual learning and few-shot continual learning. We hope that the ARIC dataset can act as a facilitator for future analysis and research for open teaching scenarios. You can download preliminary data from https://ivipclab.github.io/publication_ARIC/ARIC.
comment: arXiv admin note: text overlap with arXiv:2409.03354
☆ MC-Bench: A Benchmark for Multi-Context Visual Grounding in the Era of MLLMs
While multimodal large language models (MLLMs) have demonstrated extraordinary vision-language understanding capabilities and shown potential to serve as general-purpose assistants, their abilities to solve instance-level visual-language problems beyond a single image warrant further exploration. In order to assess these unproven abilities of MLLMs, this paper proposes a new visual grounding task called multi-context visual grounding, which aims to localize instances of interest across multiple images based on open-ended text prompts. To facilitate this research, we meticulously construct a new dataset MC-Bench for benchmarking the visual grounding capabilities of MLLMs. MC-Bench features 2K high-quality and manually annotated samples, consisting of instance-level labeled image pairs and corresponding text prompts that indicate the target instances in the images. In total, there are three distinct styles of text prompts, covering 20 practical skills. We benchmark over 20 state-of-the-art MLLMs and foundation models with potential multi-context visual grounding capabilities. Our evaluation reveals a non-trivial performance gap between existing MLLMs and humans across all metrics. We also observe that existing MLLMs typically outperform foundation models without LLMs only on image-level metrics, and the specialist MLLMs trained on single images often struggle to generalize to multi-image scenarios. Moreover, a simple stepwise baseline integrating advanced MLLM and a detector can significantly surpass prior end-to-end MLLMs. We hope our MC-Bench and empirical findings can encourage the research community to further explore and enhance the untapped potentials of MLLMs in instance-level tasks, particularly in multi-image contexts. Project page: https://xuyunqiu.github.io/MC-Bench/.
☆ Improved Anomaly Detection through Conditional Latent Space VAE Ensembles
We propose a novel Conditional Latent space Variational Autoencoder (CL-VAE) to perform improved pre-processing for anomaly detection on data with known inlier classes and unknown outlier classes. This proposed variational autoencoder (VAE) improves latent space separation by conditioning on information within the data. The method fits a unique prior distribution to each class in the dataset, effectively expanding the classic prior distribution for VAEs to include a Gaussian mixture model. An ensemble of these VAEs are merged in the latent spaces to form a group consensus that greatly improves the accuracy of anomaly detection across data sets. Our approach is compared against the capabilities of a typical VAE, a CNN, and a PCA, with regards AUC for anomaly detection. The proposed model shows increased accuracy in anomaly detection, achieving an AUC of 97.4% on the MNIST dataset compared to 95.7% for the second best model. In addition, the CL-VAE shows increased benefits from ensembling, a more interpretable latent space, and an increased ability to learn patterns in complex data with limited model sizes.
comment: 13 pages of main article, 19 pages including references and appendix, 4 figures
☆ PAPL-SLAM: Principal Axis-Anchored Monocular Point-Line SLAM
In point-line SLAM systems, the utilization of line structural information and the optimization of lines are two significant problems. The former is usually addressed through structural regularities, while the latter typically involves using minimal parameter representations of lines in optimization. However, separating these two steps leads to the loss of constraint information to each other. We anchor lines with similar directions to a principal axis and optimize them with $n+2$ parameters for $n$ lines, solving both problems together. Our method considers scene structural information, which can be easily extended to different world hypotheses while significantly reducing the number of line parameters to be optimized, enabling rapid and accurate mapping and tracking. To further enhance the system's robustness and avoid mismatch, we have modeled the line-axis probabilistic data association and provided the algorithm for axis creation, updating, and optimization. Additionally, considering that most real-world scenes conform to the Atlanta World hypothesis, we provide a structural line detection strategy based on vertical priors and vanishing points. Experimental results and ablation studies on various indoor and outdoor datasets demonstrate the effectiveness of our system.
comment: 8 pages, 4 figures
☆ FaceChain-FACT: Face Adapter with Decoupled Training for Identity-preserved Personalization
In the field of human-centric personalized image generation, the adapter-based method obtains the ability to customize and generate portraits by text-to-image training on facial data. This allows for identity-preserved personalization without additional fine-tuning in inference. Although there are improvements in efficiency and fidelity, there is often a significant performance decrease in test following ability, controllability, and diversity of generated faces compared to the base model. In this paper, we analyze that the performance degradation is attributed to the failure to decouple identity features from other attributes during extraction, as well as the failure to decouple the portrait generation training from the overall generation task. To address these issues, we propose the Face Adapter with deCoupled Training (FACT) framework, focusing on both model architecture and training strategy. To decouple identity features from others, we leverage a transformer-based face-export encoder and harness fine-grained identity features. To decouple the portrait generation training, we propose Face Adapting Increment Regularization~(FAIR), which effectively constrains the effect of face adapters on the facial region, preserving the generative ability of the base model. Additionally, we incorporate a face condition drop and shuffle mechanism, combined with curriculum learning, to enhance facial controllability and diversity. As a result, FACT solely learns identity preservation from training data, thereby minimizing the impact on the original text-to-image capabilities of the base model. Extensive experiments show that FACT has both controllability and fidelity in both text-to-image generation and inpainting solutions for portrait generation.
comment: 12 pages, 8 figures
☆ DAT: Improving Adversarial Robustness via Generative Amplitude Mix-up in Frequency Domain
To protect deep neural networks (DNNs) from adversarial attacks, adversarial training (AT) is developed by incorporating adversarial examples (AEs) into model training. Recent studies show that adversarial attacks disproportionately impact the patterns within the phase of the sample's frequency spectrum -- typically containing crucial semantic information -- more than those in the amplitude, resulting in the model's erroneous categorization of AEs. We find that, by mixing the amplitude of training samples' frequency spectrum with those of distractor images for AT, the model can be guided to focus on phase patterns unaffected by adversarial perturbations. As a result, the model's robustness can be improved. Unfortunately, it is still challenging to select appropriate distractor images, which should mix the amplitude without affecting the phase patterns. To this end, in this paper, we propose an optimized Adversarial Amplitude Generator (AAG) to achieve a better tradeoff between improving the model's robustness and retaining phase patterns. Based on this generator, together with an efficient AE production procedure, we design a new Dual Adversarial Training (DAT) strategy. Experiments on various datasets show that our proposed DAT leads to significantly improved robustness against diverse adversarial attacks.
☆ Consistency Calibration: Improving Uncertainty Calibration via Consistency among Perturbed Neighbors
Calibration is crucial in deep learning applications, especially in fields like healthcare and autonomous driving, where accurate confidence estimates are vital for decision-making. However, deep neural networks often suffer from miscalibration, with reliability diagrams and Expected Calibration Error (ECE) being the only standard perspective for evaluating calibration performance. In this paper, we introduce the concept of consistency as an alternative perspective on model calibration, inspired by uncertainty estimation literature in large language models (LLMs). We highlight its advantages over the traditional reliability-based view. Building on this concept, we propose a post-hoc calibration method called Consistency Calibration (CC), which adjusts confidence based on the model's consistency across perturbed inputs. CC is particularly effective in locally uncertainty estimation, as it requires no additional data samples or label information, instead generating input perturbations directly from the source data. Moreover, we show that performing perturbations at the logit level significantly improves computational efficiency. We validate the effectiveness of CC through extensive comparisons with various post-hoc and training-time calibration methods, demonstrating state-of-the-art performance on standard datasets such as CIFAR-10, CIFAR-100, and ImageNet, as well as on long-tailed datasets like ImageNet-LT.
☆ Fool Me Once? Contrasting Textual and Visual Explanations in a Clinical Decision-Support Setting
The growing capabilities of AI models are leading to their wider use, including in safety-critical domains. Explainable AI (XAI) aims to make these models safer to use by making their inference process more transparent. However, current explainability methods are seldom evaluated in the way they are intended to be used: by real-world end users. To address this, we conducted a large-scale user study with 85 healthcare practitioners in the context of human-AI collaborative chest X-ray analysis. We evaluated three types of explanations: visual explanations (saliency maps), natural language explanations, and a combination of both modalities. We specifically examined how different explanation types influence users depending on whether the AI advice and explanations are factually correct. We find that text-based explanations lead to significant over-reliance, which is alleviated by combining them with saliency maps. We also observe that the quality of explanations, that is, how much factually correct information they entail, and how much this aligns with AI correctness, significantly impacts the usefulness of the different explanation types.
☆ Controlled Automatic Task-Specific Synthetic Data Generation for Hallucination Detection
We present a novel approach to automatically generate non-trivial task-specific synthetic datasets for hallucination detection. Our approach features a two-step generation-selection pipeline, using hallucination pattern guidance and a language style alignment during generation. Hallucination pattern guidance leverages the most important task-specific hallucination patterns while language style alignment aligns the style of the synthetic dataset with benchmark text. To obtain robust supervised detectors from synthetic datasets, we also adopt a data mixture strategy to improve performance robustness and generalization. Our results on three datasets show that our generated hallucination text is more closely aligned with non-hallucinated text versus baselines, to train hallucination detectors with better generalization. Our hallucination detectors trained on synthetic datasets outperform in-context-learning (ICL)-based detectors by a large margin of 32%. Our extensive experiments confirm the benefits of our approach with cross-task and cross-generator generalization. Our data-mixture-based training further improves the generalization and robustness of hallucination detection.
☆ Fusion from Decomposition: A Self-Supervised Approach for Image Fusion and Beyond
Image fusion is famous as an alternative solution to generate one high-quality image from multiple images in addition to image restoration from a single degraded image. The essence of image fusion is to integrate complementary information from source images. Existing fusion methods struggle with generalization across various tasks and often require labor-intensive designs, in which it is difficult to identify and extract useful information from source images due to the diverse requirements of each fusion task. Additionally, these methods develop highly specialized features for different downstream applications, hindering the adaptation to new and diverse downstream tasks. To address these limitations, we introduce DeFusion++, a novel framework that leverages self-supervised learning (SSL) to enhance the versatility of feature representation for different image fusion tasks. DeFusion++ captures the image fusion task-friendly representations from large-scale data in a self-supervised way, overcoming the constraints of limited fusion datasets. Specifically, we introduce two innovative pretext tasks: common and unique decomposition (CUD) and masked feature modeling (MFM). CUD decomposes source images into abstract common and unique components, while MFM refines these components into robust fused features. Jointly training of these tasks enables DeFusion++ to produce adaptable representations that can effectively extract useful information from various source images, regardless of the fusion task. The resulting fused representations are also highly adaptable for a wide range of downstream tasks, including image segmentation and object detection. DeFusion++ stands out by producing versatile fused representations that can enhance both the quality of image fusion and the effectiveness of downstream high-level vision tasks, simplifying the process with the elegant fusion framework.
comment: 18page
☆ DaDiff: Domain-aware Diffusion Model for Nighttime UAV Tracking
Domain adaptation is an inspiring solution to the misalignment issue of day/night image features for nighttime UAV tracking. However, the one-step adaptation paradigm is inadequate in addressing the prevalent difficulties posed by low-resolution (LR) objects when viewed from the UAVs at night, owing to the blurry edge contour and limited detail information. Moreover, these approaches struggle to perceive LR objects disturbed by nighttime noise. To address these challenges, this work proposes a novel progressive alignment paradigm, named domain-aware diffusion model (DaDiff), aligning nighttime LR object features to the daytime by virtue of progressive and stable generations. The proposed DaDiff includes an alignment encoder to enhance the detail information of nighttime LR objects, a tracking-oriented layer designed to achieve close collaboration with tracking tasks, and a successive distribution discriminator presented to distinguish different feature distributions at each diffusion timestep successively. Furthermore, an elaborate nighttime UAV tracking benchmark is constructed for LR objects, namely NUT-LR, consisting of 100 annotated sequences. Exhaustive experiments have demonstrated the robustness and feature alignment ability of the proposed DaDiff. The source code and video demo are available at https://github.com/vision4robotics/DaDiff.
☆ LoD-Loc: Aerial Visual Localization using LoD 3D Map with Neural Wireframe Alignment NeurIPS 2024
We propose a new method named LoD-Loc for visual localization in the air. Unlike existing localization algorithms, LoD-Loc does not rely on complex 3D representations and can estimate the pose of an Unmanned Aerial Vehicle (UAV) using a Level-of-Detail (LoD) 3D map. LoD-Loc mainly achieves this goal by aligning the wireframe derived from the LoD projected model with that predicted by the neural network. Specifically, given a coarse pose provided by the UAV sensor, LoD-Loc hierarchically builds a cost volume for uniformly sampled pose hypotheses to describe pose probability distribution and select a pose with maximum probability. Each cost within this volume measures the degree of line alignment between projected and predicted wireframes. LoD-Loc also devises a 6-DoF pose optimization algorithm to refine the previous result with a differentiable Gaussian-Newton method. As no public dataset exists for the studied problem, we collect two datasets with map levels of LoD3.0 and LoD2.0, along with real RGB queries and ground-truth pose annotations. We benchmark our method and demonstrate that LoD-Loc achieves excellent performance, even surpassing current state-of-the-art methods that use textured 3D models for localization. The code and dataset are available at https://victorzoo.github.io/LoD-Loc.github.io/.
comment: Accepted by NeurIPS 2024; for Project page, see https://victorzoo.github.io/LoD-Loc.github.io/
☆ Optimizing YOLOv5s Object Detection through Knowledge Distillation algorithm
This paper explores the application of knowledge distillation technology in target detection tasks, especially the impact of different distillation temperatures on the performance of student models. By using YOLOv5l as the teacher network and a smaller YOLOv5s as the student network, we found that with the increase of distillation temperature, the student's detection accuracy gradually improved, and finally achieved mAP50 and mAP50-95 indicators that were better than the original YOLOv5s model at a specific temperature. Experimental results show that appropriate knowledge distillation strategies can not only improve the accuracy of the model but also help improve the reliability and stability of the model in practical applications. This paper also records in detail the accuracy curve and loss function descent curve during the model training process and shows that the model converges to a stable state after 150 training cycles. These findings provide a theoretical basis and technical reference for further optimizing target detection algorithms.
☆ Advancing Healthcare: Innovative ML Approaches for Improved Medical Imaging in Data-Constrained Environments
Healthcare industries face challenges when experiencing rare diseases due to limited samples. Artificial Intelligence (AI) communities overcome this situation to create synthetic data which is an ethical and privacy issue in the medical domain. This research introduces the CAT-U-Net framework as a new approach to overcome these limitations, which enhances feature extraction from medical images without the need for large datasets. The proposed framework adds an extra concatenation layer with downsampling parts, thereby improving its ability to learn from limited data while maintaining patient privacy. To validate, the proposed framework's robustness, different medical conditioning datasets were utilized including COVID-19, brain tumors, and wrist fractures. The framework achieved nearly 98% reconstruction accuracy, with a Dice coefficient close to 0.946. The proposed CAT-U-Net has the potential to make a big difference in medical image diagnostics in settings with limited data.
comment: 7 pages, 7 figures
☆ EG-HumanNeRF: Efficient Generalizable Human NeRF Utilizing Human Prior for Sparse View
Generalizable neural radiance field (NeRF) enables neural-based digital human rendering without per-scene retraining. When combined with human prior knowledge, high-quality human rendering can be achieved even with sparse input views. However, the inference of these methods is still slow, as a large number of neural network queries on each ray are required to ensure the rendering quality. Moreover, occluded regions often suffer from artifacts, especially when the input views are sparse. To address these issues, we propose a generalizable human NeRF framework that achieves high-quality and real-time rendering with sparse input views by extensively leveraging human prior knowledge. We accelerate the rendering with a two-stage sampling reduction strategy: first constructing boundary meshes around the human geometry to reduce the number of ray samples for sampling guidance regression, and then volume rendering using fewer guided samples. To improve rendering quality, especially in occluded regions, we propose an occlusion-aware attention mechanism to extract occlusion information from the human priors, followed by an image space refinement network to improve rendering quality. Furthermore, for volume rendering, we adopt a signed ray distance function (SRDF) formulation, which allows us to propose an SRDF loss at every sample position to improve the rendering quality further. Our experiments demonstrate that our method outperforms the state-of-the-art methods in rendering quality and has a competitive rendering speed compared with speed-prioritized novel view synthesis methods.
comment: project page: https://github.com/LarsPh/EG-HumanNeRF
☆ Leveraging Spatial Attention and Edge Context for Optimized Feature Selection in Visual Localization
Visual localization determines an agent's precise position and orientation within an environment using visual data. It has become a critical task in the field of robotics, particularly in applications such as autonomous navigation. This is due to the ability to determine an agent's pose using cost-effective sensors such as RGB cameras. Recent methods in visual localization employ scene coordinate regression to determine the agent's pose. However, these methods face challenges as they attempt to regress 2D-3D correspondences across the entire image region, despite not all regions providing useful information. To address this issue, we introduce an attention network that selectively targets informative regions of the image. Using this network, we identify the highest-scoring features to improve the feature selection process and combine the result with edge detection. This integration ensures that the features chosen for the training buffer are located within robust regions, thereby improving 2D-3D correspondence and overall localization performance. Our approach was tested on the outdoor benchmark dataset, demonstrating superior results compared to previous methods.
☆ Evaluating Cascaded Methods of Vision-Language Models for Zero-Shot Detection and Association of Hardhats for Increased Construction Safety
This paper evaluates the use of vision-language models (VLMs) for zero-shot detection and association of hardhats to enhance construction safety. Given the significant risk of head injuries in construction, proper enforcement of hardhat use is critical. We investigate the applicability of foundation models, specifically OWLv2, for detecting hardhats in real-world construction site images. Our contributions include the creation of a new benchmark dataset, Hardhat Safety Detection Dataset, by filtering and combining existing datasets and the development of a cascaded detection approach. Experimental results on 5,210 images demonstrate that the OWLv2 model achieves an average precision of 0.6493 for hardhat detection. We further analyze the limitations and potential improvements for real-world applications, highlighting the strengths and weaknesses of current foundation models in safety perception domains.
☆ Order-Aware Interactive Segmentation
Interactive segmentation aims to accurately segment target objects with minimal user interactions. However, current methods often fail to accurately separate target objects from the background, due to a limited understanding of order, the relative depth between objects in a scene. To address this issue, we propose OIS: order-aware interactive segmentation, where we explicitly encode the relative depth between objects into order maps. We introduce a novel order-aware attention, where the order maps seamlessly guide the user interactions (in the form of clicks) to attend to the image features. We further present an object-aware attention module to incorporate a strong object-level understanding to better differentiate objects with similar order. Our approach allows both dense and sparse integration of user clicks, enhancing both accuracy and efficiency as compared to prior works. Experimental results demonstrate that OIS achieves state-of-the-art performance, improving mIoU after one click by 7.61 on the HQSeg44K dataset and 1.32 on the DAVIS dataset as compared to the previous state-of-the-art SegNext, while also doubling inference speed compared to current leading methods. The project page is https://ukaukaaaa.github.io/projects/OIS/index.html
comment: Interactive demo can be found in project page: https://ukaukaaaa.github.io/projects/OIS/index.html
☆ Sparse Prototype Network for Explainable Pedestrian Behavior Prediction
Predicting pedestrian behavior is challenging yet crucial for applications such as autonomous driving and smart city. Recent deep learning models have achieved remarkable performance in making accurate predictions, but they fail to provide explanations of their inner workings. One reason for this problem is the multi-modal inputs. To bridge this gap, we present Sparse Prototype Network (SPN), an explainable method designed to simultaneously predict a pedestrian's future action, trajectory, and pose. SPN leverages an intermediate prototype bottleneck layer to provide sample-based explanations for its predictions. The prototypes are modality-independent, meaning that they can correspond to any modality from the input. Therefore, SPN can extend to arbitrary combinations of modalities. Regularized by mono-semanticity and clustering constraints, the prototypes learn consistent and human-understandable features and achieve state-of-the-art performance on action, trajectory and pose prediction on TITAN and PIE. Finally, we propose a metric named Top-K Mono-semanticity Scale to quantitatively evaluate the explainability. Qualitative results show the positive correlation between sparsity and explainability. Code available at https://github.com/Equinoxxxxx/SPN.
☆ Test-time adaptation for image compression with distribution regularization
Current test- or compression-time adaptation image compression (TTA-IC) approaches, which leverage both latent and decoder refinements as a two-step adaptation scheme, have potentially enhanced the rate-distortion (R-D) performance of learned image compression models on cross-domain compression tasks, \textit{e.g.,} from natural to screen content images. However, compared with the emergence of various decoder refinement variants, the latent refinement, as an inseparable ingredient, is barely tailored to cross-domain scenarios. To this end, we aim to develop an advanced latent refinement method by extending the effective hybrid latent refinement (HLR) method, which is designed for \textit{in-domain} inference improvement but shows noticeable degradation of the rate cost in \textit{cross-domain} tasks. Specifically, we first provide theoretical analyses, in a cue of marginalization approximation from in- to cross-domain scenarios, to uncover that the vanilla HLR suffers from an underlying mismatch between refined Gaussian conditional and hyperprior distributions, leading to deteriorated joint probability approximation of marginal distribution with increased rate consumption. To remedy this issue, we introduce a simple Bayesian approximation-endowed \textit{distribution regularization} to encourage learning a better joint probability approximation in a plug-and-play manner. Extensive experiments on six in- and cross-domain datasets demonstrate that our proposed method not only improves the R-D performance compared with other latent refinement counterparts, but also can be flexibly integrated into existing TTA-IC methods with incremental benefits.
☆ TransAgent: Transfer Vision-Language Foundation Models with Heterogeneous Agent Collaboration NeurIPS 2024
Vision-language foundation models (such as CLIP) have recently shown their power in transfer learning, owing to large-scale image-text pre-training. However, target domain data in the downstream tasks can be highly different from the pre-training phase, which makes it hard for such a single model to generalize well. Alternatively, there exists a wide range of expert models that contain diversified vision and/or language knowledge pre-trained on different modalities, tasks, networks, and datasets. Unfortunately, these models are "isolated agents" with heterogeneous structures, and how to integrate their knowledge for generalizing CLIP-like models has not been fully explored. To bridge this gap, we propose a general and concise TransAgent framework, which transports the knowledge of the isolated agents in a unified manner, and effectively guides CLIP to generalize with multi-source knowledge distillation. With such a distinct framework, we flexibly collaborate with 11 heterogeneous agents to empower vision-language foundation models, without further cost in the inference phase. Finally, our TransAgent achieves state-of-the-art performance on 11 visual recognition datasets. Under the same low-shot setting, it outperforms the popular CoOp with around 10% on average, and 20% on EuroSAT which contains large domain shifts.
comment: NeurIPS 2024
☆ Dual-Model Distillation for Efficient Action Classification with Hybrid Edge-Cloud Solution
As Artificial Intelligence models, such as Large Video-Language models (VLMs), grow in size, their deployment in real-world applications becomes increasingly challenging due to hardware limitations and computational costs. To address this, we design a hybrid edge-cloud solution that leverages the efficiency of smaller models for local processing while deferring to larger, more accurate cloud-based models when necessary. Specifically, we propose a novel unsupervised data generation method, Dual-Model Distillation (DMD), to train a lightweight switcher model that can predict when the edge model's output is uncertain and selectively offload inference to the large model in the cloud. Experimental results on the action classification task show that our framework not only requires less computational overhead, but also improves accuracy compared to using a large model alone. Our framework provides a scalable and adaptable solution for action classification in resource-constrained environments, with potential applications beyond healthcare. Noteworthy, while DMD-generated data is used for optimizing performance and resource usage in our pipeline, we expect the concept of DMD to further support future research on knowledge alignment across multiple models.
☆ SAM-Guided Masked Token Prediction for 3D Scene Understanding NeurIPS 2024
Foundation models have significantly enhanced 2D task performance, and recent works like Bridge3D have successfully applied these models to improve 3D scene understanding through knowledge distillation, marking considerable advancements. Nonetheless, challenges such as the misalignment between 2D and 3D representations and the persistent long-tail distribution in 3D datasets still restrict the effectiveness of knowledge distillation from 2D to 3D using foundation models. To tackle these issues, we introduce a novel SAM-guided tokenization method that seamlessly aligns 3D transformer structures with region-level knowledge distillation, replacing the traditional KNN-based tokenization techniques. Additionally, we implement a group-balanced re-weighting strategy to effectively address the long-tail problem in knowledge distillation. Furthermore, inspired by the recent success of masked feature prediction, our framework incorporates a two-stage masked token prediction process in which the student model predicts both the global embeddings and the token-wise local embeddings derived from the teacher models trained in the first stage. Our methodology has been validated across multiple datasets, including SUN RGB-D, ScanNet, and S3DIS, for tasks like 3D object detection and semantic segmentation. The results demonstrate significant improvements over current State-of-the-art self-supervised methods, establishing new benchmarks in this field.
comment: Accepted by NeurIPS 2024
☆ Unveiling the Limits of Alignment: Multi-modal Dynamic Local Fusion Network and A Benchmark for Unaligned RGBT Video Object Detection
Current RGB-Thermal Video Object Detection (RGBT VOD) methods still depend on manually aligning data at the image level, which hampers its practical application in real-world scenarios since image pairs captured by multispectral sensors often differ in both fields of view and resolution. To address this limitation, we propose a Multi-modal Dynamic Local fusion Network (MDLNet) designed to handle unaligned RGBT image pairs. Specifically, our proposed Multi-modal Dynamic Local Fusion (MDLF) module includes a set of predefined boxes, each enhanced with random Gaussian noise to generate a dynamic box. Each box selects a local region from the original high-resolution RGB image. This region is then fused with the corresponding information from another modality and reinserted into the RGB. This method adapts to various data alignment scenarios by interacting with local features across different ranges. Simultaneously, we introduce a Cascaded Temporal Scrambler (CTS) within an end-to-end architecture. This module leverages consistent spatiotemporal information from consecutive frames to enhance the representation capability of the current frame while maintaining network efficiency. We have curated an open dataset called UVT-VOD2024 for unaligned RGBT VOD. It consists of 30,494 pairs of unaligned RGBT images captured directly from a multispectral camera. We conduct a comprehensive evaluation and comparison with MDLNet and state-of-the-art (SOTA) models, demonstrating the superior effectiveness of MDLNet. We will release our code and UVT-VOD2024 to the public for further research.
☆ Task Consistent Prototype Learning for Incremental Few-shot Semantic Segmentation
Incremental Few-Shot Semantic Segmentation (iFSS) tackles a task that requires a model to continually expand its segmentation capability on novel classes using only a few annotated examples. Typical incremental approaches encounter a challenge that the objective of the base training phase (fitting base classes with sufficient instances) does not align with the incremental learning phase (rapidly adapting to new classes with less forgetting). This disconnect can result in suboptimal performance in the incremental setting. This study introduces a meta-learning-based prototype approach that encourages the model to learn how to adapt quickly while preserving previous knowledge. Concretely, we mimic the incremental evaluation protocol during the base training session by sampling a sequence of pseudo-incremental tasks. Each task in the simulated sequence is trained using a meta-objective to enable rapid adaptation without forgetting. To enhance discrimination among class prototypes, we introduce prototype space redistribution learning, which dynamically updates class prototypes to establish optimal inter-prototype boundaries within the prototype space. Extensive experiments on iFSS datasets built upon PASCAL and COCO benchmarks show the advanced performance of the proposed approach, offering valuable insights for addressing iFSS challenges.
comment: conference
☆ MMed-RAG: Versatile Multimodal RAG System for Medical Vision Language Models
Artificial Intelligence (AI) has demonstrated significant potential in healthcare, particularly in disease diagnosis and treatment planning. Recent progress in Medical Large Vision-Language Models (Med-LVLMs) has opened up new possibilities for interactive diagnostic tools. However, these models often suffer from factual hallucination, which can lead to incorrect diagnoses. Fine-tuning and retrieval-augmented generation (RAG) have emerged as methods to address these issues. However, the amount of high-quality data and distribution shifts between training data and deployment data limit the application of fine-tuning methods. Although RAG is lightweight and effective, existing RAG-based approaches are not sufficiently general to different medical domains and can potentially cause misalignment issues, both between modalities and between the model and the ground truth. In this paper, we propose a versatile multimodal RAG system, MMed-RAG, designed to enhance the factuality of Med-LVLMs. Our approach introduces a domain-aware retrieval mechanism, an adaptive retrieved contexts selection method, and a provable RAG-based preference fine-tuning strategy. These innovations make the RAG process sufficiently general and reliable, significantly improving alignment when introducing retrieved contexts. Experimental results across five medical datasets (involving radiology, ophthalmology, pathology) on medical VQA and report generation demonstrate that MMed-RAG can achieve an average improvement of 43.8% in the factual accuracy of Med-LVLMs. Our data and code are available in https://github.com/richard-peng-xia/MMed-RAG.
☆ BOXR: Body and head motion Optimization framework for eXtended Reality
The emergence of standalone XR systems has enhanced user mobility, accommodating both subtle, frequent head motions and substantial, less frequent body motions. However, the pervasively used M2D latency metric, which measures the delay between the most recent motion and its corresponding display update, only accounts for head motions. This oversight can leave users prone to motion sickness if significant body motion is involved. Although existing methods optimize M2D latency through asynchronous task scheduling and reprojection methods, they introduce challenges like resource contention between tasks and outdated pose data. These challenges are further complicated by user motion dynamics and scene changes during runtime. To address these issues, we for the first time introduce the C2D latency metric, which captures the delay caused by body motions, and present BOXR, a framework designed to co-optimize both body and head motion delays within an XR system. BOXR enhances the coordination between M2D and C2D latencies by efficiently scheduling tasks to avoid contentions while maintaining an up-to-date pose in the output frame. Moreover, BOXR incorporates a motion-driven visual inertial odometer to adjust to user motion dynamics and employs scene-dependent foveated rendering to manage changes in the scene effectively. Our evaluations show that BOXR significantly outperforms state-of-the-art solutions in 11 EuRoC MAV datasets across 4 XR applications across 3 hardware platforms. In controlled motion and scene settings, BOXR reduces M2D and C2D latencies by up to 63% and 27%, respectively and increases frame rate by up to 43%. In practical deployments, BOXR achieves substantial reductions in real-world scenarios up to 42% in M2D latency and 31% in C2D latency while maintaining remarkably low miss rates of only 1.6% for M2D requirements and 1.0% for C2D requirements.
comment: Accepted to 45th IEEE Real-Time Systems Symposium (RTSS'24)
☆ UniCoN: Universal Conditional Networks for Multi-Age Embryonic Cartilage Segmentation with Sparsely Annotated Data
Osteochondrodysplasia, affecting 2-3% of newborns globally, is a group of bone and cartilage disorders that often result in head malformations, contributing to childhood morbidity and reduced quality of life. Current research on this disease using mouse models faces challenges since it involves accurately segmenting the developing cartilage in 3D micro-CT images of embryonic mice. Tackling this segmentation task with deep learning (DL) methods is laborious due to the big burden of manual image annotation, expensive due to the high acquisition costs of 3D micro-CT images, and difficult due to embryonic cartilage's complex and rapidly changing shapes. While DL approaches have been proposed to automate cartilage segmentation, most such models have limited accuracy and generalizability, especially across data from different embryonic age groups. To address these limitations, we propose novel DL methods that can be adopted by any DL architectures -- including CNNs, Transformers, or hybrid models -- which effectively leverage age and spatial information to enhance model performance. Specifically, we propose two new mechanisms, one conditioned on discrete age categories and the other on continuous image crop locations, to enable an accurate representation of cartilage shape changes across ages and local shape details throughout the cranial region. Extensive experiments on multi-age cartilage segmentation datasets show significant and consistent performance improvements when integrating our conditional modules into popular DL segmentation architectures. On average, we achieve a 1.7% Dice score increase with minimal computational overhead and a 7.5% improvement on unseen data. These results highlight the potential of our approach for developing robust, universal models capable of handling diverse datasets with limited annotated data, a key challenge in DL-based medical image analysis.
☆ A low complexity contextual stacked ensemble-learning approach for pedestrian intent prediction
Walking as a form of active travel is essential in promoting sustainable transport. It is thus crucial to accurately predict pedestrian crossing intention and avoid collisions, especially with the advent of autonomous and advanced driver-assisted vehicles. Current research leverages computer vision and machine learning advances to predict near-misses; however, this often requires high computation power to yield reliable results. In contrast, this work proposes a low-complexity ensemble-learning approach that employs contextual data for predicting the pedestrian's intent for crossing. The pedestrian is first detected, and their image is then compressed using skeleton-ization, and contextual information is added into a stacked ensemble-learning approach. Our experiments on different datasets achieve similar pedestrian intent prediction performance as the state-of-the-art approaches with 99.7% reduction in computational complexity. Our source code and trained models will be released upon paper acceptance
☆ Synthesis and Perceptual Scaling of High Resolution Natural Images Using Stable Diffusion
Natural scenes are of key interest for visual perception. Previous work on natural scenes has frequently focused on collections of discrete images with considerable physical differences from stimulus to stimulus. For many purposes it would, however, be desirable to have sets of natural images that vary smoothly along a continuum (for example in order to measure quantitative properties such as thresholds or precisions). This problem has typically been addressed by morphing a source into a target image. However, this approach yields transitions between images that primarily follow their low-level physical features and that can be semantically unclear or ambiguous. Here, in contrast, we used a different approach (Stable Diffusion XL) to synthesise a custom stimulus set of photorealistic images that are characterized by gradual transitions where each image is a clearly interpretable but unique exemplar from the same category. We developed natural scene stimulus sets from six categories with 18 objects each. For each object we generated 10 graded variants that are ordered along a perceptual continuum. We validated the image set psychophysically in a large sample of participants, ensuring that stimuli for each exemplar have varying levels of perceptual confusability. This image set is of interest for studies on visual perception, attention and short- and long-term memory.
comment: 29 pages, 7 Figures, 5 tables
☆ Sensitivity of Generative VLMs to Semantically and Lexically Altered Prompts
Despite the significant influx of prompt-tuning techniques for generative vision-language models (VLMs), it remains unclear how sensitive these models are to lexical and semantic alterations in prompts. In this paper, we evaluate the ability of generative VLMs to understand lexical and semantic changes in text using the SugarCrepe++ dataset. We analyze the sensitivity of VLMs to lexical alterations in prompts without corresponding semantic changes. Our findings demonstrate that generative VLMs are highly sensitive to such alterations. Additionally, we show that this vulnerability affects the performance of techniques aimed at achieving consistency in their outputs.
☆ Geometric Trajectory Diffusion Models NeurIPS 2024
Generative models have shown great promise in generating 3D geometric systems, which is a fundamental problem in many natural science domains such as molecule and protein design. However, existing approaches only operate on static structures, neglecting the fact that physical systems are always dynamic in nature. In this work, we propose geometric trajectory diffusion models (GeoTDM), the first diffusion model for modeling the temporal distribution of 3D geometric trajectories. Modeling such distribution is challenging as it requires capturing both the complex spatial interactions with physical symmetries and temporal correspondence encapsulated in the dynamics. We theoretically justify that diffusion models with equivariant temporal kernels can lead to density with desired symmetry, and develop a novel transition kernel leveraging SE(3)-equivariant spatial convolution and temporal attention. Furthermore, to induce an expressive trajectory distribution for conditional generation, we introduce a generalized learnable geometric prior into the forward diffusion process to enhance temporal conditioning. We conduct extensive experiments on both unconditional and conditional generation in various scenarios, including physical simulation, molecular dynamics, and pedestrian motion. Empirical results on a wide suite of metrics demonstrate that GeoTDM can generate realistic geometric trajectories with significantly higher quality.
comment: Published at NeurIPS 2024. 29 pages, 10 figures
☆ Interpreting and Analyzing CLIP's Zero-Shot Image Classification via Mutual Knowledge NeurIPS 2024
Contrastive Language-Image Pretraining (CLIP) performs zero-shot image classification by mapping images and textual class representation into a shared embedding space, then retrieving the class closest to the image. This work provides a new approach for interpreting CLIP models for image classification from the lens of mutual knowledge between the two modalities. Specifically, we ask: what concepts do both vision and language CLIP encoders learn in common that influence the joint embedding space, causing points to be closer or further apart? We answer this question via an approach of textual concept-based explanations, showing their effectiveness, and perform an analysis encompassing a pool of 13 CLIP models varying in architecture, size and pretraining datasets. We explore those different aspects in relation to mutual knowledge, and analyze zero-shot predictions. Our approach demonstrates an effective and human-friendly way of understanding zero-shot classification decisions with CLIP.
comment: Accepted to NeurIPS 2024
☆ Hiding-in-Plain-Sight (HiPS) Attack on CLIP for Targetted Object Removal from Images NeurIPS 2024
Machine learning models are known to be vulnerable to adversarial attacks, but traditional attacks have mostly focused on single-modalities. With the rise of large multi-modal models (LMMs) like CLIP, which combine vision and language capabilities, new vulnerabilities have emerged. However, prior work in multimodal targeted attacks aim to completely change the model's output to what the adversary wants. In many realistic scenarios, an adversary might seek to make only subtle modifications to the output, so that the changes go unnoticed by downstream models or even by humans. We introduce Hiding-in-Plain-Sight (HiPS) attacks, a novel class of adversarial attacks that subtly modifies model predictions by selectively concealing target object(s), as if the target object was absent from the scene. We propose two HiPS attack variants, HiPS-cls and HiPS-cap, and demonstrate their effectiveness in transferring to downstream image captioning models, such as CLIP-Cap, for targeted object removal from image captions.
comment: Published in the 3rd Workshop on New Frontiers in Adversarial Machine Learning at NeurIPS 2024. 10 pages, 7 figures, 3 tables
☆ Configurable Embodied Data Generation for Class-Agnostic RGB-D Video Segmentation
This paper presents a method for generating large-scale datasets to improve class-agnostic video segmentation across robots with different form factors. Specifically, we consider the question of whether video segmentation models trained on generic segmentation data could be more effective for particular robot platforms if robot embodiment is factored into the data generation process. To answer this question, a pipeline is formulated for using 3D reconstructions (e.g. from HM3DSem) to generate segmented videos that are configurable based on a robot's embodiment (e.g. sensor type, sensor placement, and illumination source). A resulting massive RGB-D video panoptic segmentation dataset (MVPd) is introduced for extensive benchmarking with foundation and video segmentation models, as well as to support embodiment-focused research in video segmentation. Our experimental findings demonstrate that using MVPd for finetuning can lead to performance improvements when transferring foundation models to certain robot embodiments, such as specific camera placements. These experiments also show that using 3D modalities (depth images and camera pose) can lead to improvements in video segmentation accuracy and consistency. The project webpage is available at https://topipari.com/projects/MVPd
comment: Accepted in IEEE Robotics and Automation Letters October 2024
☆ Explainable Binary Classification of Separable Shape Ensembles
Materials scientists utilize image segmentation of micrographs to create large curve ensembles representing grain boundaries of material microstructures. Observations of these collections of shapes can facilitate inferences about material properties and manufacturing processes. We seek to bolster this application, and related engineering/scientific tasks, using novel pattern recognition formalisms and inference over large ensembles of segmented curves -- i.e., facilitate principled assessments for quantifying differences in distributions of shapes. To this end, we apply a composite integral operator to motivate accurate and efficient numerical representations of discrete planar curves over matrix manifolds. The main result is a rigid-invariant orthonormal decomposition of curve component functions into separable forms of scale variations and complementary features of undulation. We demonstrate how these separable shape tensors -- given thousands of curves in an ensemble -- can inform explainable binary classification of segmented images by utilizing a product maximum mean discrepancy to distinguish the shape distributions; absent labelled data, building interpretable feature spaces in seconds without high performance computation, and detecting discrepancies below cursory visual inspections.
comment: 20 pages, 7 figures
☆ Risk Assessment for Autonomous Landing in Urban Environments using Semantic Segmentation
In this paper, we address the vision-based autonomous landing problem in complex urban environments using deep neural networks for semantic segmentation and risk assessment. We propose employing the SegFormer, a state-of-the-art visual transformer network, for the semantic segmentation of complex, unstructured urban environments. This approach yields valuable information that can be utilized in smart autonomous landing missions, particularly in emergency landing scenarios resulting from system failures or human errors. The assessment is done in real-time flight, when images of an RGB camera at the Unmanned Aerial Vehicle (UAV) are segmented with the SegFormer into the most common classes found in urban environments. These classes are then mapped into a level of risk, considering in general, potential material damage, damaging the drone itself and endanger people. The proposed strategy is validated through several case studies, demonstrating the huge potential of semantic segmentation-based strategies to determining the safest landing areas for autonomous emergency landing, which we believe will help unleash the full potential of UAVs on civil applications within urban areas.
☆ Super-resolving Real-world Image Illumination Enhancement: A New Dataset and A Conditional Diffusion Model
Most existing super-resolution methods and datasets have been developed to improve the image quality in well-lighted conditions. However, these methods do not work well in real-world low-light conditions as the images captured in such conditions lose most important information and contain significant unknown noises. To solve this problem, we propose a SRRIIE dataset with an efficient conditional diffusion probabilistic models-based method. The proposed dataset contains 4800 paired low-high quality images. To ensure that the dataset are able to model the real-world image degradation in low-illumination environments, we capture images using an ILDC camera and an optical zoom lens with exposure levels ranging from -6 EV to 0 EV and ISO levels ranging from 50 to 12800. We comprehensively evaluate with various reconstruction and perceptual metrics and demonstrate the practicabilities of the SRRIIE dataset for deep learning-based methods. We show that most existing methods are less effective in preserving the structures and sharpness of restored images from complicated noises. To overcome this problem, we revise the condition for Raw sensor data and propose a novel time-melding condition for diffusion probabilistic model. Comprehensive quantitative and qualitative experimental results on the real-world benchmark datasets demonstrate the feasibility and effectivenesses of the proposed conditional diffusion probabilistic model on Raw sensor data. Code and dataset will be available at https://github.com/Yaofang-Liu/Super-Resolving
comment: Code and dataset at https://github.com/Yaofang-Liu/Super-Resolving
☆ MuVi: Video-to-Music Generation with Semantic Alignment and Rhythmic Synchronization
Generating music that aligns with the visual content of a video has been a challenging task, as it requires a deep understanding of visual semantics and involves generating music whose melody, rhythm, and dynamics harmonize with the visual narratives. This paper presents MuVi, a novel framework that effectively addresses these challenges to enhance the cohesion and immersive experience of audio-visual content. MuVi analyzes video content through a specially designed visual adaptor to extract contextually and temporally relevant features. These features are used to generate music that not only matches the video's mood and theme but also its rhythm and pacing. We also introduce a contrastive music-visual pre-training scheme to ensure synchronization, based on the periodicity nature of music phrases. In addition, we demonstrate that our flow-matching-based music generator has in-context learning ability, allowing us to control the style and genre of the generated music. Experimental results show that MuVi demonstrates superior performance in both audio quality and temporal synchronization. The generated music video samples are available at https://muvi-v2m.github.io.
comment: Working in progress
☆ Long-Tailed Backdoor Attack Using Dynamic Data Augmentation Operations
Recently, backdoor attack has become an increasing security threat to deep neural networks and drawn the attention of researchers. Backdoor attacks exploit vulnerabilities in third-party pretrained models during the training phase, enabling them to behave normally for clean samples and mispredict for samples with specific triggers. Existing backdoor attacks mainly focus on balanced datasets. However, real-world datasets often follow long-tailed distributions. In this paper, for the first time, we explore backdoor attack on such datasets. Specifically, we first analyze the influence of data imbalance on backdoor attack. Based on our analysis, we propose an effective backdoor attack named Dynamic Data Augmentation Operation (D$^2$AO). We design D$^2$AO selectors to select operations depending jointly on the class, sample type (clean vs. backdoored) and sample features. Meanwhile, we develop a trigger generator to generate sample-specific triggers. Through simultaneous optimization of the backdoored model and trigger generator, guided by dynamic data augmentation operation selectors, we achieve significant advancements. Extensive experiments demonstrate that our method can achieve the state-of-the-art attack performance while preserving the clean accuracy.
☆ Syn2Real Domain Generalization for Underwater Mine-like Object Detection Using Side-Scan Sonar
Underwater mine detection with deep learning suffers from limitations due to the scarcity of real-world data. This scarcity leads to overfitting, where models perform well on training data but poorly on unseen data. This paper proposes a Syn2Real (Synthetic to Real) domain generalization approach using diffusion models to address this challenge. We demonstrate that synthetic data generated with noise by DDPM and DDIM models, even if not perfectly realistic, can effectively augment real-world samples for training. The residual noise in the final sampled images improves the model's ability to generalize to real-world data with inherent noise and high variation. The baseline Mask-RCNN model when trained on a combination of synthetic and original training datasets, exhibited approximately a 60% increase in Average Precision (AP) compared to being trained solely on the original training data. This significant improvement highlights the potential of Syn2Real domain generalization for underwater mine detection tasks.
comment: 7 pages, 4 figures and 3 tables
☆ Gradient Map-Assisted Head and Neck Tumor Segmentation: A Pre-RT to Mid-RT Approach in MRI-Guided Radiotherapy
Radiation therapy (RT) is a vital part of treatment for head and neck cancer, where accurate segmentation of gross tumor volume (GTV) is essential for effective treatment planning. This study investigates the use of pre-RT tumor regions and local gradient maps to enhance mid-RT tumor segmentation for head and neck cancer in MRI-guided adaptive radiotherapy. By leveraging pre-RT images and their segmentations as prior knowledge, we address the challenge of tumor localization in mid-RT segmentation. A gradient map of the tumor region from the pre-RT image is computed and applied to mid-RT images to improve tumor boundary delineation. Our approach demonstrated improved segmentation accuracy for both primary GTV (GTVp) and nodal GTV (GTVn), though performance was limited by data constraints. The final DSCagg scores from the challenge's test set evaluation were 0.534 for GTVp, 0.867 for GTVn, and a mean score of 0.70. This method shows potential for enhancing segmentation and treatment planning in adaptive radiotherapy. Team: DCPT-Stine's group.
☆ UMambaAdj: Advancing GTV Segmentation for Head and Neck Cancer in MRI-Guided RT with UMamba and nnU-Net ResEnc Planner
Magnetic Resonance Imaging (MRI) plays a crucial role in MRI-guided adaptive radiotherapy for head and neck cancer (HNC) due to its superior soft-tissue contrast. However, accurately segmenting the gross tumor volume (GTV), which includes both the primary tumor (GTVp) and lymph nodes (GTVn), remains challenging. Recently, two deep learning segmentation innovations have shown great promise: UMamba, which effectively captures long-range dependencies, and the nnU-Net Residual Encoder (ResEnc), which enhances feature extraction through multistage residual blocks. In this study, we integrate these strengths into a novel approach, termed 'UMambaAdj'. Our proposed method was evaluated on the HNTS-MRG 2024 challenge test set using pre-RT T2-weighted MRI images, achieving an aggregated Dice Similarity Coefficient (DSCagg) of 0.751 for GTVp and 0.842 for GTVn, with a mean DSCagg of 0.796. This approach demonstrates potential for more precise tumor delineation in MRI-guided adaptive radiotherapy, ultimately improving treatment outcomes for HNC patients. Team: DCPT-Stine's group.
☆ DreamCraft3D++: Efficient Hierarchical 3D Generation with Multi-Plane Reconstruction Model
We introduce DreamCraft3D++, an extension of DreamCraft3D that enables efficient high-quality generation of complex 3D assets. DreamCraft3D++ inherits the multi-stage generation process of DreamCraft3D, but replaces the time-consuming geometry sculpting optimization with a feed-forward multi-plane based reconstruction model, speeding up the process by 1000x. For texture refinement, we propose a training-free IP-Adapter module that is conditioned on the enhanced multi-view images to enhance texture and geometry consistency, providing a 4x faster alternative to DreamCraft3D's DreamBooth fine-tuning. Experiments on diverse datasets demonstrate DreamCraft3D++'s ability to generate creative 3D assets with intricate geometry and realistic 360{\deg} textures, outperforming state-of-the-art image-to-3D methods in quality and speed. The full implementation will be open-sourced to enable new possibilities in 3D content creation.
comment: Project Page: https://dreamcraft3dplus.github.io/
☆ DEeR: Deviation Eliminating and Noise Regulating for Privacy-preserving Federated Low-rank Adaptation
Integrating low-rank adaptation (LoRA) with federated learning (FL) has received widespread attention recently, aiming to adapt pretrained foundation models (FMs) to downstream medical tasks via privacy-preserving decentralized training. However, owing to the direct combination of LoRA and FL, current methods generally undergo two problems, i.e., aggregation deviation, and differential privacy (DP) noise amplification effect. To address these problems, we propose a novel privacy-preserving federated finetuning framework called \underline{D}eviation \underline{E}liminating and Nois\underline{e} \underline{R}egulating (DEeR). Specifically, we firstly theoretically prove that the necessary condition to eliminate aggregation deviation is guaranteing the equivalence between LoRA parameters of clients. Based on the theoretical insight, a deviation eliminator is designed to utilize alternating minimization algorithm to iteratively optimize the zero-initialized and non-zero-initialized parameter matrices of LoRA, ensuring that aggregation deviation always be zeros during training. Furthermore, we also conduct an in-depth analysis of the noise amplification effect and find that this problem is mainly caused by the ``linear relationship'' between DP noise and LoRA parameters. To suppress the noise amplification effect, we propose a noise regulator that exploits two regulator factors to decouple relationship between DP and LoRA, thereby achieving robust privacy protection and excellent finetuning performance. Additionally, we perform comprehensive ablated experiments to verify the effectiveness of the deviation eliminator and noise regulator. DEeR shows better performance on public medical datasets in comparison with state-of-the-art approaches. The code is available at https://github.com/CUHK-AIM-Group/DEeR.
♻ ☆ Preserving Cardiac Integrity: A Topology-Infused Approach to Whole Heart Segmentation
Whole heart segmentation (WHS) supports cardiovascular disease (CVD) diagnosis, disease monitoring, treatment planning, and prognosis. Deep learning has become the most widely used method for WHS applications in recent years. However, segmentation of whole-heart structures faces numerous challenges including heart shape variability during the cardiac cycle, clinical artifacts like motion and poor contrast-to-noise ratio, domain shifts in multi-center data, and the distinct modalities of CT and MRI. To address these limitations and improve segmentation quality, this paper introduces a new topology-preserving module that is integrated into deep neural networks. The implementation achieves anatomically plausible segmentation by using learned topology-preserving fields, which are based entirely on 3D convolution and are therefore very effective for 3D voxel data. We incorporate natural constraints between structures into the end-to-end training and enrich the feature representation of the neural network. The effectiveness of the proposed method is validated on an open-source medical heart dataset, specifically using the WHS++ data. The results demonstrate that the architecture performs exceptionally well, achieving a Dice coefficient of 0.939 during testing. This indicates full topology preservation for individual structures and significantly outperforms other baselines in preserving the overall scene topology.
♻ ☆ AssemAI: Interpretable Image-Based Anomaly Detection for Manufacturing Pipelines ICML
Anomaly detection in manufacturing pipelines remains a critical challenge, intensified by the complexity and variability of industrial environments. This paper introduces AssemAI, an interpretable image-based anomaly detection system tailored for smart manufacturing pipelines. Utilizing a curated image dataset from an industry-focused rocket assembly pipeline, we address the challenge of imbalanced image data and demonstrate the importance of image-based methods in anomaly detection. Our primary contributions include deriving an image dataset, fine-tuning an object detection model YOLO-FF, and implementing a custom anomaly detection model for assembly pipelines. The proposed approach leverages domain knowledge in data preparation, model development and reasoning. We implement several anomaly detection models on the derived image dataset, including a Convolutional Neural Network, Vision Transformer (ViT), and pre-trained versions of these models. Additionally, we incorporate explainability techniques at both user and model levels, utilizing ontology for user-level explanations and SCORE-CAM for in-depth feature and model analysis. Finally, the best-performing anomaly detection model and YOLO-FF are deployed in a real-time setting. Our results include ablation studies on the baselines and a comprehensive evaluation of the proposed system. This work highlights the broader impact of advanced image-based anomaly detection in enhancing the reliability and efficiency of smart manufacturing processes. The image dataset, codes to reproduce the results and additional experiments are available at https://github.com/renjithk4/AssemAI.
comment: 8 Pages, 6 Figures, 4 Tables, Predictive Models in Engineering Applications special session (MLPMEA )at International Conference on Machine Learning and Applications (ICMLA) 2024
♻ ☆ Understanding Figurative Meaning through Explainable Visual Entailment
Large Vision-Language Models (VLMs) have demonstrated strong capabilities in tasks requiring a fine-grained understanding of literal meaning in images and text, such as visual question-answering or visual entailment. However, there has been little exploration of these models' capabilities when presented with images and captions containing figurative meaning, such as metaphors or humor. To close this gap, we propose a new task framing the figurative meaning understanding problem as an explainable visual entailment task, where the model has to predict whether the image (premise) entails a caption (hypothesis) and justify the predicted label with a textual explanation. The figurative phenomena can be present either in the image, the caption, or both. Utilizing a human-AI collaboration approach, we build the accompanying expert-verified dataset V-FLUTE, containing 6,027 {image, caption, label, explanation} instances spanning five diverse figurative phenomena: metaphors, similes, idioms, sarcasm, and humor. Through automatic evaluation, we find that VLMs struggle to generalize from literal to figurative meaning, particularly when it is present in images. Further, we identify common types of errors in VLM reasoning via human evaluation.
♻ ☆ Latent Inversion with Timestep-aware Sampling for Training-free Non-rigid Editing
Text-guided non-rigid editing involves complex edits for input images, such as changing motion or compositions within their surroundings. Since it requires manipulating the input structure, existing methods often struggle with preserving object identity and background, particularly when combined with Stable Diffusion. In this work, we propose a training-free approach for non-rigid editing with Stable Diffusion, aimed at improving the identity preservation quality without compromising editability. Our approach comprises three stages: text optimization, latent inversion, and timestep-aware text injection sampling. Inspired by the success of Imagic, we employ their text optimization for smooth editing. Then, we introduce latent inversion to preserve the input image's identity without additional model fine-tuning. To fully utilize the input reconstruction ability of latent inversion, we suggest timestep-aware text injection sampling. This effectively retains the structure of the input image by injecting the source text prompt in early sampling steps and then transitioning to the target prompt in subsequent sampling steps. This strategic approach seamlessly harmonizes with text optimization, facilitating complex non-rigid edits to the input without losing the original identity. We demonstrate the effectiveness of our method in terms of identity preservation, editability, and aesthetic quality through extensive experiments.
comment: This manuscript has been submitted to Pattern Recognition Letters
♻ ☆ Self-supervised Learning of LiDAR 3D Point Clouds via 2D-3D Neural Calibration
This paper introduces a novel self-supervised learning framework for enhancing 3D perception in autonomous driving scenes. Specifically, our approach, namely NCLR, focuses on 2D-3D neural calibration, a novel pretext task that estimates the rigid pose aligning camera and LiDAR coordinate systems. First, we propose the learnable transformation alignment to bridge the domain gap between image and point cloud data, converting features into a unified representation space for effective comparison and matching. Second, we identify the overlapping area between the image and point cloud with the fused features. Third, we establish dense 2D-3D correspondences to estimate the rigid pose. The framework not only learns fine-grained matching from points to pixels but also achieves alignment of the image and point cloud at a holistic level, understanding their relative pose. We demonstrate the efficacy of NCLR by applying the pre-trained backbone to downstream tasks, such as LiDAR-based 3D semantic segmentation, object detection, and panoptic segmentation. Comprehensive experiments on various datasets illustrate the superiority of NCLR over existing self-supervised methods. The results confirm that joint learning from different modalities significantly enhances the network's understanding abilities and effectiveness of learned representation. The code is publicly available at https://github.com/Eaphan/NCLR.
comment: Under review
♻ ☆ Dynamic Tuning Towards Parameter and Inference Efficiency for ViT Adaptation NeurIPS2024
Existing parameter-efficient fine-tuning (PEFT) methods have achieved significant success on vision transformers (ViTs) adaptation by improving parameter efficiency. However, the exploration of enhancing inference efficiency during adaptation remains underexplored. This limits the broader application of pre-trained ViT models, especially when the model is computationally extensive. In this paper, we propose Dynamic Tuning (DyT), a novel approach to improve both parameter and inference efficiency for ViT adaptation. Specifically, besides using the lightweight adapter modules, we propose a token dispatcher to distinguish informative tokens from less important ones, allowing the latter to dynamically skip the original block, thereby reducing the redundant computation during inference. Additionally, we explore multiple design variants to find the best practice of DyT. Finally, inspired by the mixture-of-experts (MoE) mechanism, we introduce an enhanced adapter to further boost the adaptation performance. We validate DyT across various tasks, including image/video recognition and semantic segmentation. For instance, DyT achieves superior performance compared to existing PEFT methods while evoking only 71% of their FLOPs on the VTAB-1K benchmark.
comment: Accepted to NeurIPS2024
♻ ☆ ScaleFlow++: Robust and Accurate Estimation of 3D Motion from Video
Perceiving and understanding 3D motion is a core technology in fields such as autonomous driving, robots, and motion prediction. This paper proposes a 3D motion perception method called ScaleFlow++ that is easy to generalize. With just a pair of RGB images, ScaleFlow++ can robustly estimate optical flow and motion-in-depth (MID). Most existing methods directly regress MID from two RGB frames or optical flow, resulting in inaccurate and unstable results. Our key insight is cross-scale matching, which extracts deep motion clues by matching objects in pairs of images at different scales. Unlike previous methods, ScaleFlow++ integrates optical flow and MID estimation into a unified architecture, estimating optical flow and MID end-to-end based on feature matching. Moreover, we also proposed modules such as global initialization network, global iterative optimizer, and hybrid training pipeline to integrate global motion information, reduce the number of iterations, and prevent overfitting during training. On KITTI, ScaleFlow++ achieved the best monocular scene flow estimation performance, reducing SF-all from 6.21 to 5.79. The evaluation of MID even surpasses RGBD-based methods. In addition, ScaleFlow++ has achieved stunning zero-shot generalization performance in both rigid and nonrigid scenes. Code is available at \url{https://github.com/HanLingsgjk/CSCV}.
comment: 14 pages; Previously this version appeared as arXiv:2409.12202 which was submitted as a new work by accident
♻ ☆ Semantic Token Reweighting for Interpretable and Controllable Text Embeddings in CLIP EMNLP 2024
A text encoder within Vision-Language Models (VLMs) like CLIP plays a crucial role in translating textual input into an embedding space shared with images, thereby facilitating the interpretative analysis of vision tasks through natural language. Despite the varying significance of different textual elements within a sentence depending on the context, efforts to account for variation of importance in constructing text embeddings have been lacking. We propose a framework of Semantic Token Reweighting to build Interpretable text embeddings (SToRI), which incorporates controllability as well. SToRI refines the text encoding process in CLIP by differentially weighting semantic elements based on contextual importance, enabling finer control over emphasis responsive to data-driven insights and user preferences. The efficacy of SToRI is demonstrated through comprehensive experiments on few-shot image classification and image retrieval tailored to user preferences.
comment: Accepted at EMNLP 2024 Findings
♻ ☆ Efficient and Effective Universal Adversarial Attack against Vision-Language Pre-training Models
Vision-language pre-training (VLP) models, trained on large-scale image-text pairs, have become widely used across a variety of downstream vision-and-language (V+L) tasks. This widespread adoption raises concerns about their vulnerability to adversarial attacks. Non-universal adversarial attacks, while effective, are often impractical for real-time online applications due to their high computational demands per data instance. Recently, universal adversarial perturbations (UAPs) have been introduced as a solution, but existing generator-based UAP methods are significantly time-consuming. To overcome the limitation, we propose a direct optimization-based UAP approach, termed DO-UAP, which significantly reduces resource consumption while maintaining high attack performance. Specifically, we explore the necessity of multimodal loss design and introduce a useful data augmentation strategy. Extensive experiments conducted on three benchmark VLP datasets, six popular VLP models, and three classical downstream tasks demonstrate the efficiency and effectiveness of DO-UAP. Specifically, our approach drastically decreases the time consumption by 23-fold while achieving a better attack performance.
comment: 11 pages
♻ ☆ Adaptive Convolutional Neural Network for Image Super-resolution
Convolutional neural networks can automatically learn features via deep network architectures and given input samples. However, the robustness of obtained models may face challenges in varying scenes. Bigger differences in network architecture are beneficial to extract more diversified structural information to strengthen the robustness of an obtained super-resolution model. In this paper, we proposed a adaptive convolutional neural network for image super-resolution (ADSRNet). To capture more information, ADSRNet is implemented by a heterogeneous parallel network. The upper network can enhance relation of context information, salient information relation of a kernel mapping and relations of shallow and deep layers to improve performance of image super-resolution. That can strengthen adaptability of an obtained super-resolution model for different scenes. The lower network utilizes a symmetric architecture to enhance relations of different layers to mine more structural information, which is complementary with a upper network for image super-resolution. The relevant experimental results show that the proposed ADSRNet is effective to deal with image resolving. Codes are obtained at https://github.com/hellloxiaotian/ADSRNet.
comment: 11pages, 7 figures
♻ ☆ Efficient Diffusion Models: A Comprehensive Survey from Principles to Practices
As one of the most popular and sought-after generative models in the recent years, diffusion models have sparked the interests of many researchers and steadily shown excellent advantage in various generative tasks such as image synthesis, video generation, molecule design, 3D scene rendering and multimodal generation, relying on their dense theoretical principles and reliable application practices. The remarkable success of these recent efforts on diffusion models comes largely from progressive design principles and efficient architecture, training, inference, and deployment methodologies. However, there has not been a comprehensive and in-depth review to summarize these principles and practices to help the rapid understanding and application of diffusion models. In this survey, we provide a new efficiency-oriented perspective on these existing efforts, which mainly focuses on the profound principles and efficient practices in architecture designs, model training, fast inference and reliable deployment, to guide further theoretical research, algorithm migration and model application for new scenarios in a reader-friendly way. \url{https://github.com/ponyzym/Efficient-DMs-Survey}
♻ ☆ Developing Generalist Foundation Models from a Multimodal Dataset for 3D Computed Tomography
While computer vision has achieved tremendous success with multimodal encoding and direct textual interaction with images via chat-based large language models, similar advancements in medical imaging AI, particularly in 3D imaging, have been limited due to the scarcity of comprehensive datasets. To address this critical gap, we introduce CT-RATE, the first dataset that pairs 3D medical images with corresponding textual reports. CT-RATE comprises 25,692 non-contrast 3D chest CT scans from 21,304 unique patients. Through various reconstructions, these scans are expanded to 50,188 volumes, totaling over 14.3 million 2D slices. Each scan is accompanied by its corresponding radiology report. Leveraging CT-RATE, we develop CT-CLIP, a CT-focused contrastive language-image pretraining framework designed for broad applications without the need for task-specific training. We demonstrate how CT-CLIP can be used in two tasks: multi-abnormality detection and case retrieval. Remarkably, in multi-abnormality detection, CT-CLIP outperforms state-of-the-art fully supervised models across all key metrics, effectively eliminating the need for manual annotation. In case retrieval, it efficiently retrieves relevant cases using either image or textual queries, thereby enhancing knowledge dissemination. By combining CT-CLIP's vision encoder with a pretrained large language model, we create CT-CHAT, a vision-language foundational chat model for 3D chest CT volumes. Finetuned on over 2.7 million question-answer pairs derived from the CT-RATE dataset, CT-CHAT surpasses other multimodal AI assistants, underscoring the necessity for specialized methods in 3D medical imaging. Collectively, the open-source release of CT-RATE, CT-CLIP, and CT-CHAT not only addresses critical challenges in 3D medical imaging but also lays the groundwork for future innovations in medical AI and improved patient care.
♻ ☆ Mixture of Experts Made Personalized: Federated Prompt Learning for Vision-Language Models
Prompt learning for pre-trained Vision-Language Models (VLMs) like CLIP has demonstrated potent applicability across diverse downstream tasks. This lightweight approach has quickly gained traction from federated learning (FL) researchers who seek to efficiently adapt VLMs to heterogeneous scenarios. However, current federated prompt learning methods are habitually restricted to the traditional FL paradigm, where the participating clients are generally only allowed to download a single globally aggregated model from the server. While justifiable for training full-sized models under federated settings, in this work, we argue that this paradigm is ill-suited for lightweight prompts. By facilitating the clients to download multiple pre-aggregated prompts as fixed non-local experts, we propose Personalized Federated Mixture of Adaptive Prompts (pFedMoAP), a novel FL framework that personalizes the prompt learning process through the lens of Mixture of Experts (MoE). pFedMoAP implements a local attention-based gating network that learns to generate enhanced text features for better alignment with local image data on the client, benefiting from both local and downloaded non-local adaptive prompt experts. The non-local experts are sparsely selected from a server-maintained pool, fostering collaborative learning across clients. To evaluate the proposed algorithm, we conduct extensive experiments across 9 datasets under various heterogeneous federated settings. The results show that pFedMoAP consistently outperforms the state-of-the-art alternatives, underscoring its efficacy in personalizing prompt learning for CLIP within the federated learning paradigm.
comment: 16 pages, 4 figures
♻ ☆ Interpret Your Decision: Logical Reasoning Regularization for Generalization in Visual Classification NeurIPS2024
Vision models excel in image classification but struggle to generalize to unseen data, such as classifying images from unseen domains or discovering novel categories. In this paper, we explore the relationship between logical reasoning and deep learning generalization in visual classification. A logical regularization termed L-Reg is derived which bridges a logical analysis framework to image classification. Our work reveals that L-Reg reduces the complexity of the model in terms of the feature distribution and classifier weights. Specifically, we unveil the interpretability brought by L-Reg, as it enables the model to extract the salient features, such as faces to persons, for classification. Theoretical analysis and experiments demonstrate that L-Reg enhances generalization across various scenarios, including multi-domain generalization and generalized category discovery. In complex real-world scenarios where images span unknown classes and unseen domains, L-Reg consistently improves generalization, highlighting its practical efficacy.
comment: Accepted by NeurIPS2024 as Spotlight
♻ ☆ A3D: Does Diffusion Dream about 3D Alignment?
We tackle the problem of text-driven 3D generation from a geometry alignment perspective. Given a set of text prompts, we aim to generate a collection of objects with semantically corresponding parts aligned across them. Recent methods based on Score Distillation have succeeded in distilling the knowledge from 2D diffusion models to high-quality representations of the 3D objects. These methods handle multiple text queries separately, and therefore the resulting objects have a high variability in object pose and structure. However, in some applications, such as 3D asset design, it may be desirable to obtain a set of objects aligned with each other. In order to achieve the alignment of the corresponding parts of the generated objects, we propose to embed these objects into a common latent space and optimize the continuous transitions between these objects. We enforce two kinds of properties of these transitions: smoothness of the transition and plausibility of the intermediate objects along the transition. We demonstrate that both of these properties are essential for good alignment. We provide several practical scenarios that benefit from alignment between the objects, including 3D editing and object hybridization, and experimentally demonstrate the effectiveness of our method. https://voyleg.github.io/a3d/
♻ ☆ Gaussian Primitives for Deformable Image Registration
Deformable Image Registration (DIR) is essential for aligning medical images that exhibit anatomical variations, facilitating applications such as disease tracking and radiotherapy planning. While classical iterative methods and deep learning approaches have achieved success in DIR, they are often hindered by computational inefficiency or poor generalization. In this paper, we introduce GaussianDIR, a novel, case-specific optimization DIR method inspired by 3D Gaussian splatting. In general, GaussianDIR represents image deformations using a sparse set of mobile and flexible Gaussian primitives, each defined by a center position, covariance, and local rigid transformation. This compact and explicit representation reduces noise and computational overhead while improving interpretability. Furthermore, the movement of individual voxel is derived via blending the local rigid transformation of the neighboring Gaussian primitives. By this, GaussianDIR captures both global smoothness and local rigidity as well as reduces the computational burden. To address varying levels of deformation complexity, GaussianDIR also integrates an adaptive density control mechanism that dynamically adjusts the density of Gaussian primitives. Additionally, we employ multi-scale Gaussian primitives to capture both coarse and fine deformations, reducing optimization to local minima. Experimental results on brain MRI, lung CT, and cardiac MRI datasets demonstrate that GaussianDIR outperforms existing DIR methods in both accuracy and efficiency, highlighting its potential for clinical applications. Finally, as a training-free approach, it challenges the stereotype that iterative methods are inherently slow and transcend the limitations of poor generalization.
♻ ☆ BroadWay: Boost Your Text-to-Video Generation Model in a Training-free Way
The text-to-video (T2V) generation models, offering convenient visual creation, have recently garnered increasing attention. Despite their substantial potential, the generated videos may present artifacts, including structural implausibility, temporal inconsistency, and a lack of motion, often resulting in near-static video. In this work, we have identified a correlation between the disparity of temporal attention maps across different blocks and the occurrence of temporal inconsistencies. Additionally, we have observed that the energy contained within the temporal attention maps is directly related to the magnitude of motion amplitude in the generated videos. Based on these observations, we present BroadWay, a training-free method to improve the quality of text-to-video generation without introducing additional parameters, augmenting memory or sampling time. Specifically, BroadWay is composed of two principal components: 1) Temporal Self-Guidance improves the structural plausibility and temporal consistency of generated videos by reducing the disparity between the temporal attention maps across various decoder blocks. 2) Fourier-based Motion Enhancement enhances the magnitude and richness of motion by amplifying the energy of the map. Extensive experiments demonstrate that BroadWay significantly improves the quality of text-to-video generation with negligible additional cost.
♻ ☆ VrdONE: One-stage Video Visual Relation Detection
Video Visual Relation Detection (VidVRD) focuses on understanding how entities interact over time and space in videos, a key step for gaining deeper insights into video scenes beyond basic visual tasks. Traditional methods for VidVRD, challenged by its complexity, typically split the task into two parts: one for identifying what relation categories are present and another for determining their temporal boundaries. This split overlooks the inherent connection between these elements. Addressing the need to recognize entity pairs' spatiotemporal interactions across a range of durations, we propose VrdONE, a streamlined yet efficacious one-stage model. VrdONE combines the features of subjects and objects, turning predicate detection into 1D instance segmentation on their combined representations. This setup allows for both relation category identification and binary mask generation in one go, eliminating the need for extra steps like proposal generation or post-processing. VrdONE facilitates the interaction of features across various frames, adeptly capturing both short-lived and enduring relations. Additionally, we introduce the Subject-Object Synergy (SOS) module, enhancing how subjects and objects perceive each other before combining. VrdONE achieves state-of-the-art performances on the VidOR benchmark and ImageNet-VidVRD, showcasing its superior capability in discerning relations across different temporal scales. The code is available at https://github.com/lucaspk512/vrdone.
comment: 12 pages, 8 figures, accepted by ACM Multimedia 2024
♻ ☆ Mini-Omni2: Towards Open-source GPT-4o with Vision, Speech and Duplex Capabilities
GPT-4o, an all-encompassing model, represents a milestone in the development of large multi-modal language models. It can understand visual, auditory, and textual modalities, directly output audio, and support flexible duplex interaction. Models from the open-source community often achieve some functionalities of GPT-4o, such as visual understanding and voice chat. Nevertheless, training a unified model that incorporates all modalities is challenging due to the complexities of multi-modal data, intricate model architectures, and training processes. In this paper, we introduce Mini-Omni2, a visual-audio assistant capable of providing real-time, end-to-end voice responses to visoin and audio queries. By integrating pretrained visual and auditory encoders, Mini-Omni2 maintains performance in individual modalities. We propose a three-stage training process to align modalities, allowing the language model to handle multi-modal inputs and outputs after training on a limited dataset. For interaction, we introduce a command-based interruption mechanism, enabling more flexible interaction with users. To the best of our knowledge, Mini-Omni2 is one of the closest reproductions of GPT-4o, which have similar form of functionality, and we hope it can offer valuable insights for subsequent research.
comment: 13 pages, 6 figures
♻ ☆ DNTextSpotter: Arbitrary-Shaped Scene Text Spotting via Improved Denoising Training
More and more end-to-end text spotting methods based on Transformer architecture have demonstrated superior performance. These methods utilize a bipartite graph matching algorithm to perform one-to-one optimal matching between predicted objects and actual objects. However, the instability of bipartite graph matching can lead to inconsistent optimization targets, thereby affecting the training performance of the model. Existing literature applies denoising training to solve the problem of bipartite graph matching instability in object detection tasks. Unfortunately, this denoising training method cannot be directly applied to text spotting tasks, as these tasks need to perform irregular shape detection tasks and more complex text recognition tasks than classification. To address this issue, we propose a novel denoising training method (DNTextSpotter) for arbitrary-shaped text spotting. Specifically, we decompose the queries of the denoising part into noised positional queries and noised content queries. We use the four Bezier control points of the Bezier center curve to generate the noised positional queries. For the noised content queries, considering that the output of the text in a fixed positional order is not conducive to aligning position with content, we employ a masked character sliding method to initialize noised content queries, thereby assisting in the alignment of text content and position. To improve the model's perception of the background, we further utilize an additional loss function for background characters classification in the denoising training part.Although DNTextSpotter is conceptually simple, it outperforms the state-of-the-art methods on four benchmarks (Total-Text, SCUT-CTW1500, ICDAR15, and Inverse-Text), especially yielding an improvement of 11.3% against the best approach in Inverse-Text dataset.
comment: Accepted by ACM'MM2024
♻ ☆ On Large Uni- and Multi-modal Models for Unsupervised Classification of Social Media Images: Nature's Contribution to People as a case study
Social media images have proven to be a valuable source of information for understanding human interactions with important subjects such as cultural heritage, biodiversity, and nature, among others. The task of grouping such images into a number of semantically meaningful clusters without labels is challenging due to the high diversity and complex nature of the visual content in addition to their large volume. On the other hand, recent advances in Large Visual Models (LVMs), Large Language Models (LLMs), and Large Visual Language Models (LVLMs) provide an important opportunity to explore new productive and scalable solutions. This work proposes, analyzes, and compares various approaches based on one or more state-of-the-art LVM, LLM, and LVLM, for mapping social media images into a number of predefined classes. As a case study, we consider the problem of understanding the interactions between humans and nature, also known as Nature's Contribution to People or Cultural Ecosystem Services (CES). Our experiments show that the highest-performing approaches, with accuracy above 95%, still require the creation of a small labeled dataset. These include the fine-tuned LVM DINOv2 and the LVLM LLaVA-1.5 combined with a fine-tuned LLM. The top fully unsupervised approaches, achieving accuracy above 84%, are the LVLMs, specifically the proprietary GPT-4 model and the public LLaVA-1.5 model. Additionally, the LVM DINOv2, when applied in a 10-shot learning setup, delivered competitive results with an accuracy of 83.99%, closely matching the performance of the LVLM LLaVA-1.5.
comment: 17 pages, 9 figures
♻ ☆ Topological reconstruction of sampled surfaces via Morse theory
In this work, we study the perception problem for sampled surfaces (possibly with boundary) using tools from computational topology, specifically, how to identify their underlying topology starting from point-cloud samples in space, such as those obtained with 3D scanners. We present a reconstruction algorithm based on a careful topological study of the point sample that allows us to obtain a cellular decomposition of it using a Morse function. No triangulation or local implicit equations are used as intermediate steps, avoiding in this way reconstruction-induced artifices. The algorithm can be run without any prior knowledge of the surface topology, density or regularity of the point-sample. The results consist of a piece-wise decomposition of the given surface as a union of Morse cells (i.e. topological disks), suitable for tasks such as mesh-independent reparametrization or noise-filtering, and a small-rank cellular complex determining the topology of the surface. The algorithm, which we test with several real and synthetic surfaces, can be applied to smooth surfaces with or without boundary, embedded in an ambient space of any dimension.
comment: 39 pages, 17 figures, 1 table, 1 algorithm, 1 appendix
♻ ☆ MFC-Bench: Benchmarking Multimodal Fact-Checking with Large Vision-Language Models
Large vision-language models (LVLMs) have significantly improved multimodal reasoning tasks, such as visual question answering and image captioning. These models embed multimodal facts within their parameters, rather than relying on external knowledge bases to store factual information explicitly. However, the content discerned by LVLMs may deviate from actual facts due to inherent bias or incorrect inference. To address this issue, we introduce MFC-Bench, a rigorous and comprehensive benchmark designed to evaluate the factual accuracy of LVLMs across three stages of verdict prediction for MFC: Manipulation, Out-of-Context, and Veracity Classification. Through our evaluation on MFC-Bench, we benchmarked a dozen diverse and representative LVLMs, uncovering that current models still fall short in multimodal fact-checking and demonstrate insensitivity to various forms of manipulated content. We hope that MFC-Bench could raise attention to the trustworthy AI potentially assisted by LVLMs in the future. The MFC-Bench and accompanying resources are publicly accessible at https://github.com/wskbest/MFC-Bench, contributing to ongoing research in the multimodal fact-checking field.
comment: 28 pages, 9 figures
♻ ☆ Free Video-LLM: Prompt-guided Visual Perception for Efficient Training-free Video LLMs
Vision-language large models have achieved remarkable success in various multi-modal tasks, yet applying them to video understanding remains challenging due to the inherent complexity and computational demands of video data. While training-based video-LLMs deliver high performance, they often require substantial resources for training and inference. Conversely, training-free approaches offer a more efficient alternative by adapting pre-trained image-LLMs models for video tasks without additional training, but they face inference efficiency bottlenecks due to the large number of visual tokens generated from video frames. In this work, we present a novel prompt-guided visual perception framework (abbreviated as Free Video-LLM) for efficient inference of training-free video LLMs. The proposed framework decouples spatial-temporal dimension and performs temporal frame sampling and spatial RoI cropping respectively based on task-specific prompts. Our method effectively reduces the number of visual tokens while maintaining high performance across multiple video question-answering benchmarks. Extensive experiments demonstrate that our approach achieves competitive results with significantly fewer tokens, offering an optimal trade-off between accuracy and computational efficiency compared to state-of-the-art video LLMs. The code will be available at https://github.com/contrastive/FreeVideoLLM.
comment: Tech report
♻ ☆ Enhancing Robustness of Vision-Language Models through Orthogonality Learning and Self-Regularization
Efficient fine-tuning of vision-language models (VLMs) like CLIP for specific downstream tasks is gaining significant attention. Previous works primarily focus on prompt learning to adapt the CLIP into a variety of downstream tasks, however, suffering from task overfitting when fine-tuned on a small data set. In this paper, we introduce an orthogonal fine-tuning method for efficiently fine-tuning pretrained weights and enabling enhanced robustness and generalization, while a self-regularization strategy is further exploited to maintain the stability in terms of zero-shot generalization of VLMs, dubbed OrthSR. Specifically, trainable orthogonal matrices are injected seamlessly into the transformer architecture and enforced with orthogonality constraint during the training, benefiting from the norm-preserving property and thus leading to stable and faster convergence, while keeping the pre-trained weights frozen. To alleviate deviation from fine-tuning, a self-regularization strategy is further employed to retain the generalization of the model during the training within a bypass manner. In addition, to enrich the sample diversity for downstream tasks under the small dataset scenario, we first explore attentive CutOut data augmentation to boost the efficient fine-tuning, leading to better model fitting capacity for specific downstream task. Then we support the theoretical analysis on how our approach improves the specific downstream performance and maintains the generalizability. For the first time, we revisit the CLIP and CoOp with our method to effectively improve the model on few-shot image classficiation scenario on par with the elaborated prompt learning methods.
♻ ☆ Reverse Stable Diffusion: What prompt was used to generate this image?
Text-to-image diffusion models have recently attracted the interest of many researchers, and inverting the diffusion process can play an important role in better understanding the generative process and how to engineer prompts in order to obtain the desired images. To this end, we study the task of predicting the prompt embedding given an image generated by a generative diffusion model. We consider a series of white-box and black-box models (with and without access to the weights of the diffusion network) to deal with the proposed task. We propose a novel learning framework comprising a joint prompt regression and multi-label vocabulary classification objective that generates improved prompts. To further improve our method, we employ a curriculum learning procedure that promotes the learning of image-prompt pairs with lower labeling noise (i.e. that are better aligned). We conduct experiments on the DiffusionDB data set, predicting text prompts from images generated by Stable Diffusion. In addition, we make an interesting discovery: training a diffusion model on the prompt generation task can make the model generate images that are much better aligned with the input prompts, when the model is directly reused for text-to-image generation. Our code is publicly available for download at https://github.com/CroitoruAlin/Reverse-Stable-Diffusion.
comment: Accepted for publication in Computer Vision and Image Understanding
♻ ☆ Instruction-Guided Visual Masking NeurIPS 2024
Instruction following is crucial in contemporary LLM. However, when extended to multimodal setting, it often suffers from misalignment between specific textual instruction and targeted local region of an image. To achieve more accurate and nuanced multimodal instruction following, we introduce Instruction-guided Visual Masking (IVM), a new versatile visual grounding model that is compatible with diverse multimodal models, such as LMM and robot model. By constructing visual masks for instruction-irrelevant regions, IVM-enhanced multimodal models can effectively focus on task-relevant image regions to better align with complex instructions. Specifically, we design a visual masking data generation pipeline and create an IVM-Mix-1M dataset with 1 million image-instruction pairs. We further introduce a new learning technique, Discriminator Weighted Supervised Learning (DWSL) for preferential IVM training that prioritizes high-quality data samples. Experimental results on generic multimodal tasks such as VQA and embodied robotic control demonstrate the versatility of IVM, which as a plug-and-play tool, significantly boosts the performance of diverse multimodal models, yielding new state-of-the-art results across challenging multimodal benchmarks. Code, model and data are available at https://github.com/2toinf/IVM.
comment: NeurIPS 2024
♻ ☆ InterACT: Inter-dependency Aware Action Chunking with Hierarchical Attention Transformers for Bimanual Manipulation
Bimanual manipulation presents unique challenges compared to unimanual tasks due to the complexity of coordinating two robotic arms. In this paper, we introduce InterACT: Inter-dependency aware Action Chunking with Hierarchical Attention Transformers, a novel imitation learning framework designed specifically for bimanual manipulation. InterACT leverages hierarchical attention mechanisms to effectively capture inter-dependencies between dual-arm joint states and visual inputs. The framework comprises a Hierarchical Attention Encoder, which processes multi-modal inputs through segment-wise and cross-segment attention mechanisms, and a Multi-arm Decoder that generates each arm's action predictions in parallel, while sharing information between the arms through synchronization blocks by providing the other arm's intermediate output as context. Our experiments, conducted on various simulated and real-world bimanual manipulation tasks, demonstrate that InterACT outperforms existing methods. Detailed ablation studies further validate the significance of key components, including the impact of CLS tokens, cross-segment encoders, and synchronization blocks on task performance. We provide supplementary materials and videos on our project page.
comment: Accepted at Conference on Robot Learning (CoRL) 2024
♻ ☆ No Bells, Just Whistles: Sports Field Registration by Leveraging Geometric Properties CVPR
Broadcast sports field registration is traditionally addressed as a homography estimation task, mapping the visible image area to a planar field model, predominantly focusing on the main camera shot. Addressing the shortcomings of previous approaches, we propose a novel calibration pipeline enabling camera calibration using a 3D soccer field model and extending the process to assess the multiple-view nature of broadcast videos. Our approach begins with a keypoint generation pipeline derived from SoccerNet dataset annotations, leveraging the geometric properties of the court. Subsequently, we execute classical camera calibration through DLT algorithm in a minimalist fashion, without further refinement. Through extensive experimentation on real-world soccer broadcast datasets such as SoccerNet-Calibration, WorldCup 2014 and TS- WorldCup, our method demonstrates superior performance in both multiple- and single-view 3D camera calibration while maintaining competitive results in homography estimation compared to state-of-the-art techniques.
comment: Accepted in CVPRW 2024
♻ ☆ Active Fake: DeepFake Camouflage
DeepFake technology has gained significant attention due to its ability to manipulate facial attributes with high realism, raising serious societal concerns. Face-Swap DeepFake is the most harmful among these techniques, which fabricates behaviors by swapping original faces with synthesized ones. Existing forensic methods, primarily based on Deep Neural Networks (DNNs), effectively expose these manipulations and have become important authenticity indicators. However, these methods mainly concentrate on capturing the blending inconsistency in DeepFake faces, raising a new security issue, termed Active Fake, emerges when individuals intentionally create blending inconsistency in their authentic videos to evade responsibility. This tactic is called DeepFake Camouflage. To achieve this, we introduce a new framework for creating DeepFake camouflage that generates blending inconsistencies while ensuring imperceptibility, effectiveness, and transferability. This framework, optimized via an adversarial learning strategy, crafts imperceptible yet effective inconsistencies to mislead forensic detectors. Extensive experiments demonstrate the effectiveness and robustness of our method, highlighting the need for further research in active fake detection.
♻ ☆ PIVOT-R: Primitive-Driven Waypoint-Aware World Model for Robotic Manipulation NeurIPS 2024
Language-guided robotic manipulation is a challenging task that requires an embodied agent to follow abstract user instructions to accomplish various complex manipulation tasks. Previous work trivially fitting the data without revealing the relation between instruction and low-level executable actions, these models are prone to memorizing the surficial pattern of the data instead of acquiring the transferable knowledge, and thus are fragile to dynamic environment changes. To address this issue, we propose a PrIrmitive-driVen waypOinT-aware world model for Robotic manipulation (PIVOT-R) that focuses solely on the prediction of task-relevant waypoints. Specifically, PIVOT-R consists of a Waypoint-aware World Model (WAWM) and a lightweight action prediction module. The former performs primitive action parsing and primitive-driven waypoint prediction, while the latter focuses on decoding low-level actions. Additionally, we also design an asynchronous hierarchical executor (AHE), which can use different execution frequencies for different modules of the model, thereby helping the model reduce computational redundancy and improve model execution efficiency. Our PIVOT-R outperforms state-of-the-art (SoTA) open-source models on the SeaWave benchmark, achieving an average relative improvement of 19.45% across four levels of instruction tasks. Moreover, compared to the synchronously executed PIVOT-R, the execution efficiency of PIVOT-R with AHE is increased by 28-fold, with only a 2.9% drop in performance. These results provide compelling evidence that our PIVOT-R can significantly improve both the performance and efficiency of robotic manipulation.
comment: Accepted to NeurIPS 2024
♻ ☆ Progressive Retinal Image Registration via Global and Local Deformable Transformations
Retinal image registration plays an important role in the ophthalmological diagnosis process. Since there exist variances in viewing angles and anatomical structures across different retinal images, keypoint-based approaches become the mainstream methods for retinal image registration thanks to their robustness and low latency. These methods typically assume the retinal surfaces are planar, and adopt feature matching to obtain the homography matrix that represents the global transformation between images. Yet, such a planar hypothesis inevitably introduces registration errors since retinal surface is approximately curved. This limitation is more prominent when registering image pairs with significant differences in viewing angles. To address this problem, we propose a hybrid registration framework called HybridRetina, which progressively registers retinal images with global and local deformable transformations. For that, we use a keypoint detector and a deformation network called GAMorph to estimate the global transformation and local deformable transformation, respectively. Specifically, we integrate multi-level pixel relation knowledge to guide the training of GAMorph. Additionally, we utilize an edge attention module that includes the geometric priors of the images, ensuring the deformation field focuses more on the vascular regions of clinical interest. Experiments on two widely-used datasets, FIRE and FLoRI21, show that our proposed HybridRetina significantly outperforms some state-of-the-art methods. The code is available at https://github.com/lyp-deeplearning/awesome-retinal-registration.
comment: Accepted at BIBM 2024
♻ ☆ Deciphering Cross-Modal Alignment in Large Vision-Language Models with Modality Integration Rate
We present the Modality Integration Rate (MIR), an effective, robust, and generalized metric to indicate the multi-modal pre-training quality of Large Vision Language Models (LVLMs). Large-scale pre-training plays a critical role in building capable LVLMs, while evaluating its training quality without the costly supervised fine-tuning stage is under-explored. Loss, perplexity, and in-context evaluation results are commonly used pre-training metrics for Large Language Models (LLMs), while we observed that these metrics are less indicative when aligning a well-trained LLM with a new modality. Due to the lack of proper metrics, the research of LVLMs in the critical pre-training stage is hindered greatly, including the training data choice, efficient module design, etc. In this paper, we propose evaluating the pre-training quality from the inter-modal distribution distance perspective and present MIR, the Modality Integration Rate, which is 1) \textbf{Effective} to represent the pre-training quality and show a positive relation with the benchmark performance after supervised fine-tuning. 2) \textbf{Robust} toward different training/evaluation data. 3) \textbf{Generalize} across training configurations and architecture choices. We conduct a series of pre-training experiments to explore the effectiveness of MIR and observe satisfactory results that MIR is indicative about training data selection, training strategy schedule, and model architecture design to get better pre-training results. We hope MIR could be a helpful metric for building capable LVLMs and inspire the following research about modality alignment in different areas. Our code is at: https://github.com/shikiw/Modality-Integration-Rate.
comment: Project page: https://github.com/shikiw/Modality-Integration-Rate
♻ ☆ Hyper-YOLO: When Visual Object Detection Meets Hypergraph Computation
We introduce Hyper-YOLO, a new object detection method that integrates hypergraph computations to capture the complex high-order correlations among visual features. Traditional YOLO models, while powerful, have limitations in their neck designs that restrict the integration of cross-level features and the exploitation of high-order feature interrelationships. To address these challenges, we propose the Hypergraph Computation Empowered Semantic Collecting and Scattering (HGC-SCS) framework, which transposes visual feature maps into a semantic space and constructs a hypergraph for high-order message propagation. This enables the model to acquire both semantic and structural information, advancing beyond conventional feature-focused learning. Hyper-YOLO incorporates the proposed Mixed Aggregation Network (MANet) in its backbone for enhanced feature extraction and introduces the Hypergraph-Based Cross-Level and Cross-Position Representation Network (HyperC2Net) in its neck. HyperC2Net operates across five scales and breaks free from traditional grid structures, allowing for sophisticated high-order interactions across levels and positions. This synergy of components positions Hyper-YOLO as a state-of-the-art architecture in various scale models, as evidenced by its superior performance on the COCO dataset. Specifically, Hyper-YOLO-N significantly outperforms the advanced YOLOv8-N and YOLOv9-T with 12\% $\text{AP}^{val}$ and 9\% $\text{AP}^{val}$ improvements. The source codes are at ttps://github.com/iMoonLab/Hyper-YOLO.
♻ ☆ Depth Estimation From Monocular Images With Enhanced Encoder-Decoder Architecture
Estimating depth from a single 2D image is a challenging task because of the need for stereo or multi-view data, which normally provides depth information. This paper deals with this challenge by introducing a novel deep learning-based approach using an encoder-decoder architecture, where the Inception-ResNet-v2 model is utilized as the encoder. According to the available literature, this is the first instance of using Inception-ResNet-v2 as an encoder for monocular depth estimation, illustrating better performance than previous models. The use of Inception-ResNet-v2 enables our model to capture complex objects and fine-grained details effectively that are generally difficult to predict. Besides, our model incorporates multi-scale feature extraction to enhance depth prediction accuracy across different kinds of object sizes and distances. We propose a composite loss function consisting of depth loss, gradient edge loss, and SSIM loss, where the weights are fine-tuned to optimize the weighted sum, ensuring better balance across different aspects of depth estimation. Experimental results on the NYU Depth V2 dataset show that our model achieves state-of-the-art performance, with an ARE of 0.064, RMSE of 0.228, and accuracy ($\delta$ $<1.25$) of 89.3%. These metrics demonstrate that our model effectively predicts depth, even in challenging circumstances, providing a scalable solution for real-world applications in robotics, 3D reconstruction, and augmented reality.
♻ ☆ Generative Models: What Do They Know? Do They Know Things? Let's Find Out!
Generative models excel at mimicking real scenes, suggesting they might inherently encode important intrinsic scene properties. In this paper, we aim to explore the following key questions: (1) What intrinsic knowledge do generative models like GANs, Autoregressive models, and Diffusion models encode? (2) Can we establish a general framework to recover intrinsic representations from these models, regardless of their architecture or model type? (3) How minimal can the required learnable parameters and labeled data be to successfully recover this knowledge? (4) Is there a direct link between the quality of a generative model and the accuracy of the recovered scene intrinsics? Our findings indicate that a small Low-Rank Adaptators (LoRA) can recover intrinsic images-depth, normals, albedo and shading-across different generators (Autoregressive, GANs and Diffusion) while using the same decoder head that generates the image. As LoRA is lightweight, we introduce very few learnable parameters (as few as 0.04% of Stable Diffusion model weights for a rank of 2), and we find that as few as 250 labeled images are enough to generate intrinsic images with these LoRA modules. Finally, we also show a positive correlation between the generative model's quality and the accuracy of the recovered intrinsics through control experiments.
comment: https://intrinsic-lora.github.io/
♻ ☆ Mini-Splatting: Representing Scenes with a Constrained Number of Gaussians
In this study, we explore the challenge of efficiently representing scenes with a constrained number of Gaussians. Our analysis shifts from traditional graphics and 2D computer vision to the perspective of point clouds, highlighting the inefficient spatial distribution of Gaussian representation as a key limitation in model performance. To address this, we introduce strategies for densification including blur split and depth reinitialization, and simplification through intersection preserving and sampling. These techniques reorganize the spatial positions of the Gaussians, resulting in significant improvements across various datasets and benchmarks in terms of rendering quality, resource consumption, and storage compression. Our Mini-Splatting integrates seamlessly with the original rasterization pipeline, providing a strong baseline for future research in Gaussian-Splatting-based works. \href{https://github.com/fatPeter/mini-splatting}{Code is available}.
♻ ☆ See Where You Read with Eye Gaze Tracking and Large Language Model
Losing track of reading progress during line switching can be frustrating. Eye gaze tracking technology offers a potential solution by highlighting read paragraphs, aiding users in avoiding wrong line switches. However, the gap between gaze tracking accuracy (2-3 cm) and text line spacing (3-5 mm) makes direct application impractical. Existing methods leverage the linear reading pattern but fail during jump reading. This paper presents a reading tracking and highlighting system that supports both linear and jump reading. Based on experimental insights from the gaze nature study of 16 users, two gaze error models are designed to enable both jump reading detection and relocation. The system further leverages the large language model's contextual perception capability in aiding reading tracking. A reading tracking domain-specific line-gaze alignment opportunity is also exploited to enable dynamic and frequent calibration of the gaze results. Controlled experiments demonstrate reliable linear reading tracking, as well as 84% accuracy in tracking jump reading. Furthermore, real field tests with 18 volunteers demonstrated the system's effectiveness in tracking and highlighting read paragraphs, improving reading efficiency, and enhancing user experience.
comment: 9 pages
♻ ☆ AIC MLLM: Autonomous Interactive Correction MLLM for Robust Robotic Manipulation
The ability to reflect on and correct failures is crucial for robotic systems to interact stably with real-life objects. Observing the generalization and reasoning capabilities of Multimodal Large Language Models (MLLMs), previous approaches have aimed to utilize these models to enhance robotic systems accordingly. However, these methods typically focus on high-level planning corrections using an additional MLLM, with limited utilization of failed samples to correct low-level contact poses which is particularly prone to occur during articulated object manipulation. To address this gap, we propose an Autonomous Interactive Correction (AIC) MLLM, which makes use of previous low-level interaction experiences to correct SE(3) pose predictions for articulated object. Specifically, AIC MLLM is initially fine-tuned to acquire both pose prediction and feedback prompt comprehension abilities. We design two types of prompt instructions for interactions with objects: 1) visual masks to highlight unmovable parts for position correction, and 2) textual descriptions to indicate potential directions for rotation correction. During inference, a Feedback Information Extraction module is introduced to recognize the failure cause, allowing AIC MLLM to adaptively correct the pose prediction using the corresponding prompts. To further enhance manipulation stability, we devise a Test Time Adaptation strategy that enables AIC MLLM to better adapt to the current scene configuration. Finally, extensive experiments are conducted in both simulated and real-world environments to evaluate the proposed method. The results demonstrate that our AIC MLLM can efficiently correct failure samples by leveraging interaction experience prompts. Our project website is https://sites.google.com/view/aic-mllm.
♻ ☆ Key-Grid: Unsupervised 3D Keypoints Detection using Grid Heatmap Features
Detecting 3D keypoints with semantic consistency is widely used in many scenarios such as pose estimation, shape registration and robotics. Currently, most unsupervised 3D keypoint detection methods focus on the rigid-body objects. However, when faced with deformable objects, the keypoints they identify do not preserve semantic consistency well. In this paper, we introduce an innovative unsupervised keypoint detector Key-Grid for both the rigid-body and deformable objects, which is an autoencoder framework. The encoder predicts keypoints and the decoder utilizes the generated keypoints to reconstruct the objects. Unlike previous work, we leverage the identified keypoint in formation to form a 3D grid feature heatmap called grid heatmap, which is used in the decoder section. Grid heatmap is a novel concept that represents the latent variables for grid points sampled uniformly in the 3D cubic space, where these variables are the shortest distance between the grid points and the skeleton connected by keypoint pairs. Meanwhile, we incorporate the information from each layer of the encoder into the decoder section. We conduct an extensive evaluation of Key-Grid on a list of benchmark datasets. Key-Grid achieves the state-of-the-art performance on the semantic consistency and position accuracy of keypoints. Moreover, we demonstrate the robustness of Key-Grid to noise and downsampling. In addition, we achieve SE-(3) invariance of keypoints though generalizing Key-Grid to a SE(3)-invariant backbone.
♻ ☆ MERLIN: Multimodal Embedding Refinement via LLM-based Iterative Navigation for Text-Video Retrieval-Rerank Pipeline EMNLP 2024
The rapid expansion of multimedia content has made accurately retrieving relevant videos from large collections increasingly challenging. Recent advancements in text-video retrieval have focused on cross-modal interactions, large-scale foundation model training, and probabilistic modeling, yet often neglect the crucial user perspective, leading to discrepancies between user queries and the content retrieved. To address this, we introduce MERLIN (Multimodal Embedding Refinement via LLM-based Iterative Navigation), a novel, training-free pipeline that leverages Large Language Models (LLMs) for iterative feedback learning. MERLIN refines query embeddings from a user perspective, enhancing alignment between queries and video content through a dynamic question answering process. Experimental results on datasets like MSR-VTT, MSVD, and ActivityNet demonstrate that MERLIN substantially improves Recall@1, outperforming existing systems and confirming the benefits of integrating LLMs into multimodal retrieval systems for more responsive and context-aware multimedia retrieval.
comment: EMNLP 2024 Industry Track Accepted (Camera-Ready Version)
♻ ☆ Tri-Cam: Practical Eye Gaze Tracking via Camera Network
As human eyes serve as conduits of rich information, unveiling emotions, intentions, and even aspects of an individual's health and overall well-being, gaze tracking also enables various human-computer interaction applications, as well as insights in psychological and medical research. However, existing gaze tracking solutions fall short at handling free user movement, and also require laborious user effort in system calibration. We introduce Tri-Cam, a practical deep learning-based gaze tracking system using three affordable RGB webcams. It features a split network structure for efficient training, as well as designated network designs to handle the separated gaze tracking tasks. Tri-Cam is also equipped with an implicit calibration module, which makes use of mouse click opportunities to reduce calibration overhead on the user's end. We evaluate Tri-Cam against Tobii, the state-of-the-art commercial eye tracker, achieving comparable accuracy, while supporting a wider free movement area. In conclusion, Tri-Cam provides a user-friendly, affordable, and robust gaze tracking solution that could practically enable various applications.
comment: 12 pages
♻ ☆ ViLReF: An Expert Knowledge Enabled Vision-Language Retinal Foundation Model
Subtle semantic differences in retinal image and text data present great challenges for pre-training visual-language models. Moreover, false negative samples, i.e., image-text pairs having the same semantics but incorrectly regarded as negatives, disrupt the visual-language pre-training process and affect the model's learning ability. This work aims to develop a retinal foundation model, called ViLReF, by pre-training on a paired dataset comprising 451,956 retinal images and corresponding diagnostic text reports. In our vision-language pre-training strategy, we leverage expert knowledge to facilitate the extraction of labels and propose a novel constraint, the Weighted Similarity Coupling Loss, to adjust the speed of pushing sample pairs further apart dynamically within the feature space. Furthermore, we employ a batch expansion module with dynamic memory queues, maintained by momentum encoders, to supply extra samples and compensate for the vacancies caused by eliminating false negatives. Extensive experiments are conducted on multiple datasets for downstream classification and segmentation tasks. The experimental results demonstrate the powerful zero-shot and transfer learning capabilities of ViLReF, verifying the effectiveness of our pre-training strategy. Our ViLReF model is available at: https://github.com/T6Yang/ViLReF.
♻ ☆ Scaling Up Personalized Image Aesthetic Assessment via Task Vector Customization ECCV 2024
The task of personalized image aesthetic assessment seeks to tailor aesthetic score prediction models to match individual preferences with just a few user-provided inputs. However, the scalability and generalization capabilities of current approaches are considerably restricted by their reliance on an expensive curated database. To overcome this long-standing scalability challenge, we present a unique approach that leverages readily available databases for general image aesthetic assessment and image quality assessment. Specifically, we view each database as a distinct image score regression task that exhibits varying degrees of personalization potential. By determining optimal combinations of task vectors, known to represent specific traits of each database, we successfully create personalized models for individuals. This approach of integrating multiple models allows us to harness a substantial amount of data. Our extensive experiments demonstrate the effectiveness of our approach in generalizing to previously unseen domains-a challenge previous approaches have struggled to achieve-making it highly applicable to real-world scenarios. Our novel approach significantly advances the field by offering scalable solutions for personalized aesthetic assessment and establishing high standards for future research. https://yeolj00.github.io/personal-projects/personalized-aesthetics/
comment: ECCV 2024
♻ ☆ AnimateLCM: Computation-Efficient Personalized Style Video Generation without Personalized Video Data SIGGRAPH
This paper introduces an effective method for computation-efficient personalized style video generation without requiring access to any personalized video data. It reduces the necessary generation time of similarly sized video diffusion models from 25 seconds to around 1 second while maintaining the same level of performance. The method's effectiveness lies in its dual-level decoupling learning approach: 1) separating the learning of video style from video generation acceleration, which allows for personalized style video generation without any personalized style video data, and 2) separating the acceleration of image generation from the acceleration of video motion generation, enhancing training efficiency and mitigating the negative effects of low-quality video data.
comment: Accepted as a Short Paper by SIGGRAPH ASIA 2024 Technical Communications. This is a short version of the original work. Project Page: https://animatelcm.github.io/
♻ ☆ Vision-Based Adaptive Robotics for Autonomous Surface Crack Repair
Surface cracks in infrastructure can lead to significant deterioration and costly maintenance if not efficiently repaired. Manual repair methods are labor-intensive, time-consuming, and imprecise and thus difficult to scale to large areas. While advancements in robotic perception and manipulation have progressed autonomous crack repair, existing methods still face three key challenges: accurate localization of cracks within the robot's coordinate frame, (ii) adaptability to varying crack depths and widths, and (iii) validation of the repair process under realistic conditions. This paper presents an adaptive, autonomous system for surface crack detection and repair using robotics with advanced sensing technologies to enhance precision and safety for humans. The system uses an RGB-D camera for crack detection, a laser scanner for precise measurement, and an extruder and pump for material deposition. To address one of the key challenges, the laser scanner is used to enhance the crack coordinates for accurate localization. Furthermore, our approach demonstrates that an adaptive crack-filling method is more efficient and effective than a fixed-speed approach, with experimental results confirming both precision and consistency. In addition, to ensure real-world applicability and testing repeatability, we introduce a novel validation procedure using 3D-printed crack specimens that accurately simulate real-world conditions. This research contributes to the evolving field of human-robot interaction in construction by demonstrating how adaptive robotic systems can reduce the need for manual labor, improve safety, and enhance the efficiency of maintenance operations, ultimately paving the way for more sophisticated and integrated construction robotics.
comment: 22 pages, 14 figures, submitted to Advanced Engineering Informatics
♻ ☆ Sample what you cant compress
For learned image representations, basic autoencoders often produce blurry results. Reconstruction quality can be improved by incorporating additional penalties such as adversarial (GAN) and perceptual losses. Arguably, these approaches lack a principled interpretation. Concurrently, in generative settings diffusion has demonstrated a remarkable ability to create crisp, high quality results and has solid theoretical underpinnings (from variational inference to direct study as the Fisher Divergence). Our work combines autoencoder representation learning with diffusion and is, to our knowledge, the first to demonstrate the efficacy of jointly learning a continuous encoder and decoder under a diffusion-based loss. We demonstrate that this approach yields better reconstruction quality as compared to GAN-based autoencoders while being easier to tune. We also show that the resulting representation is easier to model with a latent diffusion model as compared to the representation obtained from a state-of-the-art GAN-based loss. Since our decoder is stochastic, it can generate details not encoded in the otherwise deterministic latent representation; we therefore name our approach "Sample what you can't compress", or SWYCC for short.
♻ ☆ MuseTalk: Real-Time High Quality Lip Synchronization with Latent Space Inpainting
Achieving high-resolution, identity consistency, and accurate lip-speech synchronization in face visual dubbing presents significant challenges, particularly for real-time applications like live video streaming. We propose MuseTalk, which generates lip-sync targets in a latent space encoded by a Variational Autoencoder, enabling high-fidelity talking face video generation with efficient inference. Specifically, we project the occluded lower half of the face image and itself as an reference into a low-dimensional latent space and use a multi-scale U-Net to fuse audio and visual features at various levels. We further propose a novel sampling strategy during training, which selects reference images with head poses closely matching the target, allowing the model to focus on precise lip movement by filtering out redundant information. Additionally, we analyze the mechanism of lip-sync loss and reveal its relationship with input information volume. Extensive experiments show that MuseTalk consistently outperforms recent state-of-the-art methods in visual fidelity and achieves comparable lip-sync accuracy. As MuseTalk supports the online generation of face at 256x256 at more than 30 FPS with negligible starting latency, it paves the way for real-time applications.
comment: 15 pages, 4 figures
♻ ☆ Delta-ICM: Entropy Modeling with Delta Function for Learned Image Compression
Image Coding for Machines (ICM) is becoming more important as research in computer vision progresses. ICM is a vital research field that pursues the use of images for image recognition models, facilitating efficient image transmission and storage. The demand for recognition models is growing rapidly among the general public, and their performance continues to improve. To meet these needs, exchanging image data between consumer devices and cloud AI using ICM technology could be one possible solution. In ICM, various image compression methods have adopted Learned Image Compression (LIC). LIC includes an entropy model for estimating the bitrate of latent features, and the design of this model significantly affects its performance. Typically, LIC methods assume that the distribution of latent features follows a normal distribution. This assumption is effective for compressing images intended for human vision. However, employing an entropy model based on normal distribution is inefficient in ICM due to the limitation of image parts that require precise decoding. To address this, we propose Delta-ICM, which uses a probability distribution based on a delta function. Assuming the delta distribution as a distribution of latent features reduces the entropy of image portions unnecessary for machines. We compress the remaining portions using an entropy model based on normal distribution, similar to existing methods. Delta-ICM selects between the entropy model based on the delta distribution and the one based on the normal distribution for each latent feature. Our method outperforms existing ICM methods in image compression performance aimed at machines.
♻ ☆ Video-to-Audio Generation with Hidden Alignment
Generating semantically and temporally aligned audio content in accordance with video input has become a focal point for researchers, particularly following the remarkable breakthrough in text-to-video generation. In this work, we aim to offer insights into the video-to-audio generation paradigm, focusing on three crucial aspects: vision encoders, auxiliary embeddings, and data augmentation techniques. Beginning with a foundational model built on a simple yet surprisingly effective intuition, we explore various vision encoders and auxiliary embeddings through ablation studies. Employing a comprehensive evaluation pipeline that emphasizes generation quality and video-audio synchronization alignment, we demonstrate that our model exhibits state-of-the-art video-to-audio generation capabilities. Furthermore, we provide critical insights into the impact of different data augmentation methods on enhancing the generation framework's overall capacity. We showcase possibilities to advance the challenge of generating synchronized audio from semantic and temporal perspectives. We hope these insights will serve as a stepping stone toward developing more realistic and accurate audio-visual generation models.
comment: https://sites.google.com/view/vta-ldm
♻ ☆ Knowledge Circuits in Pretrained Transformers NeurIPS 2024
The remarkable capabilities of modern large language models are rooted in their vast repositories of knowledge encoded within their parameters, enabling them to perceive the world and engage in reasoning. The inner workings of how these models store knowledge have long been a subject of intense interest and investigation among researchers. To date, most studies have concentrated on isolated components within these models, such as the Multilayer Perceptrons and attention head. In this paper, we delve into the computation graph of the language model to uncover the knowledge circuits that are instrumental in articulating specific knowledge. The experiments, conducted with GPT2 and TinyLLAMA, have allowed us to observe how certain information heads, relation heads, and Multilayer Perceptrons collaboratively encode knowledge within the model. Moreover, we evaluate the impact of current knowledge editing techniques on these knowledge circuits, providing deeper insights into the functioning and constraints of these editing methodologies. Finally, we utilize knowledge circuits to analyze and interpret language model behaviors such as hallucinations and in-context learning. We believe the knowledge circuits hold potential for advancing our understanding of Transformers and guiding the improved design of knowledge editing. Code and data are available in https://github.com/zjunlp/KnowledgeCircuits.
comment: NeurIPS 2024, 32 pages
♻ ☆ CVCP-Fusion: On Implicit Depth Estimation for 3D Bounding Box Prediction
Combining LiDAR and Camera-view data has become a common approach for 3D Object Detection. However, previous approaches combine the two input streams at a point-level, throwing away semantic information derived from camera features. In this paper we propose Cross-View Center Point-Fusion, a state-of-the-art model to perform 3D object detection by combining camera and LiDAR-derived features in the BEV space to preserve semantic density from the camera stream while incorporating spacial data from the LiDAR stream. Our architecture utilizes aspects from previously established algorithms, Cross-View Transformers and CenterPoint, and runs their backbones in parallel, allowing efficient computation for real-time processing and application. In this paper we find that while an implicitly calculated depth-estimate may be sufficiently accurate in a 2D map-view representation, explicitly calculated geometric and spacial information is needed for precise bounding box prediction in the 3D world-view space.
comment: 7 pages, 5 figures. arXiv admin note: text overlap with arXiv:2205.02833 by other authors
♻ ☆ ReLayout: Towards Real-World Document Understanding via Layout-enhanced Pre-training
Recent approaches for visually-rich document understanding (VrDU) uses manually annotated semantic groups, where a semantic group encompasses all semantically relevant but not obviously grouped words. As OCR tools are unable to automatically identify such grouping, we argue that current VrDU approaches are unrealistic. We thus introduce a new variant of the VrDU task, real-world visually-rich document understanding (ReVrDU), that does not allow for using manually annotated semantic groups. We also propose a new method, ReLayout, compliant with the ReVrDU scenario, which learns to capture semantic grouping through arranging words and bringing the representations of words that belong to the potential same semantic group closer together. Our experimental results demonstrate the performance of existing methods is deteriorated with the ReVrDU task, while ReLayout shows superiour performance.
♻ ☆ Ultra-High-Definition Image Restoration: New Benchmarks and A Dual Interaction Prior-Driven Solution
Ultra-High-Definition (UHD) image restoration has acquired remarkable attention due to its practical demand. In this paper, we construct UHD snow and rain benchmarks, named UHD-Snow and UHD-Rain, to remedy the deficiency in this field. The UHD-Snow/UHD-Rain is established by simulating the physics process of rain/snow into consideration and each benchmark contains 3200 degraded/clear image pairs of 4K resolution. Furthermore, we propose an effective UHD image restoration solution by considering gradient and normal priors in model design thanks to these priors' spatial and detail contributions. Specifically, our method contains two branches: (a) feature fusion and reconstruction branch in high-resolution space and (b) prior feature interaction branch in low-resolution space. The former learns high-resolution features and fuses prior-guided low-resolution features to reconstruct clear images, while the latter utilizes normal and gradient priors to mine useful spatial features and detail features to guide high-resolution recovery better. To better utilize these priors, we introduce single prior feature interaction and dual prior feature interaction, where the former respectively fuses normal and gradient priors with high-resolution features to enhance prior ones, while the latter calculates the similarity between enhanced prior ones and further exploits dual guided filtering to boost the feature interaction of dual priors. We conduct experiments on both new and existing public datasets and demonstrate the state-of-the-art performance of our method on UHD image low-light enhancement, dehazing, deblurring, desonwing, and deraining. The source codes and benchmarks are available at \url{https://github.com/wlydlut/UHDDIP}.
♻ ☆ AdaMSS: Adaptive Multi-Modality Segmentation-to-Survival Learning for Survival Outcome Prediction from PET/CT Images
Survival prediction is a major concern for cancer management. Deep survival models based on deep learning have been widely adopted to perform end-to-end survival prediction from medical images. Recent deep survival models achieved promising performance by jointly performing tumor segmentation with survival prediction, where the models were guided to extract tumor-related information through Multi-Task Learning (MTL). However, these deep survival models have difficulties in exploring out-of-tumor prognostic information. In addition, existing deep survival models are unable to effectively leverage multi-modality images. Empirically-designed fusion strategies were commonly adopted to fuse multi-modality information via task-specific manually-designed networks, thus limiting the adaptability to different scenarios. In this study, we propose an Adaptive Multi-modality Segmentation-to-Survival model (AdaMSS) for survival prediction from PET/CT images. Instead of adopting MTL, we propose a novel Segmentation-to-Survival Learning (SSL) strategy, where our AdaMSS is trained for tumor segmentation and survival prediction sequentially in two stages. This strategy enables the AdaMSS to focus on tumor regions in the first stage and gradually expand its focus to include other prognosis-related regions in the second stage. We also propose a data-driven strategy to fuse multi-modality information, which realizes adaptive optimization of fusion strategies based on training data during training. With the SSL and data-driven fusion strategies, our AdaMSS is designed as an adaptive model that can self-adapt its focus regions and fusion strategy for different training stages. Extensive experiments with two large clinical datasets show that our AdaMSS outperforms state-of-the-art survival prediction methods.
comment: The extended version of this paper has been published at npj Precision Oncology as "Adaptive segmentation-to-survival learning for survival prediction from multi-modality medical images"
♻ ☆ Zero-shot Generalizable Incremental Learning for Vision-Language Object Detection NeurIPS 2024
This paper presents Incremental Vision-Language Object Detection (IVLOD), a novel learning task designed to incrementally adapt pre-trained Vision-Language Object Detection Models (VLODMs) to various specialized domains, while simultaneously preserving their zero-shot generalization capabilities for the generalized domain. To address this new challenge, we present the Zero-interference Reparameterizable Adaptation (ZiRa), a novel method that introduces Zero-interference Loss and reparameterization techniques to tackle IVLOD without incurring additional inference costs or a significant increase in memory usage. Comprehensive experiments on COCO and ODinW-13 datasets demonstrate that ZiRa effectively safeguards the zero-shot generalization ability of VLODMs while continuously adapting to new tasks. Specifically, after training on ODinW-13 datasets, ZiRa exhibits superior performance compared to CL-DETR and iDETR, boosting zero-shot generalizability by substantial 13.91 and 8.74 AP, respectively.Our code is available at https://github.com/JarintotionDin/ZiRaGroundingDINO.
comment: This paper has been accepted by NeurIPS 2024
♻ ☆ In the Eye of Transformer: Global-Local Correlation for Egocentric Gaze Estimation
In this paper, we present the first transformer-based model to address the challenging problem of egocentric gaze estimation. We observe that the connection between the global scene context and local visual information is vital for localizing the gaze fixation from egocentric video frames. To this end, we design the transformer encoder to embed the global context as one additional visual token and further propose a novel Global-Local Correlation (GLC) module to explicitly model the correlation of the global token and each local token. We validate our model on two egocentric video datasets - EGTEA Gaze+ and Ego4D. Our detailed ablation studies demonstrate the benefits of our method. In addition, our approach exceeds previous state-of-the-arts by a large margin. We also provide additional visualizations to support our claim that global-local correlation serves a key representation for predicting gaze fixation from egocentric videos. More details can be found in our website (https://bolinlai.github.io/GLC-EgoGazeEst).
comment: 23 pages
♻ ☆ Lotus: Diffusion-based Visual Foundation Model for High-quality Dense Prediction
Leveraging the visual priors of pre-trained text-to-image diffusion models offers a promising solution to enhance zero-shot generalization in dense prediction tasks. However, existing methods often uncritically use the original diffusion formulation, which may not be optimal due to the fundamental differences between dense prediction and image generation. In this paper, we provide a systemic analysis of the diffusion formulation for the dense prediction, focusing on both quality and efficiency. And we find that the original parameterization type for image generation, which learns to predict noise, is harmful for dense prediction; the multi-step noising/denoising diffusion process is also unnecessary and challenging to optimize. Based on these insights, we introduce Lotus, a diffusion-based visual foundation model with a simple yet effective adaptation protocol for dense prediction. Specifically, Lotus is trained to directly predict annotations instead of noise, thereby avoiding harmful variance. We also reformulate the diffusion process into a single-step procedure, simplifying optimization and significantly boosting inference speed. Additionally, we introduce a novel tuning strategy called detail preserver, which achieves more accurate and fine-grained predictions. Without scaling up the training data or model capacity, Lotus achieves SoTA performance in zero-shot depth and normal estimation across various datasets. It also enhances efficiency, being significantly faster than most existing diffusion-based methods. Lotus' superior quality and efficiency also enable a wide range of practical applications, such as joint estimation, single/multi-view 3D reconstruction, etc. Project page: https://lotus3d.github.io/.
comment: The first two authors contributed equally. Project page: https://lotus3d.github.io/
♻ ☆ MixedNUTS: Training-Free Accuracy-Robustness Balance via Nonlinearly Mixed Classifiers
Adversarial robustness often comes at the cost of degraded accuracy, impeding real-life applications of robust classification models. Training-based solutions for better trade-offs are limited by incompatibilities with already-trained high-performance large models, necessitating the exploration of training-free ensemble approaches. Observing that robust models are more confident in correct predictions than in incorrect ones on clean and adversarial data alike, we speculate amplifying this "benign confidence property" can reconcile accuracy and robustness in an ensemble setting. To achieve so, we propose "MixedNUTS", a training-free method where the output logits of a robust classifier and a standard non-robust classifier are processed by nonlinear transformations with only three parameters, which are optimized through an efficient algorithm. MixedNUTS then converts the transformed logits into probabilities and mixes them as the overall output. On CIFAR-10, CIFAR-100, and ImageNet datasets, experimental results with custom strong adaptive attacks demonstrate MixedNUTS's vastly improved accuracy and near-SOTA robustness -- it boosts CIFAR-100 clean accuracy by 7.86 points, sacrificing merely 0.87 points in robust accuracy.
♻ ☆ HyperDreamBooth: HyperNetworks for Fast Personalization of Text-to-Image Models
Personalization has emerged as a prominent aspect within the field of generative AI, enabling the synthesis of individuals in diverse contexts and styles, while retaining high-fidelity to their identities. However, the process of personalization presents inherent challenges in terms of time and memory requirements. Fine-tuning each personalized model needs considerable GPU time investment, and storing a personalized model per subject can be demanding in terms of storage capacity. To overcome these challenges, we propose HyperDreamBooth - a hypernetwork capable of efficiently generating a small set of personalized weights from a single image of a person. By composing these weights into the diffusion model, coupled with fast finetuning, HyperDreamBooth can generate a person's face in various contexts and styles, with high subject details while also preserving the model's crucial knowledge of diverse styles and semantic modifications. Our method achieves personalization on faces in roughly 20 seconds, 25x faster than DreamBooth and 125x faster than Textual Inversion, using as few as one reference image, with the same quality and style diversity as DreamBooth. Also our method yields a model that is 10,000x smaller than a normal DreamBooth model. Project page: https://hyperdreambooth.github.io
comment: project page: https://hyperdreambooth.github.io
♻ ☆ InterMask: 3D Human Interaction Generation via Collaborative Masked Modelling
Generating realistic 3D human-human interactions from textual descriptions remains a challenging task. Existing approaches, typically based on diffusion models, often generate unnatural and unrealistic results. In this work, we introduce InterMask, a novel framework for generating human interactions using collaborative masked modeling in discrete space. InterMask first employs a VQ-VAE to transform each motion sequence into a 2D discrete motion token map. Unlike traditional 1D VQ token maps, it better preserves fine-grained spatio-temporal details and promotes spatial awareness within each token. Building on this representation, InterMask utilizes a generative masked modeling framework to collaboratively model the tokens of two interacting individuals. This is achieved by employing a transformer architecture specifically designed to capture complex spatio-temporal interdependencies. During training, it randomly masks the motion tokens of both individuals and learns to predict them. In inference, starting from fully masked sequences, it progressively fills in the tokens for both individuals. With its enhanced motion representation, dedicated architecture, and effective learning strategy, InterMask achieves state-of-the-art results, producing high-fidelity and diverse human interactions. It outperforms previous methods, achieving an FID of $5.154$ (vs $5.535$ for in2IN) on the InterHuman dataset and $0.399$ (vs $5.207$ for InterGen) on the InterX dataset. Additionally, InterMask seamlessly supports reaction generation without the need for model redesign or fine-tuning.
comment: Project webpage: https://gohar-malik.github.io/intermask
♻ ☆ Few-Shot Classification of Interactive Activities of Daily Living (InteractADL)
Understanding Activities of Daily Living (ADLs) is a crucial step for different applications including assistive robots, smart homes, and healthcare. However, to date, few benchmarks and methods have focused on complex ADLs, especially those involving multi-person interactions in home environments. In this paper, we propose a new dataset and benchmark, InteractADL, for understanding complex ADLs that involve interaction between humans (and objects). Furthermore, complex ADLs occurring in home environments comprise a challenging long-tailed distribution due to the rarity of multi-person interactions, and pose fine-grained visual recognition tasks due to the presence of semantically and visually similar classes. To address these issues, we propose a novel method for fine-grained few-shot video classification called Name Tuning that enables greater semantic separability by learning optimal class name vectors. We show that Name Tuning can be combined with existing prompt tuning strategies to learn the entire input text (rather than only learning the prompt or class names) and demonstrate improved performance for few-shot classification on InteractADL and 4 other fine-grained visual classification benchmarks. For transparency and reproducibility, we release our code at https://github.com/zanedurante/vlm_benchmark.
♻ ☆ Standalone 16-bit Training: Missing Study for Hardware-Limited Deep Learning Practitioners
With the increasing complexity of machine learning models, managing computational resources like memory and processing power has become a critical concern. Mixed precision techniques, which leverage different numerical precisions during model training and inference to optimize resource usage, have been widely adopted. However, access to hardware that supports lower precision formats (e.g., FP8 or FP4) remains limited, especially for practitioners with hardware constraints. For many with limited resources, the available options are restricted to using 32-bit, 16-bit, or a combination of the two. While it is commonly believed that 16-bit precision can achieve results comparable to full (32-bit) precision, this study is the first to systematically validate this assumption through both rigorous theoretical analysis and extensive empirical evaluation. Our theoretical formalization of floating-point errors and classification tolerance provides new insights into the conditions under which 16-bit precision can approximate 32-bit results. This study fills a critical gap, proving for the first time that standalone 16-bit precision neural networks match 32-bit and mixed-precision in accuracy while boosting computational speed. Given the widespread availability of 16-bit across GPUs, these findings are especially valuable for machine learning practitioners with limited hardware resources to make informed decisions.
♻ ☆ Towards a Knowledge guided Multimodal Foundation Model for Spatio-Temporal Remote Sensing Applications
In recent years, there has been an increased interest in foundation models for geoscience due to the vast amount of Earth observing satellite imagery. Existing remote sensing foundation models make use of the various sources of spectral imagery to create large models pretrained on the task of masked reconstruction. In this paper, we present a foundation model framework, where the pretraining task captures the causal relationship between multiple modalities. Our framework leverages the knowledge guided principles that the spectral imagery captures the impact of the physical drivers on the environmental system, and that the relationship between them is governed by the characteristics of the system. Specifically, our method, called MultiModal Variable Step Forecasting (MM-VSF), uses forecasting of satellite imagery as a pretraining task and is able to capture the causal relationship between spectral imagery and weather. In our evaluation we show that the forecasting of satellite imagery using weather can be used as an effective pretraining task for foundation models. We further show the effectiveness of the embeddings produced by MM-VSF on the downstream tasks of pixel wise crop mapping and missing image prediction of spectral imagery, when compared with embeddings created by models trained in alternative pretraining settings including the traditional single modality input masked reconstruction.
comment: 15 pages with appendix
♻ ☆ Multilinear Mixture of Experts: Scalable Expert Specialization through Factorization NeurIPS 2024
The Mixture of Experts (MoE) paradigm provides a powerful way to decompose dense layers into smaller, modular computations often more amenable to human interpretation, debugging, and editability. However, a major challenge lies in the computational cost of scaling the number of experts high enough to achieve fine-grained specialization. In this paper, we propose the Multilinear Mixture of Experts ($\mu$MoE) layer to address this, focusing on vision models. $\mu$MoE layers enable scalable expert specialization by performing an implicit computation on prohibitively large weight tensors entirely in factorized form. Consequently, $\mu$MoEs (1) avoid the restrictively high inference-time costs of dense MoEs, yet (2) do not inherit the training issues of the popular sparse MoEs' discrete (non-differentiable) expert routing. We present both qualitative and quantitative evidence that scaling $\mu$MoE layers when fine-tuning foundation models for vision tasks leads to more specialized experts at the class-level, further enabling manual bias correction in CelebA attribute classification. Finally, we show qualitative results demonstrating the expert specialism achieved when pre-training large GPT2 and MLP-Mixer models with parameter-matched $\mu$MoE blocks at every layer, maintaining comparable accuracy. Our code is available at: https://github.com/james-oldfield/muMoE.
comment: Accepted at NeurIPS 2024. Github: https://github.com/james-oldfield/muMoE. Project page: https://james-oldfield.github.io/muMoE
♻ ☆ Evaluating Fairness in Large Vision-Language Models Across Diverse Demographic Attributes and Prompts
Large vision-language models (LVLMs) have recently achieved significant progress, demonstrating strong capabilities in open-world visual understanding. However, it is not yet clear how LVLMs address demographic biases in real life, especially the disparities across attributes such as gender, skin tone, age and race. In this paper, We empirically investigate visual fairness in several mainstream LVLMs by auditing their performance disparities across demographic attributes using public fairness benchmark datasets (e.g., FACET, UTKFace). Our fairness evaluation framework employs direct and single-choice question prompt on visual question-answering/classification tasks. Despite advancements in visual understanding, our zero-shot prompting results show that both open-source and closed-source LVLMs continue to exhibit fairness issues across different prompts and demographic groups. Furthermore, we propose a potential multi-modal Chain-of-thought (CoT) based strategy for bias mitigation, applicable to both open-source and closed-source LVLMs. This approach enhances transparency and offers a scalable solution for addressing fairness, providing a solid foundation for future bias reduction efforts.
comment: Under review
♻ ☆ D$^3$Fields: Dynamic 3D Descriptor Fields for Zero-Shot Generalizable Rearrangement
Scene representation is a crucial design choice in robotic manipulation systems. An ideal representation is expected to be 3D, dynamic, and semantic to meet the demands of diverse manipulation tasks. However, previous works often lack all three properties simultaneously. In this work, we introduce D$^3$Fields -- dynamic 3D descriptor fields. These fields are implicit 3D representations that take in 3D points and output semantic features and instance masks. They can also capture the dynamics of the underlying 3D environments. Specifically, we project arbitrary 3D points in the workspace onto multi-view 2D visual observations and interpolate features derived from visual foundational models. The resulting fused descriptor fields allow for flexible goal specifications using 2D images with varied contexts, styles, and instances. To evaluate the effectiveness of these descriptor fields, we apply our representation to rearrangement tasks in a zero-shot manner. Through extensive evaluation in real worlds and simulations, we demonstrate that D$^3$Fields are effective for zero-shot generalizable rearrangement tasks. We also compare D$^3$Fields with state-of-the-art implicit 3D representations and show significant improvements in effectiveness and efficiency.
comment: Accepted to Conference on Robot Learning (CoRL 2024) as Oral Presentation. The first three authors contributed equally. Project Page: https://robopil.github.io/d3fields/
♻ ☆ Gaussian Splatting to Real World Flight Navigation Transfer with Liquid Networks
Simulators are powerful tools for autonomous robot learning as they offer scalable data generation, flexible design, and optimization of trajectories. However, transferring behavior learned from simulation data into the real world proves to be difficult, usually mitigated with compute-heavy domain randomization methods or further model fine-tuning. We present a method to improve generalization and robustness to distribution shifts in sim-to-real visual quadrotor navigation tasks. To this end, we first build a simulator by integrating Gaussian Splatting with quadrotor flight dynamics, and then, train robust navigation policies using Liquid neural networks. In this way, we obtain a full-stack imitation learning protocol that combines advances in 3D Gaussian splatting radiance field rendering, crafty programming of expert demonstration training data, and the task understanding capabilities of Liquid networks. Through a series of quantitative flight tests, we demonstrate the robust transfer of navigation skills learned in a single simulation scene directly to the real world. We further show the ability to maintain performance beyond the training environment under drastic distribution and physical environment changes. Our learned Liquid policies, trained on single target manoeuvres curated from a photorealistic simulated indoor flight only, generalize to multi-step hikes onboard a real hardware platform outdoors.
♻ ☆ VideoTree: Adaptive Tree-based Video Representation for LLM Reasoning on Long Videos
Long-form video understanding has been a challenging task due to the high redundancy in video data and the abundance of query-irrelevant information. To tackle this challenge, we propose VideoTree, a training-free framework which builds a query-adaptive and hierarchical video representation for LLM reasoning over long-form videos. First, VideoTree extracts query-relevant information from the input video through an iterative process, progressively refining the selection of keyframes based on their relevance to the query. Furthermore, VideoTree leverages the inherent hierarchical structure of long video data, which is often overlooked by existing LLM-based methods. Specifically, we incorporate multigranularity information into a tree-based representation, allowing VideoTree to extract query-relevant details from long videos in a coarse-to-fine manner. This enables the model to effectively handle a wide range of video queries with varying levels of detail. Finally, VideoTree aggregates the hierarchical query-relevant information within the tree structure and feeds it into an LLM reasoning model to answer the query. Our experiments show that our training-free method improves both reasoning accuracy and efficiency compared to existing methods. Specifically, VideoTree outperforms the existing training-free approaches on the popular EgoSchema and NExT-QA benchmarks with less inference time, achieving 61.1% and 75.6% accuracy on the test set without additional video-specific training. Moreover, on the long split of Video-MME benchmark (average 44 minutes), the training-free VideoTree framework achieves better performance than the strong proprietary GPT-4V model and other MLLMs that were extensively trained on video data.
comment: 23 pages, first three authors contributed equally; Project page: https://videotree2024.github.io/
♻ ☆ Self-supervised learning of video representations from a child's perspective
Children learn powerful internal models of the world around them from a few years of egocentric visual experience. Can such internal models be learned from a child's visual experience with highly generic learning algorithms or do they require strong inductive biases? Recent advances in collecting large-scale, longitudinal, developmentally realistic video datasets and generic self-supervised learning (SSL) algorithms are allowing us to begin to tackle this nature vs. nurture question. However, existing work typically focuses on image-based SSL algorithms and visual capabilities that can be learned from static images (e.g. object recognition), thus ignoring temporal aspects of the world. To close this gap, here we train self-supervised video models on longitudinal, egocentric headcam recordings collected from a child over a two year period in their early development (6-31 months). The resulting models are highly effective at facilitating the learning of action concepts from a small number of labeled examples; they have favorable data size scaling properties; and they display emergent video interpolation capabilities. Video models also learn more accurate and more robust object representations than image-based models trained with the exact same data. These results suggest that important temporal aspects of a child's internal model of the world may be learnable from their visual experience using highly generic learning algorithms and without strong inductive biases.
comment: v3 updates results with significantly improved models; v2 was published as a conference paper at CogSci 2024; code & models available from https://github.com/eminorhan/video-models
♻ ☆ Motion Inversion for Video Customization
In this work, we present a novel approach for motion customization in video generation, addressing the widespread gap in the exploration of motion representation within video generative models. Recognizing the unique challenges posed by the spatiotemporal nature of video, our method introduces Motion Embeddings, a set of explicit, temporally coherent embeddings derived from a given video. These embeddings are designed to integrate seamlessly with the temporal transformer modules of video diffusion models, modulating self-attention computations across frames without compromising spatial integrity. Our approach provides a compact and efficient solution to motion representation, utilizing two types of embeddings: a Motion Query-Key Embedding to modulate the temporal attention map and a Motion Value Embedding to modulate the attention values. Additionally, we introduce an inference strategy that excludes spatial dimensions from the Motion Query-Key Embedding and applies a differential operation to the Motion Value Embedding, both designed to debias appearance and ensure the embeddings focus solely on motion. Our contributions include the introduction of a tailored motion embedding for customization tasks and a demonstration of the practical advantages and effectiveness of our method through extensive experiments.
comment: https://wileewang.github.io/MotionInversion/
♻ ☆ MMFusion: Combining Image Forensic Filters for Visual Manipulation Detection and Localization
Recent image manipulation localization and detection techniques typically leverage forensic artifacts and traces that are produced by a noise-sensitive filter, such as SRM or Bayar convolution. In this paper, we showcase that different filters commonly used in such approaches excel at unveiling different types of manipulations and provide complementary forensic traces. Thus, we explore ways of combining the outputs of such filters to leverage the complementary nature of the produced artifacts for performing image manipulation localization and detection (IMLD). We assess two distinct combination methods: one that produces independent features from each forensic filter and then fuses them (this is referred to as late fusion) and one that performs early mixing of different modal outputs and produces combined features (this is referred to as early fusion). We use the latter as a feature encoding mechanism, accompanied by a new decoding mechanism that encompasses feature re-weighting, for formulating the proposed MMFusion architecture. We demonstrate that MMFusion achieves competitive performance for both image manipulation localization and detection, outperforming state-of-the-art models across several image and video datasets. We also investigate further the contribution of each forensic filter within MMFusion for addressing different types of manipulations, building on recent AI explainability measures.
comment: This version (v2): extended journal version, submitted for publication. Initial version (v1), arXiv:2312.01790v1 , presented and published in the 30th Int. Conf. on MultiMedia Modeling (MMM 2024), Amsterdam, NL, Jan.-Feb. 2024. This is the "submitted manuscript" version
Information Retrieval 23
☆ RosePO: Aligning LLM-based Recommenders with Human Values
Recently, there has been a growing interest in leveraging Large Language Models (LLMs) for recommendation systems, which usually adapt a pre-trained LLM to the recommendation scenario through supervised fine-tuning (SFT). However, both the pre-training and SFT stages fail to explicitly model the comparative relationships of a user's preferences on different items. To construct a "helpful and harmless" LLM-based recommender, we propose a general framework -- Recommendation with smoothing personalized Preference Optimization (RosePO), which better aligns with customized human values during the post-training stage. Specifically, in addition to the input and chosen response that naturally align with SFT data, we design a rejected sampling strategy tailored for enhancing helpfulness, along with two strategies aimed at mitigating biases to promote harmlessness. To ensure robustness against uncertain labels present in automatically constructed preference data, we introduce a personalized smoothing factor predicted by a preference oracle into the optimization objective. Evaluation on three real-world datasets demonstrates the effectiveness of our method, showcasing not only improved recommendation performance but also mitigation of semantic hallucination and popularity bias.
☆ Unifying Economic and Language Models for Enhanced Sentiment Analysis of the Oil Market
Crude oil, a critical component of the global economy, has its prices influenced by various factors such as economic trends, political events, and natural disasters. Traditional prediction methods based on historical data have their limits in forecasting, but recent advancements in natural language processing bring new possibilities for event-based analysis. In particular, Language Models (LM) and their advancement, the Generative Pre-trained Transformer (GPT), have shown potential in classifying vast amounts of natural language. However, these LMs often have difficulty with domain-specific terminology, limiting their effectiveness in the crude oil sector. Addressing this gap, we introduce CrudeBERT, a fine-tuned LM specifically for the crude oil market. The results indicate that CrudeBERT's sentiment scores align more closely with the WTI Futures curve and significantly enhance price predictions, underscoring the crucial role of integrating economic principles into LMs.
☆ Mitigating Dual Latent Confounding Biases in Recommender Systems
Recommender systems are extensively utilised across various areas to predict user preferences for personalised experiences and enhanced user engagement and satisfaction. Traditional recommender systems, however, are complicated by confounding bias, particularly in the presence of latent confounders that affect both item exposure and user feedback. Existing debiasing methods often fail to capture the complex interactions caused by latent confounders in interaction data, especially when dual latent confounders affect both the user and item sides. To address this, we propose a novel debiasing method that jointly integrates the Instrumental Variables (IV) approach and identifiable Variational Auto-Encoder (iVAE) for Debiased representation learning in Recommendation systems, referred to as IViDR. Specifically, IViDR leverages the embeddings of user features as IVs to address confounding bias caused by latent confounders between items and user feedback, and reconstructs the embedding of items to obtain debiased interaction data. Moreover, IViDR employs an Identifiable Variational Auto-Encoder (iVAE) to infer identifiable representations of latent confounders between item exposure and user feedback from both the original and debiased interaction data. Additionally, we provide theoretical analyses of the soundness of using IV and the identifiability of the latent representations. Extensive experiments on both synthetic and real-world datasets demonstrate that IViDR outperforms state-of-the-art models in reducing bias and providing reliable recommendations.
☆ QUIDS: Query Intent Generation via Dual Space Modeling
Query understanding is a crucial component of Information Retrieval (IR), aimed at identifying the underlying search intent of textual queries. However, most existing approaches oversimplify this task into query classification or clustering, which fails to fully capture the nuanced intent behind the query. In this paper, we address the task of query intent generation: to automatically generate detailed and precise intent descriptions for search queries using relevant and irrelevant documents given a query. These intent descriptions can help users understand why the search engine considered the top-ranked documents relevant, and provide more transparency to the retrieval process. We propose a dual-space model that uses semantic relevance and irrelevance information in the returned documents to explain the understanding of the query intent. Specifically, in the encoding process, we project, separate, and distinguish relevant and irrelevant documents in the representation space. Then, we introduce a semantic decoupling model in the novel disentangling space, where the semantics of irrelevant information are removed from the relevant space, ensuring that only the essential and relevant intent is captured. This process refines the understanding of the query and provides more accurate explanations for the search results. Experiments on benchmark data demonstrate that our methods produce high-quality query intent descriptions, outperforming existing methods for this task, as well as state-of-the-art query-based summarization methods. A token-level visualization of attention scores reveals that our model effectively reduces the focus on irrelevant intent topics. Our findings open up promising research and application directions for query intent generation, particularly in exploratory search.
☆ Multi-Cause Deconfounding for Recommender Systems with Latent Confounders
In recommender systems, various latent confounding factors (e.g., user social environment and item public attractiveness) can affect user behavior, item exposure, and feedback in distinct ways. These factors may directly or indirectly impact user feedback and are often shared across items or users, making them multi-cause latent confounders. However, existing methods typically fail to account for latent confounders between users and their feedback, as well as those between items and user feedback simultaneously. To address the problem of multi-cause latent confounders, we propose a multi-cause deconfounding method for recommender systems with latent confounders (MCDCF). MCDCF leverages multi-cause causal effect estimation to learn substitutes for latent confounders associated with both users and items, using user behaviour data. Specifically, MCDCF treats the multiple items that users interact with and the multiple users that interact with items as treatment variables, enabling it to learn substitutes for the latent confounders that influence the estimation of causality between users and their feedback, as well as between items and user feedback. Additionally, we theoretically demonstrate the soundness of our MCDCF method. Extensive experiments on three real-world datasets demonstrate that our MCDCF method effectively recovers latent confounders related to users and items, reducing bias and thereby improving recommendation accuracy.
☆ Comprehending Knowledge Graphs with Large Language Models for Recommender Systems
Recently, the introduction of knowledge graphs (KGs) has significantly advanced recommender systems by facilitating the discovery of potential associations between items. However, existing methods still face several limitations. First, most KGs suffer from missing facts or limited scopes. This can lead to biased knowledge representations, thereby constraining the model's performance. Second, existing methods typically convert textual information into IDs, resulting in the loss of natural semantic connections between different items. Third, existing methods struggle to capture high-order relationships in global KGs due to their inefficient layer-by-layer information propagation mechanisms, which are prone to introducing significant noise. To address these limitations, we propose a novel method called CoLaKG, which leverages large language models (LLMs) for knowledge-aware recommendation. The extensive world knowledge and remarkable reasoning capabilities of LLMs enable them to supplement KGs. Additionally, the strong text comprehension abilities of LLMs allow for a better understanding of semantic information. Based on this, we first extract subgraphs centered on each item from the KG and convert them into textual inputs for the LLM. The LLM then outputs its comprehension of these item-centered subgraphs, which are subsequently transformed into semantic embeddings. Furthermore, to utilize the global information of the KG, we construct an item-item graph using these semantic embeddings, which can directly capture higher-order associations between items. Both the semantic embeddings and the structural information from the item-item graph are effectively integrated into the recommendation model through our designed representation alignment and neighbor augmentation modules. Extensive experiments on four real-world datasets demonstrate the superiority of our method.
☆ Triple Modality Fusion: Aligning Visual, Textual, and Graph Data with Large Language Models for Multi-Behavior Recommendations
Integrating diverse data modalities is crucial for enhancing the performance of personalized recommendation systems. Traditional models, which often rely on singular data sources, lack the depth needed to accurately capture the multifaceted nature of item features and user behaviors. This paper introduces a novel framework for multi-behavior recommendations, leveraging the fusion of triple-modality, which is visual, textual, and graph data through alignment with large language models (LLMs). By incorporating visual information, we capture contextual and aesthetic item characteristics; textual data provides insights into user interests and item features in detail; and graph data elucidates relationships within the item-behavior heterogeneous graphs. Our proposed model called Triple Modality Fusion (TMF) utilizes the power of LLMs to align and integrate these three modalities, achieving a comprehensive representation of user behaviors. The LLM models the user's interactions including behaviors and item features in natural languages. Initially, the LLM is warmed up using only natural language-based prompts. We then devise the modality fusion module based on cross-attention and self-attention mechanisms to integrate different modalities from other models into the same embedding space and incorporate them into an LLM. Extensive experiments demonstrate the effectiveness of our approach in improving recommendation accuracy. Further ablation studies validate the effectiveness of our model design and benefits of the TMF.
☆ Is Semantic Chunking Worth the Computational Cost?
Recent advances in Retrieval-Augmented Generation (RAG) systems have popularized semantic chunking, which aims to improve retrieval performance by dividing documents into semantically coherent segments. Despite its growing adoption, the actual benefits over simpler fixed-size chunking, where documents are split into consecutive, fixed-size segments, remain unclear. This study systematically evaluates the effectiveness of semantic chunking using three common retrieval-related tasks: document retrieval, evidence retrieval, and retrieval-based answer generation. The results show that the computational costs associated with semantic chunking are not justified by consistent performance gains. These findings challenge the previous assumptions about semantic chunking and highlight the need for more efficient chunking strategies in RAG systems.
☆ Supply Chain Network Extraction and Entity Classification Leveraging Large Language Models
Supply chain networks are critical to the operational efficiency of industries, yet their increasing complexity presents significant challenges in mapping relationships and identifying the roles of various entities. Traditional methods for constructing supply chain networks rely heavily on structured datasets and manual data collection, limiting their scope and efficiency. In contrast, recent advancements in Natural Language Processing (NLP) and large language models (LLMs) offer new opportunities for discovering and analyzing supply chain networks using unstructured text data. This paper proposes a novel approach that leverages LLMs to extract and process raw textual information from publicly available sources to construct a comprehensive supply chain graph. We focus on the civil engineering sector as a case study, demonstrating how LLMs can uncover hidden relationships among companies, projects, and other entities. Additionally, we fine-tune an LLM to classify entities within the supply chain graph, providing detailed insights into their roles and relationships. The results show that domain-specific fine-tuning improves classification accuracy, highlighting the potential of LLMs for industry-specific supply chain analysis. Our contributions include the development of a supply chain graph for the civil engineering sector, as well as a fine-tuned LLM model that enhances entity classification and understanding of supply chain networks.
comment: 11 pages, 4 figures
☆ LLM Confidence Evaluation Measures in Zero-Shot CSS Classification
Assessing classification confidence is critical for leveraging large language models (LLMs) in automated labeling tasks, especially in the sensitive domains presented by Computational Social Science (CSS) tasks. In this paper, we make three key contributions: (1) we propose an uncertainty quantification (UQ) performance measure tailored for data annotation tasks, (2) we compare, for the first time, five different UQ strategies across three distinct LLMs and CSS data annotation tasks, (3) we introduce a novel UQ aggregation strategy that effectively identifies low-confidence LLM annotations and disproportionately uncovers data incorrectly labeled by the LLMs. Our results demonstrate that our proposed UQ aggregation strategy improves upon existing methods andcan be used to significantly improve human-in-the-loop data annotation processes.
☆ LFOSum: Summarizing Long-form Opinions with Large Language Models
Online reviews play a pivotal role in influencing consumer decisions across various domains, from purchasing products to selecting hotels or restaurants. However, the sheer volume of reviews -- often containing repetitive or irrelevant content -- leads to information overload, making it challenging for users to extract meaningful insights. Traditional opinion summarization models face challenges in handling long inputs and large volumes of reviews, while newer Large Language Model (LLM) approaches often fail to generate accurate and faithful summaries. To address those challenges, this paper introduces (1) a new dataset of long-form user reviews, each entity comprising over a thousand reviews, (2) two training-free LLM-based summarization approaches that scale to long inputs, and (3) automatic evaluation metrics. Our dataset of user reviews is paired with in-depth and unbiased critical summaries by domain experts, serving as a reference for evaluation. Additionally, our novel reference-free evaluation metrics provide a more granular, context-sensitive assessment of summary faithfulness. We benchmark several open-source and closed-source LLMs using our methods. Our evaluation reveals that LLMs still face challenges in balancing sentiment and format adherence in long-form summaries, though open-source models can narrow the gap when relevant information is retrieved in a focused manner.
☆ Towards Computational Analysis of Pansori Singing
Pansori is one of the most representative vocal genres of Korean traditional music, which has an elaborated vocal melody line with strong vibrato. Although the music is transmitted orally without any music notation, transcribing pansori music in Western staff notation has been introduced for several purposes, such as documentation of music, education, or research. In this paper, we introduce computational analysis of pansori based on both audio and corresponding transcription, how modern Music Information Retrieval tasks can be used in analyzing traditional music and how it revealed different audio characteristics of what pansori contains.
comment: Late-Breaking Demo Session of the 25th International Society for Music Information Retrieval (ISMIR) Conference, 2024
☆ REFINE on Scarce Data: Retrieval Enhancement through Fine-Tuning via Model Fusion of Embedding Models
Retrieval augmented generation (RAG) pipelines are commonly used in tasks such as question-answering (QA), relying on retrieving relevant documents from a vector store computed using a pretrained embedding model. However, if the retrieved context is inaccurate, the answers generated using the large language model (LLM) may contain errors or hallucinations. Although pretrained embedding models have advanced, adapting them to new domains remains challenging. Fine-tuning is a potential solution, but industry settings often lack the necessary fine-tuning data. To address these challenges, we propose REFINE, a novel technique that generates synthetic data from available documents and then uses a model fusion approach to fine-tune embeddings for improved retrieval performance in new domains, while preserving out-of-domain capability. We conducted experiments on the two public datasets: SQUAD and RAG-12000 and a proprietary TOURISM dataset. Results demonstrate that even the standard fine-tuning with the proposed data augmentation technique outperforms the vanilla pretrained model. Furthermore, when combined with model fusion, the proposed approach achieves superior performance, with a 5.76% improvement in recall on the TOURISM dataset, and 6.58 % and 0.32% enhancement on SQUAD and RAG-12000 respectively.
comment: Accepted in AJCAI'24
☆ AT-RAG: An Adaptive RAG Model Enhancing Query Efficiency with Topic Filtering and Iterative Reasoning
Recent advancements in QA with LLM, like GPT-4, have shown limitations in handling complex multi-hop queries. We propose AT-RAG, a novel multistep RAG incorporating topic modeling for efficient document retrieval and reasoning. Using BERTopic, our model dynamically assigns topics to queries, improving retrieval accuracy and efficiency. We evaluated AT-RAG on multihop benchmark datasets QA and a medical case study QA. Results show significant improvements in correctness, completeness, and relevance compared to existing methods. AT-RAG reduces retrieval time while maintaining high precision, making it suitable for general tasks QA and complex domain-specific challenges such as medical QA. The integration of topic filtering and iterative reasoning enables our model to handle intricate queries efficiently, which makes it suitable for applications that require nuanced information retrieval and decision-making.
☆ P4GCN: Vertical Federated Social Recommendation with Privacy-Preserving Two-Party Graph Convolution Networks
In recent years, graph neural networks (GNNs) have been commonly utilized for social recommendation systems. However, real-world scenarios often present challenges related to user privacy and business constraints, inhibiting direct access to valuable social information from other platforms. While many existing methods have tackled matrix factorization-based social recommendations without direct social data access, developing GNN-based federated social recommendation models under similar conditions remains largely unexplored. To address this issue, we propose a novel vertical federated social recommendation method leveraging privacy-preserving two-party graph convolution networks (P4GCN) to enhance recommendation accuracy without requiring direct access to sensitive social information. First, we introduce a Sandwich-Encryption module to ensure comprehensive data privacy during the collaborative computing process. Second, we provide a thorough theoretical analysis of the privacy guarantees, considering the participation of both curious and honest parties. Extensive experiments on four real-world datasets demonstrate that P4GCN outperforms state-of-the-art methods in terms of recommendation accuracy. The code is available at https://github.com/WwZzz/P4GCN.
♻ ☆ PromptDSI: Prompt-based Rehearsal-free Instance-wise Incremental Learning for Document Retrieval
Differentiable Search Index (DSI) utilizes Pre-trained Language Models (PLMs) for efficient document retrieval without relying on external indexes. However, DSI needs full re-training to handle updates in dynamic corpora, causing significant computational inefficiencies. We introduce PromptDSI, a prompt-based rehearsal-free approach for instance-wise incremental learning document retrieval. PromptDSI attaches prompts to the frozen PLM's encoder of DSI, leveraging its powerful representation to efficiently index new corpora while maintaining a balance between stability and plasticity. We eliminate the initial forward pass of prompt-based continual learning methods that doubles training and inference time. Moreover, we propose a topic-aware prompt pool that employs neural topic embeddings as fixed keys. This strategy ensures diverse and effective prompt usage, addressing the challenge of parameter underutilization caused by the collapse of the query-key matching mechanism. Our empirical evaluations demonstrate that BERT-based PromptDSI matches IncDSI in managing forgetting while improving new corpora performance by more than 4% Hits@10 on NQ320k and upto 3% MRR@10 on MS MARCO 300k.
comment: 20 pages
♻ ☆ DIRAS: Efficient LLM Annotation of Document Relevance in Retrieval Augmented Generation
Retrieval Augmented Generation (RAG) is widely employed to ground responses to queries on domain-specific documents. But do RAG implementations leave out important information when answering queries that need an integrated analysis of information (e.g., Tell me good news in the stock market today.)? To address these concerns, RAG developers need to annotate information retrieval (IR) data for their domain of interest, which is challenging because (1) domain-specific queries usually need nuanced definitions of relevance beyond shallow semantic relevance; and (2) human or GPT-4 annotation is costly and cannot cover all (query, document) pairs (i.e., annotation selection bias), thus harming the effectiveness in evaluating IR recall. To address these challenges, we propose DIRAS (Domain-specific Information Retrieval Annotation with Scalability), a manual-annotation-free schema that fine-tunes open-sourced LLMs to consider nuanced relevance definition and annotate (partial) relevance labels with calibrated relevance scores. Extensive evaluation shows that DIRAS enables smaller (8B) LLMs to achieve GPT-4-level performance on annotating and ranking unseen (query, document) pairs, and is helpful for real-world RAG development. All code, LLM generations, and human annotations can be found in \url{https://github.com/EdisonNi-hku/DIRAS}.
♻ ☆ FLEX: Expert-level False-Less EXecution Metric for Reliable Text-to-SQL Benchmark
Text-to-SQL systems have become crucial for translating natural language into SQL queries in various industries, enabling non-technical users to perform complex data operations. The need for accurate evaluation methods has increased as these systems have grown more sophisticated. However, the Execution Accuracy (EX), the most prevalent evaluation metric, still shows many false positives and negatives. Thus, this paper introduces FLEX (False-Less EXecution), a novel approach to evaluating text-to-SQL systems using large language models (LLMs) to emulate human expert-level evaluation of SQL queries. Our metric improves agreement with human experts (from 62 to 87.04 in Cohen's kappa) with comprehensive context and sophisticated criteria. Our extensive experiments yield several key insights: (1) Models' performance increases by over 2.6 points on average, substantially affecting rankings on Spider and BIRD benchmarks; (2) The underestimation of models in EX primarily stems from annotation quality issues; and (3) Model performance on particularly challenging questions tends to be overestimated. This work contributes to a more accurate and nuanced evaluation of text-to-SQL systems, potentially reshaping our understanding of state-of-the-art performance in this field.
comment: preprint, under review
♻ ☆ A Practice-Friendly LLM-Enhanced Paradigm with Preference Parsing for Sequential Recommendation
The training paradigm integrating large language models (LLM) is gradually reshaping sequential recommender systems (SRS) and has shown promising results. However, most existing LLM-enhanced methods rely on rich textual information on the item side and instance-level supervised fine-tuning (SFT) to inject collaborative information into LLM, which is inefficient and limited in many applications. To alleviate these problems, this paper proposes a practice-friendly LLM-enhanced paradigm with preference parsing (P2Rec) for SRS. Specifically, in the information reconstruction stage, we design a new user-level SFT task for collaborative information injection with the assistance of a pre-trained SRS model, which is more efficient and compatible with limited text information. Our goal is to let LLM learn to reconstruct a corresponding prior preference distribution from each user's interaction sequence, where LLM needs to effectively parse the latent category of each item and the relationship between different items to accomplish this task. In the information augmentation stage, we feed each item into LLM to obtain a set of enhanced embeddings that combine collaborative information and LLM inference capabilities. These embeddings can then be used to help train various future SRS models. Finally, we verify the effectiveness and efficiency of our TSLRec on three SRS benchmark datasets.
♻ ☆ Knowledge Circuits in Pretrained Transformers NeurIPS 2024
The remarkable capabilities of modern large language models are rooted in their vast repositories of knowledge encoded within their parameters, enabling them to perceive the world and engage in reasoning. The inner workings of how these models store knowledge have long been a subject of intense interest and investigation among researchers. To date, most studies have concentrated on isolated components within these models, such as the Multilayer Perceptrons and attention head. In this paper, we delve into the computation graph of the language model to uncover the knowledge circuits that are instrumental in articulating specific knowledge. The experiments, conducted with GPT2 and TinyLLAMA, have allowed us to observe how certain information heads, relation heads, and Multilayer Perceptrons collaboratively encode knowledge within the model. Moreover, we evaluate the impact of current knowledge editing techniques on these knowledge circuits, providing deeper insights into the functioning and constraints of these editing methodologies. Finally, we utilize knowledge circuits to analyze and interpret language model behaviors such as hallucinations and in-context learning. We believe the knowledge circuits hold potential for advancing our understanding of Transformers and guiding the improved design of knowledge editing. Code and data are available in https://github.com/zjunlp/KnowledgeCircuits.
comment: NeurIPS 2024, 32 pages
♻ ☆ $\textit{lucie}$: An Improved Python Package for Loading Datasets from the UCI Machine Learning Repository
The University of California--Irvine (UCI) Machine Learning (ML) Repository (UCIMLR) is consistently cited as one of the most popular dataset repositories, hosting hundreds of high-impact datasets. However, a significant portion, including 28.4% of the top 250, cannot be imported via the $\textit{ucimlrepo}$ package that is provided and recommended by the UCIMLR website. Instead, they are hosted as .zip files, containing nonstandard formats that are difficult to import without additional ad hoc processing. To address this issue, here we present $\textit{lucie}$ -- $\underline{l}oad$ $\underline{U}niversity$ $\underline{C}alifornia$ $\underline{I}rvine$ $\underline{e}xamples$ -- a utility that automatically determines the data format and imports many of these previously non-importable datasets, while preserving as much of a tabular data structure as possible. $\textit{lucie}$ was designed using the top 100 most popular datasets and benchmarked on the next 130, where it resulted in a success rate of 95.4% vs. 73.1% for $\textit{ucimlrepo}$. $\textit{lucie}$ is available as a Python package on PyPI with 98% code coverage.
comment: 5 pages, 3 figures
♻ ☆ Beyond Inter-Item Relations: Dynamic Adaption for Enhancing LLM-Based Sequential Recommendation
Sequential recommender systems (SRS) predict the next items that users may prefer based on user historical interaction sequences. Inspired by the rise of large language models (LLMs) in various AI applications, there is a surge of work on LLM-based SRS. Despite their attractive performance, existing LLM-based SRS still exhibit some limitations, including neglecting intra-item relations, ignoring long-term collaborative knowledge and using inflexible architecture designs for adaption. To alleviate these issues, we propose an LLM-based sequential recommendation model named DARec. Built on top of coarse-grained adaption for capturing inter-item relations, DARec is further enhanced with (1) context masking that models intra-item relations to help LLM better understand token and item semantics in the context of SRS, (2) collaborative knowledge injection that helps LLM incorporate long-term collaborative knowledge, and (3) a dynamic adaption mechanism that uses Bayesian optimization to flexibly choose layer-wise adapter architectures in order to better incorporate different sequential information. Extensive experiments demonstrate that DARec can effectively handle sequential recommendation in a dynamic and adaptive manner.
comment: 11 pages, 14 figures
♻ ☆ LitSearch: A Retrieval Benchmark for Scientific Literature Search EMNLP 2024
Literature search questions, such as "Where can I find research on the evaluation of consistency in generated summaries?" pose significant challenges for modern search engines and retrieval systems. These questions often require a deep understanding of research concepts and the ability to reason across entire articles. In this work, we introduce LitSearch, a retrieval benchmark comprising 597 realistic literature search queries about recent ML and NLP papers. LitSearch is constructed using a combination of (1) questions generated by GPT-4 based on paragraphs containing inline citations from research papers and (2) questions manually written by authors about their recently published papers. All LitSearch questions were manually examined or edited by experts to ensure high quality. We extensively benchmark state-of-the-art retrieval models and also evaluate two LLM-based reranking pipelines. We find a significant performance gap between BM25 and state-of-the-art dense retrievers, with a 24.8% absolute difference in recall@5. The LLM-based reranking strategies further improve the best-performing dense retriever by 4.4%. Additionally, commercial search engines and research tools like Google Search perform poorly on LitSearch, lagging behind the best dense retriever by up to 32 recall points. Taken together, these results show that LitSearch is an informative new testbed for retrieval systems while catering to a real-world use case.
comment: Accepted by EMNLP 2024. Dataset and code are available at https://github.com/princeton-nlp/LitSearch
Machine Learning 200
☆ Dual Prototype Evolving for Test-Time Generalization of Vision-Language Models NeurIPS 2024
Test-time adaptation, which enables models to generalize to diverse data with unlabeled test samples, holds significant value in real-world scenarios. Recently, researchers have applied this setting to advanced pre-trained vision-language models (VLMs), developing approaches such as test-time prompt tuning to further extend their practical applicability. However, these methods typically focus solely on adapting VLMs from a single modality and fail to accumulate task-specific knowledge as more samples are processed. To address this, we introduce Dual Prototype Evolving (DPE), a novel test-time adaptation approach for VLMs that effectively accumulates task-specific knowledge from multi-modalities. Specifically, we create and evolve two sets of prototypes--textual and visual--to progressively capture more accurate multi-modal representations for target classes during test time. Moreover, to promote consistent multi-modal representations, we introduce and optimize learnable residuals for each test sample to align the prototypes from both modalities. Extensive experimental results on 15 benchmark datasets demonstrate that our proposed DPE consistently outperforms previous state-of-the-art methods while also exhibiting competitive computational efficiency. Code is available at https://github.com/zhangce01/DPE-CLIP.
comment: Accepted by NeurIPS 2024. Project page: https://zhangce01.github.io/DPE-CLIP
☆ Metal Price Spike Prediction via a Neurosymbolic Ensemble Approach
Predicting price spikes in critical metals such as Cobalt, Copper, Magnesium, and Nickel is crucial for mitigating economic risks associated with global trends like the energy transition and reshoring of manufacturing. While traditional models have focused on regression-based approaches, our work introduces a neurosymbolic ensemble framework that integrates multiple neural models with symbolic error detection and correction rules. This framework is designed to enhance predictive accuracy by correcting individual model errors and offering interpretability through rule-based explanations. We show that our method provides up to 6.42% improvement in precision, 29.41% increase in recall at 13.24% increase in F1 over the best performing neural models. Further, our method, as it is based on logical rules, has the benefit of affording an explanation as to which combination of neural models directly contribute to a given prediction.
☆ JudgeBench: A Benchmark for Evaluating LLM-based Judges
LLM-based judges have emerged as a scalable alternative to human evaluation and are increasingly used to assess, compare, and improve models. However, the reliability of LLM-based judges themselves is rarely scrutinized. As LLMs become more advanced, their responses grow more sophisticated, requiring stronger judges to evaluate them. Existing benchmarks primarily focus on a judge's alignment with human preferences, but often fail to account for more challenging tasks where crowdsourced human preference is a poor indicator of factual and logical correctness. To address this, we propose a novel evaluation framework to objectively evaluate LLM-based judges. Based on this framework, we propose JudgeBench, a benchmark for evaluating LLM-based judges on challenging response pairs spanning knowledge, reasoning, math, and coding. JudgeBench leverages a novel pipeline for converting existing difficult datasets into challenging response pairs with preference labels reflecting objective correctness. Our comprehensive evaluation on a collection of prompted judges, fine-tuned judges, multi-agent judges, and reward models shows that JudgeBench poses a significantly greater challenge than previous benchmarks, with many strong models (e.g., GPT-4o) performing just slightly better than random guessing. Overall, JudgeBench offers a reliable platform for assessing increasingly advanced LLM-based judges. Data and code are available at https://github.com/ScalerLab/JudgeBench .
comment: preprint
☆ Context-Scaling versus Task-Scaling in In-Context Learning
Transformers exhibit In-Context Learning (ICL), where these models solve new tasks by using examples in the prompt without additional training. In our work, we identify and analyze two key components of ICL: (1) context-scaling, where model performance improves as the number of in-context examples increases and (2) task-scaling, where model performance improves as the number of pre-training tasks increases. While transformers are capable of both context-scaling and task-scaling, we empirically show that standard Multi-Layer Perceptrons (MLPs) with vectorized input are only capable of task-scaling. To understand how transformers are capable of context-scaling, we first propose a significantly simplified transformer architecture without key, query, value weights. We show that it performs ICL comparably to the original GPT-2 model in various statistical learning tasks including linear regression, teacher-student settings. Furthermore, a single block of our simplified transformer can be viewed as data dependent feature map followed by an MLP. This feature map on its own is a powerful predictor that is capable of context-scaling but is not capable of task-scaling. We show empirically that concatenating the output of this feature map with vectorized data as an input to MLPs enables both context-scaling and task-scaling. This finding provides a simple setting to study context and task-scaling for ICL.
☆ Geometry-Aware Generative Autoencoders for Warped Riemannian Metric Learning and Generative Modeling on Data Manifolds
Rapid growth of high-dimensional datasets in fields such as single-cell RNA sequencing and spatial genomics has led to unprecedented opportunities for scientific discovery, but it also presents unique computational and statistical challenges. Traditional methods struggle with geometry-aware data generation, interpolation along meaningful trajectories, and transporting populations via feasible paths. To address these issues, we introduce Geometry-Aware Generative Autoencoder (GAGA), a novel framework that combines extensible manifold learning with generative modeling. GAGA constructs a neural network embedding space that respects the intrinsic geometries discovered by manifold learning and learns a novel warped Riemannian metric on the data space. This warped metric is derived from both the points on the data manifold and negative samples off the manifold, allowing it to characterize a meaningful geometry across the entire latent space. Using this metric, GAGA can uniformly sample points on the manifold, generate points along geodesics, and interpolate between populations across the learned manifold. GAGA shows competitive performance in simulated and real world datasets, including a 30% improvement over the state-of-the-art methods in single-cell population-level trajectory inference.
☆ Meta-Unlearning on Diffusion Models: Preventing Relearning Unlearned Concepts
With the rapid progress of diffusion-based content generation, significant efforts are being made to unlearn harmful or copyrighted concepts from pretrained diffusion models (DMs) to prevent potential model misuse. However, it is observed that even when DMs are properly unlearned before release, malicious finetuning can compromise this process, causing DMs to relearn the unlearned concepts. This occurs partly because certain benign concepts (e.g., "skin") retained in DMs are related to the unlearned ones (e.g., "nudity"), facilitating their relearning via finetuning. To address this, we propose meta-unlearning on DMs. Intuitively, a meta-unlearned DM should behave like an unlearned DM when used as is; moreover, if the meta-unlearned DM undergoes malicious finetuning on unlearned concepts, the related benign concepts retained within it will be triggered to self-destruct, hindering the relearning of unlearned concepts. Our meta-unlearning framework is compatible with most existing unlearning methods, requiring only the addition of an easy-to-implement meta objective. We validate our approach through empirical experiments on meta-unlearning concepts from Stable Diffusion models (SD-v1-4 and SDXL), supported by extensive ablation studies. Our code is available at https://github.com/sail-sg/Meta-Unlearning.
☆ The Non-Local Model Merging Problem: Permutation Symmetries and Variance Collapse
Model merging aims to efficiently combine the weights of multiple expert models, each trained on a specific task, into a single multi-task model, with strong performance across all tasks. When applied to all but the last layer of weights, existing methods -- such as Task Arithmetic, TIES-merging, and TALL mask merging -- work well to combine expert models obtained by fine-tuning a common foundation model, operating within a "local" neighborhood of the foundation model. This work explores the more challenging scenario of "non-local" merging, which we find arises when an expert model changes significantly during pretraining or where the expert models do not even share a common foundation model. We observe that standard merging techniques often fail to generalize effectively in this non-local setting, even when accounting for permutation symmetries using standard techniques. We identify that this failure is, in part, due to "variance collapse", a phenomenon identified also in the setting of linear mode connectivity by Jordan et al. (2023). To address this, we propose a multi-task technique to re-scale and shift the output activations of the merged model for each task, aligning its output statistics with those of the corresponding task-specific expert models. Our experiments demonstrate that this correction significantly improves the performance of various model merging approaches in non-local settings, providing a strong baseline for future research on this problem.
☆ SAFREE: Training-Free and Adaptive Guard for Safe Text-to-Image And Video Generation
Recent advances in diffusion models have significantly enhanced their ability to generate high-quality images and videos, but they have also increased the risk of producing unsafe content. Existing unlearning/editing-based methods for safe generation remove harmful concepts from models but face several challenges: (1) They cannot instantly remove harmful concepts without training. (2) Their safe generation capabilities depend on collected training data. (3) They alter model weights, risking degradation in quality for content unrelated to toxic concepts. To address these, we propose SAFREE, a novel, training-free approach for safe T2I and T2V, that does not alter the model's weights. Specifically, we detect a subspace corresponding to a set of toxic concepts in the text embedding space and steer prompt embeddings away from this subspace, thereby filtering out harmful content while preserving intended semantics. To balance the trade-off between filtering toxicity and preserving safe concepts, SAFREE incorporates a novel self-validating filtering mechanism that dynamically adjusts the denoising steps when applying the filtered embeddings. Additionally, we incorporate adaptive re-attention mechanisms within the diffusion latent space to selectively diminish the influence of features related to toxic concepts at the pixel level. In the end, SAFREE ensures coherent safety checking, preserving the fidelity, quality, and safety of the output. SAFREE achieves SOTA performance in suppressing unsafe content in T2I generation compared to training-free baselines and effectively filters targeted concepts while maintaining high-quality images. It also shows competitive results against training-based methods. We extend SAFREE to various T2I backbones and T2V tasks, showcasing its flexibility and generalization. SAFREE provides a robust and adaptable safeguard for ensuring safe visual generation.
comment: The first two authors contributed equally; Project page: https://safree-safe-t2i-t2v.github.io/
☆ StyleDistance: Stronger Content-Independent Style Embeddings with Synthetic Parallel Examples
Style representations aim to embed texts with similar writing styles closely and texts with different styles far apart, regardless of content. However, the contrastive triplets often used for training these representations may vary in both style and content, leading to potential content leakage in the representations. We introduce StyleDistance, a novel approach to training stronger content-independent style embeddings. We use a large language model to create a synthetic dataset of near-exact paraphrases with controlled style variations, and produce positive and negative examples across 40 distinct style features for precise contrastive learning. We assess the quality of our synthetic data and embeddings through human and automatic evaluations. StyleDistance enhances the content-independence of style embeddings, which generalize to real-world benchmarks and outperform leading style representations in downstream applications. Our model can be found at https://huggingface.co/StyleDistance/styledistance .
☆ Initialization Method for Factorization Machine Based on Low-Rank Approximation for Constructing a Corrected Approximate Ising Model
This paper presents an initialization method that can approximate a given approximate Ising model with a high degree of accuracy using the Factorization Machine (FM), a machine learning model. The construction of Ising models using FM is applied to the combinatorial optimization problem using the factorization machine with quantum annealing. It is anticipated that the optimization performance of FMQA will be enhanced through the implementation of the warm-start method. Nevertheless, the optimal initialization method for leveraging the warm-start approach in FMQA remains undetermined. Consequently, the present study compares a number of initialization methods and identifies the most appropriate for use with a warm-start in FMQA through numerical experimentation. Furthermore, the properties of the proposed FM initialization method are analyzed using random matrix theory, demonstrating that the approximation accuracy of the proposed method is not significantly influenced by the specific Ising model under consideration. The findings of this study will facilitate the advancement of combinatorial optimization problem-solving through the use of Ising machines.
comment: 25 pages, 5 figures
☆ CREAM: Consistency Regularized Self-Rewarding Language Models
Recent self-rewarding large language models (LLM) have successfully applied LLM-as-a-Judge to iteratively improve the alignment performance without the need of human annotations for preference data. These methods commonly utilize the same LLM to act as both the policy model (which generates responses) and the reward model (which scores and ranks those responses). The ranked responses are then used as preference pairs to train the LLM via direct alignment technologies (e.g. DPO). However, it is noteworthy that throughout this process, there is no guarantee of accuracy in the rewarding and ranking, which is critical for ensuring accurate rewards and high-quality preference data. Empirical results from relatively small LLMs (e.g., 7B parameters) also indicate that improvements from self-rewarding may diminish after several iterations in certain situations, which we hypothesize is due to accumulated bias in the reward system. This bias can lead to unreliable preference data for training the LLM. To address this issue, we first formulate and analyze the generalized iterative preference fine-tuning framework for self-rewarding language model. We then introduce the regularization to this generalized framework to mitigate the overconfident preference labeling in the self-rewarding process. Based on this theoretical insight, we propose a Consistency Regularized sElf-rewarding lAnguage Model (CREAM) that leverages the rewarding consistency across different iterations to regularize the self-rewarding training, helping the model to learn from more reliable preference data. With this explicit regularization, our empirical results demonstrate the superiority of CREAM in improving both reward consistency and alignment performance. The code is publicly available at https://github.com/Raibows/CREAM.
☆ Counterfactual Generative Modeling with Variational Causal Inference
Estimating an individual's potential outcomes under counterfactual treatments is a challenging task for traditional causal inference and supervised learning approaches when the outcome is high-dimensional (e.g. gene expressions, facial images) and covariates are relatively limited. In this case, to predict one's outcomes under counterfactual treatments, it is crucial to leverage individual information contained in its high-dimensional observed outcome in addition to the covariates. Prior works using variational inference in counterfactual generative modeling have been focusing on neural adaptations and model variants within the conditional variational autoencoder formulation, which we argue is fundamentally ill-suited to the notion of counterfactual in causal inference. In this work, we present a novel variational Bayesian causal inference framework and its theoretical backings to properly handle counterfactual generative modeling tasks, through which we are able to conduct counterfactual supervision end-to-end during training without any counterfactual samples, and encourage latent disentanglement that aids the correct identification of causal effect in counterfactual generations. In experiments, we demonstrate the advantage of our framework compared to state-of-the-art models in counterfactual generative modeling on multiple benchmarks.
Transformer based super-resolution downscaling for regional reanalysis: Full domain vs tiling approaches
Super-resolution (SR) is a promising cost-effective downscaling methodology for producing high-resolution climate information from coarser counterparts. A particular application is downscaling regional reanalysis outputs (predictand) from the driving global counterparts (predictor). This study conducts an intercomparison of various SR downscaling methods focusing on temperature and using the CERRA reanalysis (5.5 km resolution, produced with a regional atmospheric model driven by ERA5) as example. The method proposed in this work is the Swin transformer and two alternative methods are used as benchmark (fully convolutional U-Net and convolutional and dense DeepESD) as well as the simple bicubic interpolation. We compare two approaches, the standard one using the full domain as input and a more scalable tiling approach, dividing the full domain into tiles that are used as input. The methods are trained to downscale CERRA surface temperature, based on temperature information from the driving ERA5; in addition, the tiling approach includes static orographic information. We show that the tiling approach, which requires spatial transferability, comes at the cost of a lower performance (although it outperforms some full-domain benchmarks), but provides an efficient scalable solution that allows SR reduction on a pan-European scale and is valuable for real-time applications.
☆ Optimizing 3D Geometry Reconstruction from Implicit Neural Representations
Implicit neural representations have emerged as a powerful tool in learning 3D geometry, offering unparalleled advantages over conventional representations like mesh-based methods. A common type of INR implicitly encodes a shape's boundary as the zero-level set of the learned continuous function and learns a mapping from a low-dimensional latent space to the space of all possible shapes represented by its signed distance function. However, most INRs struggle to retain high-frequency details, which are crucial for accurate geometric depiction, and they are computationally expensive. To address these limitations, we present a novel approach that both reduces computational expenses and enhances the capture of fine details. Our method integrates periodic activation functions, positional encodings, and normals into the neural network architecture. This integration significantly enhances the model's ability to learn the entire space of 3D shapes while preserving intricate details and sharp features, areas where conventional representations often fall short.
☆ How Does Variance Shape the Regret in Contextual Bandits? NeurIPS 2024
We consider realizable contextual bandits with general function approximation, investigating how small reward variance can lead to better-than-minimax regret bounds. Unlike in minimax bounds, we show that the eluder dimension $d_\text{elu}$$-$a complexity measure of the function class$-$plays a crucial role in variance-dependent bounds. We consider two types of adversary: (1) Weak adversary: The adversary sets the reward variance before observing the learner's action. In this setting, we prove that a regret of $\Omega(\sqrt{\min\{A,d_\text{elu}\}\Lambda}+d_\text{elu})$ is unavoidable when $d_{\text{elu}}\leq\sqrt{AT}$, where $A$ is the number of actions, $T$ is the total number of rounds, and $\Lambda$ is the total variance over $T$ rounds. For the $A\leq d_\text{elu}$ regime, we derive a nearly matching upper bound $\tilde{O}(\sqrt{A\Lambda}+d_\text{elu})$ for the special case where the variance is revealed at the beginning of each round. (2) Strong adversary: The adversary sets the reward variance after observing the learner's action. We show that a regret of $\Omega(\sqrt{d_\text{elu}\Lambda}+d_\text{elu})$ is unavoidable when $\sqrt{d_\text{elu}\Lambda}+d_\text{elu}\leq\sqrt{AT}$. In this setting, we provide an upper bound of order $\tilde{O}(d_\text{elu}\sqrt{\Lambda}+d_\text{elu})$. Furthermore, we examine the setting where the function class additionally provides distributional information of the reward, as studied by Wang et al. (2024). We demonstrate that the regret bound $\tilde{O}(\sqrt{d_\text{elu}\Lambda}+d_\text{elu})$ established in their work is unimprovable when $\sqrt{d_{\text{elu}}\Lambda}+d_\text{elu}\leq\sqrt{AT}$. However, with a slightly different definition of the total variance and with the assumption that the reward follows a Gaussian distribution, one can achieve a regret of $\tilde{O}(\sqrt{A\Lambda}+d_\text{elu})$.
comment: NeurIPS 2024
☆ On the sample complexity of purity and inner product estimation
We study the sample complexity of the prototypical tasks quantum purity estimation and quantum inner product estimation. In purity estimation, we are to estimate $tr(\rho^2)$ of an unknown quantum state $\rho$ to additive error $\epsilon$. Meanwhile, for quantum inner product estimation, Alice and Bob are to estimate $tr(\rho\sigma)$ to additive error $\epsilon$ given copies of unknown quantum state $\rho$ and $\sigma$ using classical communication and restricted quantum communication. In this paper, we show a strong connection between the sample complexity of purity estimation with bounded quantum memory and inner product estimation with bounded quantum communication and unentangled measurements. We propose a protocol that solves quantum inner product estimation with $k$-qubit one-way quantum communication and unentangled local measurements using $O(median\{1/\epsilon^2,2^{n/2}/\epsilon,2^{n-k}/\epsilon^2\})$ copies of $\rho$ and $\sigma$. Our protocol can be modified to estimate the purity of an unknown quantum state $\rho$ using $k$-qubit quantum memory with the same complexity. We prove that arbitrary protocols with $k$-qubit quantum memory that estimate purity to error $\epsilon$ require $\Omega(median\{1/\epsilon^2,2^{n/2}/\sqrt{\epsilon},2^{n-k}/\epsilon^2\})$ copies of $\rho$. This indicates the same lower bound for quantum inner product estimation with one-way $k$-qubit quantum communication and classical communication, and unentangled local measurements. For purity estimation, we further improve the lower bound to $\Omega(\max\{1/\epsilon^2,2^{n/2}/\epsilon\})$ for any protocols using an identical single-copy projection-valued measurement. Additionally, we investigate a decisional variant of quantum distributed inner product estimation without quantum communication for mixed state and provide a lower bound on the sample complexity.
comment: 33 pages, 1 figure
☆ FusionLLM: A Decentralized LLM Training System on Geo-distributed GPUs with Adaptive Compression
To alleviate hardware scarcity in training large deep neural networks (DNNs), particularly large language models (LLMs), we present FusionLLM, a decentralized training system designed and implemented for training DNNs using geo-distributed GPUs across different computing clusters or individual devices. Decentralized training faces significant challenges regarding system design and efficiency, including: 1) the need for remote automatic differentiation (RAD), 2) support for flexible model definitions and heterogeneous software, 3) heterogeneous hardware leading to low resource utilization or the straggler problem, and 4) slow network communication. To address these challenges, in the system design, we represent the model as a directed acyclic graph of operators (OP-DAG). Each node in the DAG represents the operator in the DNNs, while the edge represents the data dependency between operators. Based on this design, 1) users are allowed to customize any DNN without caring low-level operator implementation; 2) we enable the task scheduling with the more fine-grained sub-tasks, offering more optimization space; 3) a DAG runtime executor can implement RAD withour requiring the consistent low-level ML framework versions. To enhance system efficiency, we implement a workload estimator and design an OP-Fence scheduler to cluster devices with similar bandwidths together and partition the DAG to increase throughput. Additionally, we propose an AdaTopK compressor to adaptively compress intermediate activations and gradients at the slowest communication links. To evaluate the convergence and efficiency of our system and algorithms, we train ResNet-101 and GPT-2 on three real-world testbeds using 48 GPUs connected with 8 Mbps~10 Gbps networks. Experimental results demonstrate that our system and method can achieve 1.45 - 9.39x speedup compared to baseline methods while ensuring convergence.
☆ Sarcasm Detection in a Less-Resourced Language
The sarcasm detection task in natural language processing tries to classify whether an utterance is sarcastic or not. It is related to sentiment analysis since it often inverts surface sentiment. Because sarcastic sentences are highly dependent on context, and they are often accompanied by various non-verbal cues, the task is challenging. Most of related work focuses on high-resourced languages like English. To build a sarcasm detection dataset for a less-resourced language, such as Slovenian, we leverage two modern techniques: a machine translation specific medium-size transformer model, and a very large generative language model. We explore the viability of translated datasets and how the size of a pretrained transformer affects its ability to detect sarcasm. We train ensembles of detection models and evaluate models' performance. The results show that larger models generally outperform smaller ones and that ensembling can slightly improve sarcasm detection performance. Our best ensemble approach achieves an $\text{F}_1$-score of 0.765 which is close to annotators' agreement in the source language.
comment: 4 pages, published in the Slovenian Conference on Artificial Intelligence
☆ Neural-based Control for CubeSat Docking Maneuvers
Autonomous Rendezvous and Docking (RVD) have been extensively studied in recent years, addressing the stringent requirements of spacecraft dynamics variations and the limitations of GNC systems. This paper presents an innovative approach employing Artificial Neural Networks (ANN) trained through Reinforcement Learning (RL) for autonomous spacecraft guidance and control during the final phase of the rendezvous maneuver. The proposed strategy is easily implementable onboard and offers fast adaptability and robustness to disturbances by learning control policies from experience rather than relying on predefined models. Extensive Monte Carlo simulations within a relevant environment are conducted in 6DoF settings to validate our approach, along with hardware tests that demonstrate deployment feasibility. Our findings highlight the efficacy of RL in assuring the adaptability and efficiency of spacecraft RVD, offering insights into future mission expectations.
☆ Embedding an Ethical Mind: Aligning Text-to-Image Synthesis via Lightweight Value Optimization
Recent advancements in diffusion models trained on large-scale data have enabled the generation of indistinguishable human-level images, yet they often produce harmful content misaligned with human values, e.g., social bias, and offensive content. Despite extensive research on Large Language Models (LLMs), the challenge of Text-to-Image (T2I) model alignment remains largely unexplored. Addressing this problem, we propose LiVO (Lightweight Value Optimization), a novel lightweight method for aligning T2I models with human values. LiVO only optimizes a plug-and-play value encoder to integrate a specified value principle with the input prompt, allowing the control of generated images over both semantics and values. Specifically, we design a diffusion model-tailored preference optimization loss, which theoretically approximates the Bradley-Terry model used in LLM alignment but provides a more flexible trade-off between image quality and value conformity. To optimize the value encoder, we also develop a framework to automatically construct a text-image preference dataset of 86k (prompt, aligned image, violating image, value principle) samples. Without updating most model parameters and through adaptive value selection from the input prompt, LiVO significantly reduces harmful outputs and achieves faster convergence, surpassing several strong baselines and taking an initial step towards ethically aligned T2I models.
comment: Accepted by ACM Multimedia 2024. The dataset and code can be found at https://github.com/achernarwang/LiVO
☆ Machine Learning Approach to Brain Tumor Detection and Classification
Brain tumor detection and classification are critical tasks in medical image analysis, particularly in early-stage diagnosis, where accurate and timely detection can significantly improve treatment outcomes. In this study, we apply various statistical and machine learning models to detect and classify brain tumors using brain MRI images. We explore a variety of statistical models including linear, logistic, and Bayesian regressions, and the machine learning models including decision tree, random forest, single-layer perceptron, multi-layer perceptron, convolutional neural network (CNN), recurrent neural network, and long short-term memory. Our findings show that CNN outperforms other models, achieving the best performance. Additionally, we confirm that the CNN model can also work for multi-class classification, distinguishing between four categories of brain MRI images such as normal, glioma, meningioma, and pituitary tumor images. This study demonstrates that machine learning approaches are suitable for brain tumor detection and classification, facilitating real-world medical applications in assisting radiologists with early and accurate diagnosis.
comment: 7 pages, 2 figures, 2 tables
☆ Local transfer learning Gaussian process modeling, with applications to surrogate modeling of expensive computer simulators
A critical bottleneck for scientific progress is the costly nature of computer simulations for complex systems. Surrogate models provide an appealing solution: such models are trained on simulator evaluations, then used to emulate and quantify uncertainty on the expensive simulator at unexplored inputs. In many applications, one often has available data on related systems. For example, in designing a new jet turbine, there may be existing studies on turbines with similar configurations. A key question is how information from such "source" systems can be transferred for effective surrogate training on the "target" system of interest. We thus propose a new LOcal transfer Learning Gaussian Process (LOL-GP) model, which leverages a carefully-designed Gaussian process to transfer such information for surrogate modeling. The key novelty of the LOL-GP is a latent regularization model, which identifies regions where transfer should be performed and regions where it should be avoided. This "local transfer" property is desirable in scientific systems: at certain parameters, such systems may behave similarly and thus transfer is beneficial; at other parameters, they may behave differently and thus transfer is detrimental. By accounting for local transfer, the LOL-GP can rectify a critical limitation of "negative transfer" in existing transfer learning models, where the transfer of information worsens predictive performance. We derive a Gibbs sampling algorithm for efficient posterior predictive sampling on the LOL-GP, for both the multi-source and multi-fidelity transfer settings. We then show, via a suite of numerical experiments and an application for jet turbine design, the improved surrogate performance of the LOL-GP over existing methods.
☆ A distance function for stochastic matrices
Motivated by information geometry, a distance function on the space of stochastic matrices is advocated. Starting with sequences of Markov chains the Bhattacharyya angle is advocated as the natural tool for comparing both short and long term Markov chain runs. Bounds on the convergence of the distance and mixing times are derived. Guided by the desire to compare different Markov chain models, especially in the setting of healthcare processes, a new distance function on the space of stochastic matrices is presented. It is a true distance measure which has a closed form and is efficient to implement for numerical evaluation. In the case of ergodic Markov chains, it is shown that considering either the Bhattacharyya angle on Markov sequences or the new stochastic matrix distance leads to the same distance between models.
comment: 9 pages, 2 figures
☆ Automatic Mapping of Anatomical Landmarks from Free-Text Using Large Language Models: Insights from Llama-2
Anatomical landmarks are vital in medical imaging for navigation and anomaly detection. Modern large language models (LLMs), like Llama-2, offer promise for automating the mapping of these landmarks in free-text radiology reports to corresponding positions in image data. Recent studies propose LLMs may develop coherent representations of generative processes. Motivated by these insights, we investigated whether LLMs accurately represent the spatial positions of anatomical landmarks. Through experiments with Llama-2 models, we found that they can linearly represent anatomical landmarks in space with considerable robustness to different prompts. These results underscore the potential of LLMs to enhance the efficiency and accuracy of medical imaging workflows.
comment: 6 pages, 2 figures, 1 table
☆ Generative Neural Reparameterization for Differentiable PDE-constrained Optimization NeurIPS 2024
Partial-differential-equation (PDE)-constrained optimization is a well-worn technique for acquiring optimal parameters of systems governed by PDEs. However, this approach is limited to providing a single set of optimal parameters per optimization. Given a differentiable PDE solver, if the free parameters are reparameterized as the output of a neural network, that neural network can be trained to learn a map from a probability distribution to the distribution of optimal parameters. This proves useful in the case where there are many well performing local minima for the PDE. We apply this technique to train a neural network that generates optimal parameters that minimize laser-plasma instabilities relevant to laser fusion and show that the neural network generates many well performing and diverse minima.
comment: Accepted to D3S3: Data-driven and Differentiable Simulations, Surrogates, and Solvers - Workshop @ NeurIPS 2024
☆ Optimizing Multi-Task Learning for Accurate Spacecraft Pose Estimation
Accurate satellite pose estimation is crucial for autonomous guidance, navigation, and control (GNC) systems in in-orbit servicing (IOS) missions. This paper explores the impact of different tasks within a multi-task learning (MTL) framework for satellite pose estimation using monocular images. By integrating tasks such as direct pose estimation, keypoint prediction, object localization, and segmentation into a single network, the study aims to evaluate the reciprocal influence between tasks by testing different multi-task configurations thanks to the modularity of the convolutional neural network (CNN) used in this work. The trends of mutual bias between the analyzed tasks are found by employing different weighting strategies to further test the robustness of the findings. A synthetic dataset was developed to train and test the MTL network. Results indicate that direct pose estimation and heatmap-based pose estimation positively influence each other in general, while both the bounding box and segmentation tasks do not provide significant contributions and tend to degrade the overall estimation accuracy.
☆ Efficient Optimization Algorithms for Linear Adversarial Training
Adversarial training can be used to learn models that are robust against perturbations. For linear models, it can be formulated as a convex optimization problem. Compared to methods proposed in the context of deep learning, leveraging the optimization structure allows significantly faster convergence rates. Still, the use of generic convex solvers can be inefficient for large-scale problems. Here, we propose tailored optimization algorithms for the adversarial training of linear models, which render large-scale regression and classification problems more tractable. For regression problems, we propose a family of solvers based on iterative ridge regression and, for classification, a family of solvers based on projected gradient descent. The methods are based on extended variable reformulations of the original problem. We illustrate their efficiency in numerical examples.
☆ Context Matters: Leveraging Contextual Features for Time Series Forecasting
Time series forecasts are often influenced by exogenous contextual features in addition to their corresponding history. For example, in financial settings, it is hard to accurately predict a stock price without considering public sentiments and policy decisions in the form of news articles, tweets, etc. Though this is common knowledge, the current state-of-the-art (SOTA) forecasting models fail to incorporate such contextual information, owing to its heterogeneity and multimodal nature. To address this, we introduce ContextFormer, a novel plug-and-play method to surgically integrate multimodal contextual information into existing pre-trained forecasting models. ContextFormer effectively distills forecast-specific information from rich multimodal contexts, including categorical, continuous, time-varying, and even textual information, to significantly enhance the performance of existing base forecasters. ContextFormer outperforms SOTA forecasting models by up to 30% on a range of real-world datasets spanning energy, traffic, environmental, and financial domains.
☆ New Paradigm of Adversarial Training: Breaking Inherent Trade-Off between Accuracy and Robustness via Dummy Classes
Adversarial Training (AT) is one of the most effective methods to enhance the robustness of DNNs. However, existing AT methods suffer from an inherent trade-off between adversarial robustness and clean accuracy, which seriously hinders their real-world deployment. While this problem has been widely studied within the current AT paradigm, existing AT methods still typically experience a reduction in clean accuracy by over 10% to date, without significant improvements in robustness compared with simple baselines like PGD-AT. This inherent trade-off raises a question: whether the current AT paradigm, which assumes to learn the corresponding benign and adversarial samples as the same class, inappropriately combines clean and robust objectives that may be essentially inconsistent. In this work, we surprisingly reveal that up to 40% of CIFAR-10 adversarial samples always fail to satisfy such an assumption across various AT methods and robust models, explicitly indicating the improvement room for the current AT paradigm. Accordingly, to relax the tension between clean and robust learning derived from this overstrict assumption, we propose a new AT paradigm by introducing an additional dummy class for each original class, aiming to accommodate the hard adversarial samples with shifted distribution after perturbation. The robustness w.r.t. these adversarial samples can be achieved by runtime recovery from the predicted dummy classes to their corresponding original ones, eliminating the compromise with clean learning. Building on this new paradigm, we propose a novel plug-and-play AT technology named DUmmy Classes-based Adversarial Training (DUCAT). Extensive experiments on CIFAR-10, CIFAR-100, and Tiny-ImageNet demonstrate that the DUCAT concurrently improves clean accuracy and adversarial robustness compared with state-of-the-art benchmarks, effectively breaking the existing inherent trade-off.
comment: Preprint. Work in progress. The code is available at https://github.com/FlaAI/DUCAT
☆ Explanation-Preserving Augmentation for Semi-Supervised Graph Representation Learning
Graph representation learning (GRL), enhanced by graph augmentation methods, has emerged as an effective technique achieving performance improvements in wide tasks such as node classification and graph classification. In self-supervised GRL, paired graph augmentations are generated from each graph. Its objective is to infer similar representations for augmentations of the same graph, but maximally distinguishable representations for augmentations of different graphs. Analogous to image and language domains, the desiderata of an ideal augmentation method include both (1) semantics-preservation; and (2) data-perturbation; i.e., an augmented graph should preserve the semantics of its original graph while carrying sufficient variance. However, most existing (un-)/self-supervised GRL methods focus on data perturbation but largely neglect semantics preservation. To address this challenge, in this paper, we propose a novel method, Explanation-Preserving Augmentation (EPA), that leverages graph explanation techniques for generating augmented graphs that can bridge the gap between semantics-preservation and data-perturbation. EPA first uses a small number of labels to train a graph explainer to infer the sub-structures (explanations) that are most relevant to a graph's semantics. These explanations are then used to generate semantics-preserving augmentations for self-supervised GRL, namely EPA-GRL. We demonstrate theoretically, using an analytical example, and through extensive experiments on a variety of benchmark datasets that EPA-GRL outperforms the state-of-the-art (SOTA) GRL methods, which are built upon semantics-agnostic data augmentations.
comment: 16 pages, 7 figures, 7 tables
☆ Position Specific Scoring Is All You Need? Revisiting Protein Sequence Classification Tasks
Understanding the structural and functional characteristics of proteins are crucial for developing preventative and curative strategies that impact fields from drug discovery to policy development. An important and popular technique for examining how amino acids make up these characteristics of the protein sequences with position-specific scoring (PSS). While the string kernel is crucial in natural language processing (NLP), it is unclear if string kernels can extract biologically meaningful information from protein sequences, despite the fact that they have been shown to be effective in the general sequence analysis tasks. In this work, we propose a weighted PSS kernel matrix (or W-PSSKM), that combines a PSS representation of protein sequences, which encodes the frequency information of each amino acid in a sequence, with the notion of the string kernel. This results in a novel kernel function that outperforms many other approaches for protein sequence classification. We perform extensive experimentation to evaluate the proposed method. Our findings demonstrate that the W-PSSKM significantly outperforms existing baselines and state-of-the-art methods and achieves up to 45.1\% improvement in classification accuracy.
☆ Constrained Posterior Sampling: Time Series Generation with Hard Constraints
Generating realistic time series samples is crucial for stress-testing models and protecting user privacy by using synthetic data. In engineering and safety-critical applications, these samples must meet certain hard constraints that are domain-specific or naturally imposed by physics or nature. Consider, for example, generating electricity demand patterns with constraints on peak demand times. This can be used to stress-test the functioning of power grids during adverse weather conditions. Existing approaches for generating constrained time series are either not scalable or degrade sample quality. To address these challenges, we introduce Constrained Posterior Sampling (CPS), a diffusion-based sampling algorithm that aims to project the posterior mean estimate into the constraint set after each denoising update. Notably, CPS scales to a large number of constraints (~100) without requiring additional training. We provide theoretical justifications highlighting the impact of our projection step on sampling. Empirically, CPS outperforms state-of-the-art methods in sample quality and similarity to real time series by around 10% and 42%, respectively, on real-world stocks, traffic, and air quality datasets.
☆ Optimization and Application of Cloud-based Deep Learning Architecture for Multi-Source Data Prediction
This study develops a cloud-based deep learning system for early prediction of diabetes, leveraging the distributed computing capabilities of the AWS cloud platform and deep learning technologies to achieve efficient and accurate risk assessment. The system utilizes EC2 p3.8xlarge GPU instances to accelerate model training, reducing training time by 93.2% while maintaining a prediction accuracy of 94.2%. With an automated data processing and model training pipeline built using Apache Airflow, the system can complete end-to-end updates within 18.7 hours. In clinical applications, the system demonstrates a prediction accuracy of 89.8%, sensitivity of 92.3%, and specificity of 95.1%. Early interventions based on predictions lead to a 37.5% reduction in diabetes incidence among the target population. The system's high performance and scalability provide strong support for large-scale diabetes prevention and management, showcasing significant public health value.
comment: 6 Pages, 5 Figures, 3 Tables. The final version will be published in the proceedings of the IEEE conference
☆ Towards Arbitrary QUBO Optimization: Analysis of Classical and Quantum-Activated Feedforward Neural Networks
Quadratic Unconstrained Binary Optimization (QUBO) sits at the heart of many industries and academic fields such as logistics, supply chain, finance, pharmaceutical science, chemistry, IT, and energy sectors, among others. These problems typically involve optimizing a large number of binary variables, which makes finding exact solutions exponentially more difficult. Consequently, most QUBO problems are classified as NP-hard. To address this challenge, we developed a powerful feedforward neural network (FNN) optimizer for arbitrary QUBO problems. In this work, we demonstrate that the FNN optimizer can provide high-quality approximate solutions for large problems, including dense 80-variable weighted MaxCut and random QUBOs, achieving an average accuracy of over 99% in less than 1.1 seconds on an 8-core CPU. Additionally, the FNN optimizer outperformed the Gurobi optimizer by 72% on 200-variable random QUBO problems within a 100-second computation time limit, exhibiting strong potential for real-time optimization tasks. Building on this model, we explored the novel approach of integrating FNNs with a quantum annealer-based activation function to create a quantum-classical encoder-decoder (QCED) optimizer, aiming to further enhance the performance of FNNs in QUBO optimization.
☆ An Exact Finite-dimensional Explicit Feature Map for Kernel Functions
Kernel methods in machine learning use a kernel function that takes two data points as input and returns their inner product after mapping them to a Hilbert space, implicitly and without actually computing the mapping. For many kernel functions, such as Gaussian and Laplacian kernels, the feature space is known to be infinite-dimensional, making operations in this space possible only implicitly. This implicit nature necessitates algorithms to be expressed using dual representations and the kernel trick. In this paper, given an arbitrary kernel function, we introduce an explicit, finite-dimensional feature map for any arbitrary kernel function that ensures the inner product of data points in the feature space equals the kernel function value, during both training and testing. The existence of this explicit mapping allows for kernelized algorithms to be formulated in their primal form, without the need for the kernel trick or the dual representation. As a first application, we demonstrate how to derive kernelized machine learning algorithms directly, without resorting to the dual representation, and apply this method specifically to PCA. As another application, without any changes to the t-SNE algorithm and its implementation, we use it for visualizing the feature space of kernel functions.
☆ Explainable Moral Values: a neuro-symbolic approach to value classification ESWC24
This work explores the integration of ontology-based reasoning and Machine Learning techniques for explainable value classification. By relying on an ontological formalization of moral values as in the Moral Foundations Theory, relying on the DnS Ontology Design Pattern, the \textit{sandra} neuro-symbolic reasoner is used to infer values (fomalized as descriptions) that are \emph{satisfied by} a certain sentence. Sentences, alongside their structured representation, are automatically generated using an open-source Large Language Model. The inferred descriptions are used to automatically detect the value associated with a sentence. We show that only relying on the reasoner's inference results in explainable classification comparable to other more complex approaches. We show that combining the reasoner's inferences with distributional semantics methods largely outperforms all the baselines, including complex models based on neural network architectures. Finally, we build a visualization tool to explore the potential of theory-based values classification, which is publicly available at http://xmv.geomeaning.com/.
comment: Published at ESWC24 Satellite Event
☆ Weak-to-Strong Generalization beyond Accuracy: a Pilot Study in Safety, Toxicity, and Legal Reasoning
As large language models (LLMs) continue to advance, ensuring their alignment with human values becomes increasingly critical. Traditional alignment methods heavily rely on human feedback to fine-tune models. With the emergence of superhuman models whose outputs may surpass human understanding, evaluating and aligning these models using human judgments poses significant challenges. To address the challenges, recent works use weak supervisors to elicit knowledge from much stronger models. However, there are important disanalogies between the empirical setup in the existing works and the genuine goal of alignment. We remark that existing works investigate the phenomenon of weak-to-strong generation in analogous setup (i.e., binary classification), rather than practical alignment-relevant tasks (e.g., safety). In this paper, we bridge this gap by extending weak-to-strong generation to the context of practical alignment. We empirically demonstrate the widespread phenomenon of weak-to-strong generation in three complicated alignment tasks: safety, toxicity, and legal reasoning}. Furthermore, we explore efficient strategies for improving alignment performance to enhance the quality of model outcomes. Lastly, we summarize and analyze the challenges and potential solutions in regard to specific alignment tasks, which we hope to catalyze the research progress on the topic of weak-to-strong generalization. Our code is released at https://github.com/yeruimeng/WTS.git.
☆ Exploring Model Kinship for Merging Large Language Models
Model merging has become one of the key technologies for enhancing the capabilities and efficiency of Large Language Models (LLMs). However, our understanding of the expected performance gains and principles when merging any two models remains limited. In this work, we introduce model kinship, the degree of similarity or relatedness between LLMs, analogous to biological evolution. With comprehensive empirical analysis, we find that there is a certain relationship between model kinship and the performance gains after model merging, which can help guide our selection of candidate models. Inspired by this, we propose a new model merging strategy: Top-k Greedy Merging with Model Kinship, which can yield better performance on benchmark datasets. Specifically, we discover that using model kinship as a criterion can assist us in continuously performing model merging, alleviating the degradation (local optima) in model evolution, whereas model kinship can serve as a guide to escape these traps. Code is available at https://github.com/zjunlp/ModelKinship.
comment: Ongoing work
☆ Towards Graph Foundation Models: The Perspective of Zero-shot Reasoning on Knowledge Graphs
Inspired by the success of artificial general intelligence, there is a trend towards developing Graph Foundation Models that excel in generalization across various graph tasks and domains. However, current models often require extensive training or fine-tuning to capture structural and semantic insights on new graphs, which limits their versatility. In this work, we explore graph foundation models from the perspective of zero-shot reasoning on Knowledge Graphs (KGs). Our focus is on utilizing KGs as a unified topological structure to tackle diverse tasks, while addressing semantic isolation challenges in KG reasoning to effectively integrate diverse semantic and structural features. This brings us new methodological insights into KG reasoning, as well as high generalizability towards foundation models in practice. Methodologically, we introduce SCORE, a unified graph reasoning framework that effectively generalizes diverse graph tasks using zero-shot learning. At the core of SCORE is semantic conditional message passing, a technique designed to capture both structural and semantic invariances in graphs, with theoretical backing for its expressive power. Practically, we evaluate the zero-shot reasoning capability of SCORE using 38 diverse graph datasets, covering node-level, link-level, and graph-level tasks across multiple domains. Our experiments reveal a substantial performance improvement over prior foundation models and supervised baselines, highlighting the efficacy and adaptability of our approach.
comment: 17 Pages, 5 figures
☆ Low-Rank Adversarial PGD Attack
Adversarial attacks on deep neural network models have seen rapid development and are extensively used to study the stability of these networks. Among various adversarial strategies, Projected Gradient Descent (PGD) is a widely adopted method in computer vision due to its effectiveness and quick implementation, making it suitable for adversarial training. In this work, we observe that in many cases, the perturbations computed using PGD predominantly affect only a portion of the singular value spectrum of the original image, suggesting that these perturbations are approximately low-rank. Motivated by this observation, we propose a variation of PGD that efficiently computes a low-rank attack. We extensively validate our method on a range of standard models as well as robust models that have undergone adversarial training. Our analysis indicates that the proposed low-rank PGD can be effectively used in adversarial training due to its straightforward and fast implementation coupled with competitive performance. Notably, we find that low-rank PGD often performs comparably to, and sometimes even outperforms, the traditional full-rank PGD attack, while using significantly less memory.
Self-Supervised Learning of Disentangled Representations for Multivariate Time-Series NeurIPS 2024
Multivariate time-series data in fields like healthcare and industry are informative but challenging due to high dimensionality and lack of labels. Recent self-supervised learning methods excel in learning rich representations without labels but struggle with disentangled embeddings and inductive bias issues like transformation-invariance. To address these challenges, we introduce TimeDRL, a framework for multivariate time-series representation learning with dual-level disentangled embeddings. TimeDRL features: (i) disentangled timestamp-level and instance-level embeddings using a [CLS] token strategy; (ii) timestamp-predictive and instance-contrastive tasks for representation learning; and (iii) avoidance of augmentation methods to eliminate inductive biases. Experiments on forecasting and classification datasets show TimeDRL outperforms existing methods, with further validation in semi-supervised settings with limited labeled data.
comment: NeurIPS 2024 Workshop: Self-Supervised Learning - Theory and Practice
☆ The Bayesian Confidence (BACON) Estimator for Deep Neural Networks
This paper introduces the Bayesian Confidence Estimator (BACON) for deep neural networks. Current practice of interpreting Softmax values in the output layer as probabilities of outcomes is prone to extreme predictions of class probability. In this work we extend Waagen's method of representing the terminal layers with a geometric model, where the probability associated with an output vector is estimated with Bayes' Rule using validation data to provide likelihood and normalization values. This estimator provides superior ECE and ACE calibration error compared to Softmax for ResNet-18 at 85% network accuracy, and EfficientNet-B0 at 95% network accuracy, on the CIFAR-10 dataset with an imbalanced test set, except for very high accuracy edge cases. In addition, when using the ACE metric, BACON demonstrated improved calibration error when estimating probabilities for the imbalanced test set when using actual class distribution fractions.
comment: 14 pages, 15 figures (10 of which include sub-figures)
☆ Dynamic Learning Rate for Deep Reinforcement Learning: A Bandit Approach
In Deep Reinforcement Learning models trained using gradient-based techniques, the choice of optimizer and its learning rate are crucial to achieving good performance: higher learning rates can prevent the model from learning effectively, while lower ones might slow convergence. Additionally, due to the non-stationarity of the objective function, the best-performing learning rate can change over the training steps. To adapt the learning rate, a standard technique consists of using decay schedulers. However, these schedulers assume that the model is progressively approaching convergence, which may not always be true, leading to delayed or premature adjustments. In this work, we propose dynamic Learning Rate for deep Reinforcement Learning (LRRL), a meta-learning approach that selects the learning rate based on the agent's performance during training. LRRL is based on a multi-armed bandit algorithm, where each arm represents a different learning rate, and the bandit feedback is provided by the cumulative returns of the RL policy to update the arms' probability distribution. Our empirical results demonstrate that LRRL can substantially improve the performance of deep RL algorithms.
☆ Personalized Prediction Models for Changes in Knee Pain among Patients with Osteoarthritis Participating in Supervised Exercise and Education
Knee osteoarthritis (OA) is a widespread chronic condition that impairs mobility and diminishes quality of life. Despite the proven benefits of exercise therapy and patient education in managing the OA symptoms pain and functional limitations, these strategies are often underutilized. Personalized outcome prediction models can help motivate and engage patients, but the accuracy of existing models in predicting changes in knee pain remains insufficiently examined. To validate existing models and introduce a concise personalized model predicting changes in knee pain before to after participating in a supervised education and exercise therapy program (GLA:D) for knee OA patients. Our models use self-reported patient information and functional measures. To refine the number of variables, we evaluated the variable importance and applied clinical reasoning. We trained random forest regression models and compared the rate of true predictions of our models with those utilizing average values. We evaluated the performance of a full, continuous, and concise model including all 34, all 11 continuous, and the six most predictive variables respectively. All three models performed similarly and were comparable to the existing model, with R-squares of 0.31-0.32 and RMSEs of 18.65-18.85 - despite our increased sample size. Allowing a deviation of 15 VAS points from the true change in pain, our concise model and utilizing the average values estimated the change in pain at 58% and 51% correctly, respectively. Our supplementary analysis led to similar outcomes. Our concise personalized prediction model more accurately predicts changes in knee pain following the GLA:D program compared to average pain improvement values. Neither the increase in sample size nor the inclusion of additional variables improved previous models. To improve predictions, new variables beyond those in the GLA:D are required.
☆ Expand and Compress: Exploring Tuning Principles for Continual Spatio-Temporal Graph Forecasting
The widespread deployment of sensing devices leads to a surge in data for spatio-temporal forecasting applications such as traffic flow, air quality, and wind energy. Although spatio-temporal graph neural networks have achieved success in modeling various static spatio-temporal forecasting scenarios, real-world spatio-temporal data are typically received in a streaming manner, and the network continuously expands with the installation of new sensors. Thus, spatio-temporal forecasting in streaming scenarios faces dual challenges: the inefficiency of retraining models over newly arrived data and the detrimental effects of catastrophic forgetting over long-term history. To address these challenges, we propose a novel prompt tuning-based continuous forecasting method, following two fundamental tuning principles guided by empirical and theoretical analysis: expand and compress, which effectively resolve the aforementioned problems with lightweight tuning parameters. Specifically, we integrate the base spatio-temporal graph neural network with a continuous prompt pool, utilizing stored prompts (i.e., few learnable parameters) in memory, and jointly optimize them with the base spatio-temporal graph neural network. This method ensures that the model sequentially learns from the spatio-temporal data stream to accomplish tasks for corresponding periods. Extensive experimental results on multiple real-world datasets demonstrate the multi-faceted superiority of our method over the state-of-the-art baselines, including effectiveness, efficiency, universality, etc.
☆ Cocoon: Robust Multi-Modal Perception with Uncertainty-Aware Sensor Fusion
An important paradigm in 3D object detection is the use of multiple modalities to enhance accuracy in both normal and challenging conditions, particularly for long-tail scenarios. To address this, recent studies have explored two directions of adaptive approaches: MoE-based adaptive fusion, which struggles with uncertainties arising from distinct object configurations, and late fusion for output-level adaptive fusion, which relies on separate detection pipelines and limits comprehensive understanding. In this work, we introduce Cocoon, an object- and feature-level uncertainty-aware fusion framework. The key innovation lies in uncertainty quantification for heterogeneous representations, enabling fair comparison across modalities through the introduction of a feature aligner and a learnable surrogate ground truth, termed feature impression. We also define a training objective to ensure that their relationship provides a valid metric for uncertainty quantification. Cocoon consistently outperforms existing static and adaptive methods in both normal and challenging conditions, including those with natural and artificial corruptions. Furthermore, we show the validity and efficacy of our uncertainty metric across diverse datasets.
comment: 23 pages
☆ From Lab to Pocket: A Novel Continual Learning-based Mobile Application for Screening COVID-19
Artificial intelligence (AI) has emerged as a promising tool for predicting COVID-19 from medical images. In this paper, we propose a novel continual learning-based approach and present the design and implementation of a mobile application for screening COVID-19. Our approach demonstrates the ability to adapt to evolving datasets, including data collected from different locations or hospitals, varying virus strains, and diverse clinical presentations, without retraining from scratch. We have evaluated state-of-the-art continual learning methods for detecting COVID-19 from chest X-rays and selected the best-performing model for our mobile app. We evaluated various deep learning architectures to select the best-performing one as a foundation model for continual learning. Both regularization and memory-based methods for continual learning were tested, using different memory sizes to develop the optimal continual learning model for our app. DenseNet161 emerged as the best foundation model with 96.87\% accuracy, and Learning without Forgetting (LwF) was the top continual learning method with an overall performance of 71.99\%. The mobile app design considers both patient and doctor perspectives. It incorporates the continual learning DenseNet161 LwF model on a cloud server, enabling the model to learn from new instances of chest X-rays and their classifications as they are submitted. The app is designed, implemented, and evaluated to ensure it provides an efficient tool for COVID-19 screening. The app is available to download from https://github.com/DannyFGitHub/COVID-19PneumoCheckApp.
comment: 31 pages
☆ Self-DenseMobileNet: A Robust Framework for Lung Nodule Classification using Self-ONN and Stacking-based Meta-Classifier
In this study, we propose a novel and robust framework, Self-DenseMobileNet, designed to enhance the classification of nodules and non-nodules in chest radiographs (CXRs). Our approach integrates advanced image standardization and enhancement techniques to optimize the input quality, thereby improving classification accuracy. To enhance predictive accuracy and leverage the strengths of multiple models, the prediction probabilities from Self-DenseMobileNet were transformed into tabular data and used to train eight classical machine learning (ML) models; the top three performers were then combined via a stacking algorithm, creating a robust meta-classifier that integrates their collective insights for superior classification performance. To enhance the interpretability of our results, we employed class activation mapping (CAM) to visualize the decision-making process of the best-performing model. Our proposed framework demonstrated remarkable performance on internal validation data, achieving an accuracy of 99.28\% using a Meta-Random Forest Classifier. When tested on an external dataset, the framework maintained strong generalizability with an accuracy of 89.40\%. These results highlight a significant improvement in the classification of CXRs with lung nodules.
comment: 31 pages
☆ On the Role of Activation Functions in EEG-To-Text Decoder
In recent years, much interdisciplinary research has been conducted exploring potential use cases of neuroscience to advance the field of information retrieval. Initial research concentrated on the use of fMRI data, but fMRI was deemed to be not suitable for real-world applications, and soon, research shifted towards using EEG data. In this paper, we try to improve the original performance of a first attempt at generating text using EEG by focusing on the less explored area of optimising neural network performance. We test a set of different activation functions and compare their performance. Our results show that introducing a higher degree polynomial activation function can enhance model performance without changing the model architecture. We also show that the learnable 3rd-degree activation function performs better on the 1-gram evaluation compared to a 3rd-degree non-learnable function. However, when evaluating the model on 2-grams and above, the polynomial function lacks in performance, whilst the leaky ReLU activation function outperforms the baseline.
☆ One Step Diffusion via Shortcut Models
Diffusion models and flow-matching models have enabled generating diverse and realistic images by learning to transfer noise to data. However, sampling from these models involves iterative denoising over many neural network passes, making generation slow and expensive. Previous approaches for speeding up sampling require complex training regimes, such as multiple training phases, multiple networks, or fragile scheduling. We introduce shortcut models, a family of generative models that use a single network and training phase to produce high-quality samples in a single or multiple sampling steps. Shortcut models condition the network not only on the current noise level but also on the desired step size, allowing the model to skip ahead in the generation process. Across a wide range of sampling step budgets, shortcut models consistently produce higher quality samples than previous approaches, such as consistency models and reflow. Compared to distillation, shortcut models reduce complexity to a single network and training phase and additionally allow varying step budgets at inference time.
☆ Investigating Sensitive Directions in GPT-2: An Improved Baseline and Comparative Analysis of SAEs
Sensitive directions experiments attempt to understand the computational features of Language Models (LMs) by measuring how much the next token prediction probabilities change by perturbing activations along specific directions. We extend the sensitive directions work by introducing an improved baseline for perturbation directions. We demonstrate that KL divergence for Sparse Autoencoder (SAE) reconstruction errors are no longer pathologically high compared to the improved baseline. We also show that feature directions uncovered by SAEs have varying impacts on model outputs depending on the SAE's sparsity, with lower L0 SAE feature directions exerting a greater influence. Additionally, we find that end-to-end SAE features do not exhibit stronger effects on model outputs compared to traditional SAEs.
☆ Evaluating Utility of Memory Efficient Medical Image Generation: A Study on Lung Nodule Segmentation
The scarcity of publicly available medical imaging data limits the development of effective AI models. This work proposes a memory-efficient patch-wise denoising diffusion probabilistic model (DDPM) for generating synthetic medical images, focusing on CT scans with lung nodules. Our approach generates high-utility synthetic images with nodule segmentation while efficiently managing memory constraints, enabling the creation of training datasets. We evaluate the method in two scenarios: training a segmentation model exclusively on synthetic data, and augmenting real-world training data with synthetic images. In the first case, models trained solely on synthetic data achieve Dice scores comparable to those trained on real-world data benchmarks. In the second case, augmenting real-world data with synthetic images significantly improves segmentation performance. The generated images demonstrate their potential to enhance medical image datasets in scenarios with limited real-world data.
☆ Is Complex Query Answering Really Complex?
Complex query answering (CQA) on knowledge graphs (KGs) is gaining momentum as a challenging reasoning task. In this paper, we show that the current benchmarks for CQA are not really complex, and the way they are built distorts our perception of progress in this field. For example, we find that in these benchmarks, most queries (up to 98% for some query types) can be reduced to simpler problems, e.g., link prediction, where only one link needs to be predicted. The performance of state-of-the-art CQA models drops significantly when such models are evaluated on queries that cannot be reduced to easier types. Thus, we propose a set of more challenging benchmarks, composed of queries that require models to reason over multiple hops and better reflect the construction of real-world KGs. In a systematic empirical investigation, the new benchmarks show that current methods leave much to be desired from current CQA methods.
☆ SiFiSinger: A High-Fidelity End-to-End Singing Voice Synthesizer based on Source-filter Model ICASSP 2024
This paper presents an advanced end-to-end singing voice synthesis (SVS) system based on the source-filter mechanism that directly translates lyrical and melodic cues into expressive and high-fidelity human-like singing. Similarly to VISinger 2, the proposed system also utilizes training paradigms evolved from VITS and incorporates elements like the fundamental pitch (F0) predictor and waveform generation decoder. To address the issue that the coupling of mel-spectrogram features with F0 information may introduce errors during F0 prediction, we consider two strategies. Firstly, we leverage mel-cepstrum (mcep) features to decouple the intertwined mel-spectrogram and F0 characteristics. Secondly, inspired by the neural source-filter models, we introduce source excitation signals as the representation of F0 in the SVS system, aiming to capture pitch nuances more accurately. Meanwhile, differentiable mcep and F0 losses are employed as the waveform decoder supervision to fortify the prediction accuracy of speech envelope and pitch in the generated speech. Experiments on the Opencpop dataset demonstrate efficacy of the proposed model in synthesis quality and intonation accuracy.
comment: Accepted by ICASSP 2024, Synthesized audio samples are available at: https://sounddemos.github.io/sifisinger
☆ Disentangling data distribution for Federated Learning
Federated Learning (FL) facilitates collaborative training of a global model whose performance is boosted by private data owned by distributed clients, without compromising data privacy. Yet the wide applicability of FL is hindered by entanglement of data distributions across different clients. This paper demonstrates for the first time that by disentangling data distributions FL can in principle achieve efficiencies comparable to those of distributed systems, requiring only one round of communication. To this end, we propose a novel FedDistr algorithm, which employs stable diffusion models to decouple and recover data distributions. Empirical results on the CIFAR100 and DomainNet datasets show that FedDistr significantly enhances model utility and efficiency in both disentangled and near-disentangled scenarios while ensuring privacy, outperforming traditional federated learning methods.
☆ MING: A Functional Approach to Learning Molecular Generative Models
Traditional molecule generation methods often rely on sequence or graph-based representations, which can limit their expressive power or require complex permutation-equivariant architectures. This paper introduces a novel paradigm for learning molecule generative models based on functional representations. Specifically, we propose Molecular Implicit Neural Generation (MING), a diffusion-based model that learns molecular distributions in function space. Unlike standard diffusion processes in data space, MING employs a novel functional denoising probabilistic process, which jointly denoises the information in both the function's input and output spaces by leveraging an expectation-maximization procedure for latent implicit neural representations of data. This approach allows for a simple yet effective model design that accurately captures underlying function distributions. Experimental results on molecule-related datasets demonstrate MING's superior performance and ability to generate plausible molecular samples, surpassing state-of-the-art data-space methods while offering a more streamlined architecture and significantly faster generation times.
☆ End-to-end Planner Training for Language Modeling
Through end-to-end training to predict the next token, LLMs have become valuable tools for various tasks. Enhancing their core training in language modeling can improve numerous downstream applications. A successful approach to enhance language modeling uses a separate planning module to predict abstract labels of future sentences and conditions the LM on these predictions. However, this method is non-differentiable, preventing joint end-to-end tuning of the planner with the LM. We propose an effective method to improve this approach by enabling joint fine-tuning of the planner and the LM. We show that a naive way of approximating the gradient of selecting a label via the straight-through estimator is not effective. Instead, we propose to use the predicted label probabilities as mixing weights to condition the LM on a weighted average of label embeddings in a differentiable manner. This not only enables joint fine-tuning of the planner and the LM, but also allows the LM to draw on the full label distribution predicted by the planner, retaining more information. Our experimental results show consistent improvements in perplexity.
comment: 14 pages
☆ Data-Driven Gyroscope Calibration
Gyroscopes are inertial sensors that measure the angular velocity of the platforms to which they are attached. To estimate the gyroscope deterministic error terms prior mission start, a calibration procedure is performed. When considering low-cost gyroscopes, the calibration requires a turntable as the gyros are incapable of sensing the Earth turn rate. In this paper, we propose a data-driven framework to estimate the scale factor and bias of a gyroscope. To train and validate our approach, a dataset of 56 minutes was recorded using a turntable. We demonstrated that our proposed approach outperforms the model-based approach, in terms of accuracy and convergence time. Specifically, we improved the scale factor and bias estimation by an average of 72% during six seconds of calibration time, demonstrating an average of 75% calibration time improvement. That is, instead of minutes, our approach requires only several seconds for the calibration.
comment: 19 Pages, 5 Figures, 3 Tables
☆ SAC-GLAM: Improving Online RL for LLM agents with Soft Actor-Critic and Hindsight Relabeling
The past years have seen Large Language Models (LLMs) strive not only as generative models but also as agents solving textual sequential decision-making tasks. When facing complex environments where their zero-shot abilities are insufficient, recent work showed online Reinforcement Learning (RL) could be used for the LLM agent to discover and learn efficient strategies interactively. However, most prior work sticks to on-policy algorithms, which greatly reduces the scope of methods such agents could use for both exploration and exploitation, such as experience replay and hindsight relabeling. Yet, such methods may be key for LLM learning agents, and in particular when designing autonomous intrinsically motivated agents sampling and pursuing their own goals (i.e. autotelic agents). This paper presents and studies an adaptation of Soft Actor-Critic and hindsight relabeling to LLM agents. Our method not only paves the path towards autotelic LLM agents that learn online but can also outperform on-policy methods in more classic multi-goal RL environments.
☆ KcMF: A Knowledge-compliant Framework for Schema and Entity Matching with Fine-tuning-free LLMs
Schema and entity matching tasks are crucial for data integration and management. While large language models (LLMs) have shown promising results in these tasks, they suffer from hallucinations and confusion about task instructions. In this paper, we present the Knowledge-Compliant Matching Framework (KcMF), an LLM-based approach that addresses these issues without the need for domain-specific fine-tuning. KcMF employs a pseudo-code-based task decomposition strategy to adopt task-specific natural language statements that guide LLM reasoning and reduce confusion. We also propose two mechanisms, Dataset as Knowledge (DaK) and Example as Knowledge (EaK), to build domain knowledge sets when unstructured domain knowledge is lacking. Additionally, we introduce a result-ensembling strategy to leverage multiple knowledge sources and suppress poorly formatted outputs. Comprehensive evaluations on schema and entity matching tasks demonstrate that KcMF outperforms previous non-LLM state-of-the-art (SOTA) methods by an average F1 score of 22.9% and competes effectively with SOTA fine-tuned LLMs. Moreover, KcMF generalizes well across different LLMs.
☆ Retrieval-Reasoning Large Language Model-based Synthetic Clinical Trial Generation
Machine learning (ML) exhibits promise in the clinical domain. However, it is constrained by data scarcity and ethical considerations, as the generation of clinical trials presents significant challenges due to stringent privacy regulations, high costs, and the extended duration required for conducting studies with human participants. Despite the advancements of large language models (LLMs) in general generation tasks, their potential in facilitating the generation of synthetic clinical trials is under-explored. To address this gap, we introduce a novel Retrieval-Reasoning few-shot framework that leverages LLMs to generate artificial yet realistic and diverse clinical trials with binary success/failure labels. Experiments conducted on real clinical trials from the \url{ClinicalTrials.gov} database demonstrate that our synthetic data can effectively augment real datasets. Furthermore, by fine-tuning a pre-trained model as a binary classifier on synthetic clinical trial datasets, we demonstrate that this augmentation enhances model training for downstream tasks such as trial outcome prediction. Our findings suggest that LLMs for synthetic clinical trial generation hold promise for accelerating clinical research and upholding ethical standards for patient privacy. The code is publicly available at https://anonymous.4open.science/r/Retrieval_Reasoning_Clinical_Trial_Generation-3EC4.
☆ Mind the Gap Between Prototypes and Images in Cross-domain Finetuning
In cross-domain few-shot classification (CFC), recent works mainly focus on adapting a simple transformation head on top of a frozen pre-trained backbone with few labeled data to project embeddings into a task-specific metric space where classification can be performed by measuring similarities between image instance and prototype representations. Technically, an assumption implicitly adopted in such a framework is that the prototype and image instance embeddings share the same representation transformation. However, in this paper, we find that there naturally exists a gap, which resembles the modality gap, between the prototype and image instance embeddings extracted from the frozen pre-trained backbone, and simply applying the same transformation during the adaptation phase constrains exploring the optimal representations and shrinks the gap between prototype and image representations. To solve this problem, we propose a simple yet effective method, contrastive prototype-image adaptation (CoPA), to adapt different transformations respectively for prototypes and images similarly to CLIP by treating prototypes as text prompts. Extensive experiments on Meta-Dataset demonstrate that CoPA achieves the state-of-the-art performance more efficiently. Meanwhile, further analyses also indicate that CoPA can learn better representation clusters, enlarge the gap, and achieve minimal validation loss at the enlarged gap.
☆ Challenges, Methods, Data -- a Survey of Machine Learning in Water Distribution Networks ICANN 2024
Research on methods for planning and controlling water distribution networks gains increasing relevance as the availability of drinking water will decrease as a consequence of climate change. So far, the majority of approaches is based on hydraulics and engineering expertise. However, with the increasing availability of sensors, machine learning techniques constitute a promising tool. This work presents the main tasks in water distribution networks, discusses how they relate to machine learning and analyses how the particularities of the domain pose challenges to and can be leveraged by machine learning approaches. Besides, it provides a technical toolkit by presenting evaluation benchmarks and a structured survey of the exemplary task of leakage detection and localization.
comment: This preprint has not undergone any post-submission improvements or corrections. The Version of Record of this contribution is published in Artificial Neural Networks and Machine Learning -- ICANN 2024
☆ HELM: Hierarchical Encoding for mRNA Language Modeling
Messenger RNA (mRNA) plays a crucial role in protein synthesis, with its codon structure directly impacting biological properties. While Language Models (LMs) have shown promise in analyzing biological sequences, existing approaches fail to account for the hierarchical nature of mRNA's codon structure. We introduce Hierarchical Encoding for mRNA Language Modeling (HELM), a novel pre-training strategy that incorporates codon-level hierarchical structure into language model training. HELM modulates the loss function based on codon synonymity, aligning the model's learning process with the biological reality of mRNA sequences. We evaluate HELM on diverse mRNA datasets and tasks, demonstrating that HELM outperforms standard language model pre-training as well as existing foundation model baselines on six diverse downstream property prediction tasks and an antibody region annotation tasks on average by around 8\%. Additionally, HELM enhances the generative capabilities of language model, producing diverse mRNA sequences that better align with the underlying true data distribution compared to non-hierarchical baselines.
☆ Sharpness-Aware Black-Box Optimization
Black-box optimization algorithms have been widely used in various machine learning problems, including reinforcement learning and prompt fine-tuning. However, directly optimizing the training loss value, as commonly done in existing black-box optimization methods, could lead to suboptimal model quality and generalization performance. To address those problems in black-box optimization, we propose a novel Sharpness-Aware Black-box Optimization (SABO) algorithm, which applies a sharpness-aware minimization strategy to improve the model generalization. Specifically, the proposed SABO method first reparameterizes the objective function by its expectation over a Gaussian distribution. Then it iteratively updates the parameterized distribution by approximated stochastic gradients of the maximum objective value within a small neighborhood around the current solution in the Gaussian distribution space. Theoretically, we prove the convergence rate and generalization bound of the proposed SABO algorithm. Empirically, extensive experiments on the black-box prompt fine-tuning tasks demonstrate the effectiveness of the proposed SABO method in improving model generalization performance.
comment: 27 pages, 5 figures
☆ Training Neural Samplers with Reverse Diffusive KL Divergence
Training generative models to sample from unnormalized density functions is an important and challenging task in machine learning. Traditional training methods often rely on the reverse Kullback-Leibler (KL) divergence due to its tractability. However, the mode-seeking behavior of reverse KL hinders effective approximation of multi-modal target distributions. To address this, we propose to minimize the reverse KL along diffusion trajectories of both model and target densities. We refer to this objective as the reverse diffusive KL divergence, which allows the model to capture multiple modes. Leveraging this objective, we train neural samplers that can efficiently generate samples from the target distribution in one step. We demonstrate that our method enhances sampling performance across various Boltzmann distributions, including both synthetic multi-modal densities and n-body particle systems.
comment: 23 pages, 6 figures, 3 tables, 1 algorithm
☆ Loss Landscape Characterization of Neural Networks without Over-Parametrziation
Optimization methods play a crucial role in modern machine learning, powering the remarkable empirical achievements of deep learning models. These successes are even more remarkable given the complex non-convex nature of the loss landscape of these models. Yet, ensuring the convergence of optimization methods requires specific structural conditions on the objective function that are rarely satisfied in practice. One prominent example is the widely recognized Polyak-Lojasiewicz (PL) inequality, which has gained considerable attention in recent years. However, validating such assumptions for deep neural networks entails substantial and often impractical levels of over-parametrization. In order to address this limitation, we propose a novel class of functions that can characterize the loss landscape of modern deep models without requiring extensive over-parametrization and can also include saddle points. Crucially, we prove that gradient-based optimizers possess theoretical guarantees of convergence under this assumption. Finally, we validate the soundness of our new function class through both theoretical analysis and empirical experimentation across a diverse range of deep learning models.
☆ FairGLVQ: Fairness in Partition-Based Classification
Fairness is an important objective throughout society. From the distribution of limited goods such as education, over hiring and payment, to taxes, legislation, and jurisprudence. Due to the increasing importance of machine learning approaches in all areas of daily life including those related to health, security, and equity, an increasing amount of research focuses on fair machine learning. In this work, we focus on the fairness of partition- and prototype-based models. The contribution of this work is twofold: 1) we develop a general framework for fair machine learning of partition-based models that does not depend on a specific fairness definition, and 2) we derive a fair version of learning vector quantization (LVQ) as a specific instantiation. We compare the resulting algorithm against other algorithms from the literature on theoretical and real-world data showing its practical relevance.
comment: This preprint has not undergone any post-submission improvements or corrections. The Version of Record of this contribution is published in Advances in Self-Organizing Maps, Learning Vector Quantization, Interpretable Machine Learning, and Beyond
☆ Reconstruction of Differentially Private Text Sanitization via Large Language Models
Differential privacy (DP) is the de facto privacy standard against privacy leakage attacks, including many recently discovered ones against large language models (LLMs). However, we discovered that LLMs could reconstruct the altered/removed privacy from given DP-sanitized prompts. We propose two attacks (black-box and white-box) based on the accessibility to LLMs and show that LLMs could connect the pair of DP-sanitized text and the corresponding private training data of LLMs by giving sample text pairs as instructions (in the black-box attacks) or fine-tuning data (in the white-box attacks). To illustrate our findings, we conduct comprehensive experiments on modern LLMs (e.g., LLaMA-2, LLaMA-3, ChatGPT-3.5, ChatGPT-4, ChatGPT-4o, Claude-3, Claude-3.5, OPT, GPT-Neo, GPT-J, Gemma-2, and Pythia) using commonly used datasets (such as WikiMIA, Pile-CC, and Pile-Wiki) against both word-level and sentence-level DP. The experimental results show promising recovery rates, e.g., the black-box attacks against the word-level DP over WikiMIA dataset gave 72.18% on LLaMA-2 (70B), 82.39% on LLaMA-3 (70B), 75.35% on Gemma-2, 91.2% on ChatGPT-4o, and 94.01% on Claude-3.5 (Sonnet). More urgently, this study indicates that these well-known LLMs have emerged as a new security risk for existing DP text sanitization approaches in the current environment.
☆ ConLUX: Concept-Based Local Unified Explanations
With the rapid advancements of various machine learning models, there is a significant demand for model-agnostic explanation techniques, which can explain these models across different architectures. Mainstream model-agnostic explanation techniques generate local explanations based on basic features (e.g., words for text models and (super-)pixels for image models). However, these explanations often do not align with the decision-making processes of the target models and end-users, resulting in explanations that are unfaithful and difficult for users to understand. On the other hand, concept-based techniques provide explanations based on high-level features (e.g., topics for text models and objects for image models), but most are model-specific or require additional pre-defined external concept knowledge. To address this limitation, we propose \toolname, a general framework to provide concept-based local explanations for any machine learning models. Our key insight is that we can automatically extract high-level concepts from large pre-trained models, and uniformly extend existing local model-agnostic techniques to provide unified concept-based explanations. We have instantiated \toolname on four different types of explanation techniques: LIME, Kernel SHAP, Anchor, and LORE, and applied these techniques to text and image models. Our evaluation results demonstrate that 1) compared to the vanilla versions, \toolname offers more faithful explanations and makes them more understandable to users, and 2) by offering multiple forms of explanations, \toolname outperforms state-of-the-art concept-based explanation techniques specifically designed for text and image models, respectively.
☆ Approaching Metaheuristic Deep Learning Combos for Automated Data Mining
Lack of data on which to perform experimentation is a recurring issue in many areas of research, particularly in machine learning. The inability of most automated data mining techniques to be generalized to all types of data is inherently related with their dependency on those types which deems them ineffective against anything slightly different. Meta-heuristics are algorithms which attempt to optimize some solution independently of the type of data used, whilst classifiers or neural networks focus on feature extrapolation and dimensionality reduction to fit some model onto data arranged in a particular way. These two algorithmic fields encompass a group of characteristics which when combined are seemingly capable of achieving data mining regardless of how it is arranged. To this end, this work proposes a means of combining meta-heuristic methods with conventional classifiers and neural networks in order to perform automated data mining. Experiments on the MNIST dataset for handwritten digit recognition were performed and it was empirically observed that using a ground truth labeled dataset's validation accuracy is inadequate for correcting labels of other previously unseen data instances.
comment: Tentative submission for data mining and knowledge discovery
☆ Perseus: Leveraging Common Data Patterns with Curriculum Learning for More Robust Graph Neural Networks
Graph Neural Networks (GNNs) excel at handling graph data but remain vulnerable to adversarial attacks. Existing defense methods typically rely on assumptions like graph sparsity and homophily to either preprocess the graph or guide structure learning. However, preprocessing methods often struggle to accurately distinguish between normal edges and adversarial perturbations, leading to suboptimal results due to the loss of valuable edge information. Robust graph neural network models train directly on graph data affected by adversarial perturbations, without preprocessing. This can cause the model to get stuck in poor local optima, negatively affecting its performance. To address these challenges, we propose Perseus, a novel adversarial defense method based on curriculum learning. Perseus assesses edge difficulty using global homophily and applies a curriculum learning strategy to adjust the learning order, guiding the model to learn the full graph structure while adaptively focusing on common data patterns. This approach mitigates the impact of adversarial perturbations. Experiments show that models trained with Perseus achieve superior performance and are significantly more robust to adversarial attacks.
☆ Nonlinear bayesian tomography of ion temperature and velocity for Doppler coherence imaging spectroscopy in RT-1
We present a novel Bayesian tomography approach for Coherence Imaging Spectroscopy (CIS) that simultaneously reconstructs ion temperature and velocity distributions in plasmas. Utilizing nonlinear Gaussian Process Tomography (GPT) with the Laplace approximation, we model prior distributions of log-emissivity, temperature, and velocity as Gaussian processes. This framework rigorously incorporates nonlinear effects and temperature dependencies often neglected in conventional CIS tomography, enabling robust reconstruction even in the region of high temperature and velocity. By applying a log-Gaussian process, we also address issues like velocity divergence in low-emissivity regions. Validated with phantom simulations and experimental data from the RT-1 device, our method reveals detailed spatial structures of ion temperature and toroidal ion flow characteristic of magnetospheric plasma. This work significantly broadens the scope of CIS tomography, offering a robust tool for plasma diagnostics and facilitating integration with complementary measurement techniques.
comment: 13 page, 9 figures
☆ Tracking Universal Features Through Fine-Tuning and Model Merging
We study how features emerge, disappear, and persist across models fine-tuned on different domains of text. More specifically, we start from a base one-layer Transformer language model that is trained on a combination of the BabyLM corpus, and a collection of Python code from The Stack. This base model is adapted to two new domains of text: TinyStories, and the Lua programming language, respectively; and then these two models are merged using these two models using spherical linear interpolation. Our exploration aims to provide deeper insights into the stability and transformation of features across typical transfer-learning scenarios using small-scale models and sparse auto-encoders.
☆ Adaptive and Stratified Subsampling Techniques for High Dimensional Non-Standard Data Environments
This paper addresses the challenge of estimating high-dimensional parameters in non-standard data environments, where traditional methods often falter due to issues such as heavy-tailed distributions, data contamination, and dependent observations. We propose robust subsampling techniques, specifically Adaptive Importance Sampling (AIS) and Stratified Subsampling, designed to enhance the reliability and efficiency of parameter estimation. Under some clearly outlined conditions, we establish consistency and asymptotic normality for the proposed estimators, providing non-asymptotic error bounds that quantify their performance. Our theoretical foundations are complemented by controlled experiments demonstrating the superiority of our methods over conventional approaches. By bridging the gap between theory and practice, this work offers significant contributions to robust statistical estimation, paving the way for advancements in various applied domains.
☆ Towards Neural Scaling Laws for Time Series Foundation Models
Scaling laws offer valuable insights into the design of time series foundation models (TSFMs). However, previous research has largely focused on the scaling laws of TSFMs for in-distribution (ID) data, leaving their out-of-distribution (OOD) scaling behavior and the influence of model architectures less explored. In this work, we examine two common TSFM architectures, encoder-only and decoder-only Transformers, and investigate their scaling behavior on both ID and OOD data. These models are trained and evaluated across varying parameter counts, compute budgets, and dataset sizes. Our experiments reveal that the log-likelihood loss of TSFMs exhibits similar scaling behavior in both OOD and ID settings. We further compare the scaling properties across different architectures, incorporating two state-of-the-art TSFMs as case studies, showing that model architecture plays a significant role in scaling. The encoder-only Transformers demonstrate better scalability than the decoder-only Transformers, while the architectural enhancements in the two advanced TSFMs primarily improve ID performance but reduce OOD scalability. While scaling up TSFMs is expected to drive performance breakthroughs, the lack of a comprehensive understanding of TSFM scaling laws has hindered the development of a robust framework to guide model scaling. We fill this gap in this work by synthesizing our findings and providing practical guidelines for designing and scaling larger TSFMs with enhanced model capabilities.
☆ Federated Temporal Graph Clustering
Temporal graph clustering is a complex task that involves discovering meaningful structures in dynamic graphs where relationships and entities change over time. Existing methods typically require centralized data collection, which poses significant privacy and communication challenges. In this work, we introduce a novel Federated Temporal Graph Clustering (FTGC) framework that enables decentralized training of graph neural networks (GNNs) across multiple clients, ensuring data privacy throughout the process. Our approach incorporates a temporal aggregation mechanism to effectively capture the evolution of graph structures over time and a federated optimization strategy to collaboratively learn high-quality clustering representations. By preserving data privacy and reducing communication overhead, our framework achieves competitive performance on temporal graph datasets, making it a promising solution for privacy-sensitive, real-world applications involving dynamic data.
comment: 8 pages, 1 figure
☆ MAX: Masked Autoencoder for X-ray Fluorescence in Geological Investigation
Pre-training foundation models has become the de-facto procedure for deep learning approaches, yet its application remains limited in the geological studies, where in needs of the model transferability to break the shackle of data scarcity. Here we target on the X-ray fluorescence (XRF) scanning data, a standard high-resolution measurement in extensive scientific drilling projects. We propose a scalable self-supervised learner, masked autoencoders on XRF spectra (MAX), to pre-train a foundation model covering geological records from multiple regions of the Pacific and Southern Ocean. In pre-training, we find that masking a high proportion of the input spectrum (50\%) yields a nontrivial and meaningful self-supervisory task. For downstream tasks, we select the quantification of XRF spectra into two costly geochemical measurements, CaCO$_3$ and total organic carbon, due to their importance in understanding the paleo-oceanic carbon system. Our results show that MAX, requiring only one-third of the data, outperforms models without pre-training in terms of quantification accuracy. Additionally, the model's generalizability improves by more than 60\% in zero-shot tests on new materials, with explainability further ensuring its robustness. Thus, our approach offers a promising pathway to overcome data scarcity in geological discovery by leveraging the self-supervised foundation model and fast-acquired XRF scanning data.
☆ Improved Anomaly Detection through Conditional Latent Space VAE Ensembles
We propose a novel Conditional Latent space Variational Autoencoder (CL-VAE) to perform improved pre-processing for anomaly detection on data with known inlier classes and unknown outlier classes. This proposed variational autoencoder (VAE) improves latent space separation by conditioning on information within the data. The method fits a unique prior distribution to each class in the dataset, effectively expanding the classic prior distribution for VAEs to include a Gaussian mixture model. An ensemble of these VAEs are merged in the latent spaces to form a group consensus that greatly improves the accuracy of anomaly detection across data sets. Our approach is compared against the capabilities of a typical VAE, a CNN, and a PCA, with regards AUC for anomaly detection. The proposed model shows increased accuracy in anomaly detection, achieving an AUC of 97.4% on the MNIST dataset compared to 95.7% for the second best model. In addition, the CL-VAE shows increased benefits from ensembling, a more interpretable latent space, and an increased ability to learn patterns in complex data with limited model sizes.
comment: 13 pages of main article, 19 pages including references and appendix, 4 figures
☆ Revisited Large Language Model for Time Series Analysis through Modality Alignment
Large Language Models have demonstrated impressive performance in many pivotal web applications such as sensor data analysis. However, since LLMs are not designed for time series tasks, simpler models like linear regressions can often achieve comparable performance with far less complexity. In this study, we perform extensive experiments to assess the effectiveness of applying LLMs to key time series tasks, including forecasting, classification, imputation, and anomaly detection. We compare the performance of LLMs against simpler baseline models, such as single-layer linear models and randomly initialized LLMs. Our results reveal that LLMs offer minimal advantages for these core time series tasks and may even distort the temporal structure of the data. In contrast, simpler models consistently outperform LLMs while requiring far fewer parameters. Furthermore, we analyze existing reprogramming techniques and show, through data manifold analysis, that these methods fail to effectively align time series data with language and display pseudo-alignment behaviour in embedding space. Our findings suggest that the performance of LLM-based methods in time series tasks arises from the intrinsic characteristics and structure of time series data, rather than any meaningful alignment with the language model architecture.
☆ TPFL: A Trustworthy Personalized Federated Learning Framework via Subjective Logic
Federated learning (FL) enables collaborative model training across distributed clients while preserving data privacy. Despite its widespread adoption, most FL approaches focusing solely on privacy protection fall short in scenarios where trustworthiness is crucial, necessitating advancements in secure training, dependable decision-making mechanisms, robustness on corruptions, and enhanced performance with Non-IID data. To bridge this gap, we introduce Trustworthy Personalized Federated Learning (TPFL) framework designed for classification tasks via subjective logic in this paper. Specifically, TPFL adopts a unique approach by employing subjective logic to construct federated models, providing probabilistic decisions coupled with an assessment of uncertainty rather than mere probability assignments. By incorporating a trainable heterogeneity prior to the local training phase, TPFL effectively mitigates the adverse effects of data heterogeneity. Model uncertainty and instance uncertainty are further utilized to ensure the safety and reliability of the training and inference stages. Through extensive experiments on widely recognized federated learning benchmarks, we demonstrate that TPFL not only achieves competitive performance compared with advanced methods but also exhibits resilience against prevalent malicious attacks, robustness on domain shifts, and reliability in high-stake scenarios.
comment: 17 Pages with Appendix
☆ DAT: Improving Adversarial Robustness via Generative Amplitude Mix-up in Frequency Domain
To protect deep neural networks (DNNs) from adversarial attacks, adversarial training (AT) is developed by incorporating adversarial examples (AEs) into model training. Recent studies show that adversarial attacks disproportionately impact the patterns within the phase of the sample's frequency spectrum -- typically containing crucial semantic information -- more than those in the amplitude, resulting in the model's erroneous categorization of AEs. We find that, by mixing the amplitude of training samples' frequency spectrum with those of distractor images for AT, the model can be guided to focus on phase patterns unaffected by adversarial perturbations. As a result, the model's robustness can be improved. Unfortunately, it is still challenging to select appropriate distractor images, which should mix the amplitude without affecting the phase patterns. To this end, in this paper, we propose an optimized Adversarial Amplitude Generator (AAG) to achieve a better tradeoff between improving the model's robustness and retaining phase patterns. Based on this generator, together with an efficient AE production procedure, we design a new Dual Adversarial Training (DAT) strategy. Experiments on various datasets show that our proposed DAT leads to significantly improved robustness against diverse adversarial attacks.
☆ Continuous Pupillography: A Case for Visual Health Ecosystem
This article aims to cover pupillography, and its potential use in a number of ophthalmological diagnostic applications in biomedical space. With the ever-increasing incorporation of technology within our daily lives and an ever-growing active research into smart devices and technologies, we try to make a case for a health ecosystem that revolves around continuous eye monitoring. We tend to summarize the design constraints & requirements for an IoT-based continuous pupil detection system, with an attempt at developing a pipeline for wearable pupillographic device, while comparing two compact mini-camera modules currently available in the market. We use a light algorithm that can be directly adopted to current micro-controllers, and share our results for different lighting conditions, and scenarios. Lastly, we present our findings, along with an analysis on the challenges faced and a way ahead towards successfully building this ecosystem.
☆ Two Birds with One Stone: Multi-Task Semantic Communications Systems over Relay Channel
In this paper, we propose a novel multi-task, multi-link relay semantic communications (MTML-RSC) scheme that enables the destination node to simultaneously perform image reconstruction and classification with one transmission from the source node. In the MTML-RSC scheme, the source node broadcasts a signal using semantic communications, and the relay node forwards the signal to the destination. We analyze the coupling relationship between the two tasks and the two links (source-to-relay and source-to-destination) and design a semantic-focused forward method for the relay node, where it selectively forwards only the semantics of the relevant class while ignoring others. At the destination, the node combines signals from both the source node and the relay node to perform classification, and then uses the classification result to assist in decoding the signal from the relay node for image reconstructing. Experimental results demonstrate that the proposed MTML-RSC scheme achieves significant performance gains, e.g., $1.73$ dB improvement in peak-signal-to-noise ratio (PSNR) for image reconstruction and increasing the accuracy from $64.89\%$ to $70.31\%$ for classification.
comment: submitted to IEEE WCNC
☆ Conjunction Subspaces Test for Conformal and Selective Classification
In this paper, we present a new classifier, which integrates significance testing results over different random subspaces to yield consensus p-values for quantifying the uncertainty of classification decision. The null hypothesis is that the test sample has no association with the target class on a randomly chosen subspace, and hence the classification problem can be formulated as a problem of testing for the conjunction of hypotheses. The proposed classifier can be easily deployed for the purpose of conformal prediction and selective classification with reject and refine options by simply thresholding the consensus p-values. The theoretical analysis on the generalization error bound of the proposed classifier is provided and empirical studies on real data sets are conducted as well to demonstrate its effectiveness.
comment: 36 pages, 9 figures
☆ Consistency Calibration: Improving Uncertainty Calibration via Consistency among Perturbed Neighbors
Calibration is crucial in deep learning applications, especially in fields like healthcare and autonomous driving, where accurate confidence estimates are vital for decision-making. However, deep neural networks often suffer from miscalibration, with reliability diagrams and Expected Calibration Error (ECE) being the only standard perspective for evaluating calibration performance. In this paper, we introduce the concept of consistency as an alternative perspective on model calibration, inspired by uncertainty estimation literature in large language models (LLMs). We highlight its advantages over the traditional reliability-based view. Building on this concept, we propose a post-hoc calibration method called Consistency Calibration (CC), which adjusts confidence based on the model's consistency across perturbed inputs. CC is particularly effective in locally uncertainty estimation, as it requires no additional data samples or label information, instead generating input perturbations directly from the source data. Moreover, we show that performing perturbations at the logit level significantly improves computational efficiency. We validate the effectiveness of CC through extensive comparisons with various post-hoc and training-time calibration methods, demonstrating state-of-the-art performance on standard datasets such as CIFAR-10, CIFAR-100, and ImageNet, as well as on long-tailed datasets like ImageNet-LT.
☆ Towards LLM-based Cognitive Models of Students with Misconceptions
Accurately modeling student cognition is crucial for developing effective AI-driven educational technologies. A key challenge is creating realistic student models that satisfy two essential properties: (1) accurately replicating specific misconceptions, and (2) correctly solving problems where these misconceptions are not applicable. This dual requirement reflects the complex nature of student understanding, where misconceptions coexist with correct knowledge. This paper investigates whether Large Language Models (LLMs) can be instruction-tuned to meet this dual requirement and effectively simulate student thinking in algebra. We introduce MalAlgoPy, a novel Python library that generates datasets reflecting authentic student solution patterns through a graph-based representation of algebraic problem-solving. Utilizing MalAlgoPy, we define and examine Cognitive Student Models (CSMs) - LLMs instruction tuned to faithfully emulate realistic student behavior. Our findings reveal that LLMs trained on misconception examples can efficiently learn to replicate errors. However, the training diminishes the model's ability to solve problems correctly, particularly for problem types where the misconceptions are not applicable, thus failing to satisfy second property of CSMs. We demonstrate that by carefully calibrating the ratio of correct to misconception examples in the training data - sometimes as low as 0.25 - it is possible to develop CSMs that satisfy both properties. Our insights enhance our understanding of AI-based student models and pave the way for effective adaptive learning systems.
☆ Discovering Leitmotifs in Multidimensional Time Series
A leitmotif is a recurring theme in literature, movies or music that carries symbolic significance for the piece it is contained in. When this piece can be represented as a multi-dimensional time series (MDTS), such as acoustic or visual observations, finding a leitmotif is equivalent to the pattern discovery problem, which is an unsupervised and complex problem in time series analytics. Compared to the univariate case, it carries additional complexity because patterns typically do not occur in all dimensions but only in a few - which are, however, unknown and must be detected by the method itself. In this paper, we present the novel, efficient and highly effective leitmotif discovery algorithm LAMA for MDTS. LAMA rests on two core principals: (a) a leitmotif manifests solely given a yet unknown number of sub-dimensions - neither too few, nor too many, and (b) the set of sub-dimensions are not independent from the best pattern found therein, necessitating both problems to be approached in a joint manner. In contrast to most previous methods, LAMA tackles both problems jointly - instead of independently selecting dimensions (or leitmotifs) and finding the best leitmotifs (or dimensions). Our experimental evaluation on a novel ground-truth annotated benchmark of 14 distinct real-life data sets shows that LAMA, when compared to four state-of-the-art baselines, shows superior performance in detecting meaningful patterns without increased computational complexity.
☆ AI-Aided Kalman Filters
The Kalman filter (KF) and its variants are among the most celebrated algorithms in signal processing. These methods are used for state estimation of dynamic systems by relying on mathematical representations in the form of simple state-space (SS) models, which may be crude and inaccurate descriptions of the underlying dynamics. Emerging data-centric artificial intelligence (AI) techniques tackle these tasks using deep neural networks (DNNs), which are model-agnostic. Recent developments illustrate the possibility of fusing DNNs with classic Kalman-type filtering, obtaining systems that learn to track in partially known dynamics. This article provides a tutorial-style overview of design approaches for incorporating AI in aiding KF-type algorithms. We review both generic and dedicated DNN architectures suitable for state estimation, and provide a systematic presentation of techniques for fusing AI tools with KFs and for leveraging partial SS modeling and data, categorizing design approaches into task-oriented and SS model-oriented. The usefulness of each approach in preserving the individual strengths of model-based KFs and data-driven DNNs is investigated in a qualitative and quantitative study, whose code is publicly available, illustrating the gains of hybrid model-based/data-driven designs. We also discuss existing challenges and future research directions that arise from fusing AI and Kalman-type algorithms.
comment: Submitted to IEEE Signal Processing Magazine
☆ A Numerical Study of Chaotic Dynamics of K-S Equation with FNOs
Solving non-linear partial differential equations which exhibit chaotic dynamics is an important problem with a wide-range of applications such as predicting weather extremes and financial market risk. Fourier neural operators (FNOs) have been shown to be efficient in solving partial differential equations (PDEs). In this work we demonstrate simulation of dynamics in the chaotic regime of the two-dimensional (2d) Kuramoto-Sivashinsky equation using FNOs. Particularly, we analyze the effect of Fourier mode cutoff on the results obtained by using FNOs vs those obtained using traditional PDE solvers. We compare the outputs using metrics such as the 2d power spectrum and the radial power spectrum. In addition we propose the normalised error power spectrum which measures the percentage error in the FNO model outputs. We conclude that FNOs capture the dynamics in the chaotic regime of the 2d K-S equation, provided the Fourier mode cutoff is kept sufficiently high.
comment: 8 pages, 5 figures. Submitted to CASML 2024
☆ Stress Assessment with Convolutional Neural Network Using PPG Signals
Stress is one of the main issues of nowadays lifestyle. If it becomes chronic it can have adverse effects on the human body. Thus, the early detection of stress is crucial to prevent its hurting effects on the human body and have a healthier life. Stress can be assessed using physiological signals. To this end, Photoplethysmography (PPG) is one of the most favorable physiological signals for stress assessment. This research is focused on developing a novel technique to assess stressful events using raw PPG signals recorded by Empatica E4 sensor. To achieve this goal, an adaptive convolutional neural network (CNN) combined with Multilayer Perceptron (MLP) has been utilized to realize the detection of stressful events. This research will use a dataset that is publicly available and named wearable stress and effect detection (WESAD). This dataset will be used to simulate the proposed model and to examine the advantages of the proposed developed model. The proposed model in this research will be able to distinguish between normal events and stressful events. This model will be able to detect stressful events with an accuracy of 96.7%.
comment: 5 figures, 2 tables
☆ iFuzzyTL: Interpretable Fuzzy Transfer Learning for SSVEP BCI System
The rapid evolution of Brain-Computer Interfaces (BCIs) has significantly influenced the domain of human-computer interaction, with Steady-State Visual Evoked Potentials (SSVEP) emerging as a notably robust paradigm. This study explores advanced classification techniques leveraging interpretable fuzzy transfer learning (iFuzzyTL) to enhance the adaptability and performance of SSVEP-based systems. Recent efforts have strengthened to reduce calibration requirements through innovative transfer learning approaches, which refine cross-subject generalizability and minimize calibration through strategic application of domain adaptation and few-shot learning strategies. Pioneering developments in deep learning also offer promising enhancements, facilitating robust domain adaptation and significantly improving system responsiveness and accuracy in SSVEP classification. However, these methods often require complex tuning and extensive data, limiting immediate applicability. iFuzzyTL introduces an adaptive framework that combines fuzzy logic principles with neural network architectures, focusing on efficient knowledge transfer and domain adaptation. iFuzzyTL refines input signal processing and classification in a human-interpretable format by integrating fuzzy inference systems and attention mechanisms. This approach bolsters the model's precision and aligns with real-world operational demands by effectively managing the inherent variability and uncertainty of EEG data. The model's efficacy is demonstrated across three datasets: 12JFPM (89.70% accuracy for 1s with an information transfer rate (ITR) of 149.58), Benchmark (85.81% accuracy for 1s with an ITR of 213.99), and eldBETA (76.50% accuracy for 1s with an ITR of 94.63), achieving state-of-the-art results and setting new benchmarks for SSVEP BCI performance.
☆ Game Theory Meets Statistical Mechanics in Deep Learning Design
We present a novel deep graphical representation that seamlessly merges principles of game theory with laws of statistical mechanics. It performs feature extraction, dimensionality reduction, and pattern classification within a single learning framework. Our approach draws an analogy between neurons in a network and players in a game theory model. Furthermore, each neuron viewed as a classical particle (subject to statistical physics' laws) is mapped to a set of actions representing specific activation value, and neural network layers are conceptualized as games in a sequential cooperative game theory setting. The feed-forward process in deep learning is interpreted as a sequential game, where each game comprises a set of players. During training, neurons are iteratively evaluated and filtered based on their contributions to a payoff function, which is quantified using the Shapley value driven by an energy function. Each set of neurons that significantly contributes to the payoff function forms a strong coalition. These neurons are the only ones permitted to propagate the information forward to the next layers. We applied this methodology to the task of facial age estimation and gender classification. Experimental results demonstrate that our approach outperforms both multi-layer perceptron and convolutional neural network models in terms of efficiency and accuracy.
☆ CATCH: Channel-Aware multivariate Time Series Anomaly Detection via Frequency Patching
Anomaly detection in multivariate time series is challenging as heterogeneous subsequence anomalies may occur. Reconstruction-based methods, which focus on learning nomral patterns in the frequency domain to detect diverse abnormal subsequences, achieve promising resutls, while still falling short on capturing fine-grained frequency characteristics and channel correlations. To contend with the limitations, we introduce CATCH, a framework based on frequency patching. We propose to patchify the frequency domain into frequency bands, which enhances its ability to capture fine-grained frequency characteristics. To perceive appropriate channel correlations, we propose a Channel Fusion Module (CFM), which features a patch-wise mask generator and a masked-attention mechanism. Driven by a bi-level multi-objective optimization algorithm, the CFM is encouraged to iteratively discover appropriate patch-wise channel correlations, and to cluster relevant channels while isolating adverse effects from irrelevant channels. Extensive experiments on 9 real-world datasets and 12 synthetic datasets demonstrate that CATCH achieves state-of-the-art performance.
☆ Optimizing YOLOv5s Object Detection through Knowledge Distillation algorithm
This paper explores the application of knowledge distillation technology in target detection tasks, especially the impact of different distillation temperatures on the performance of student models. By using YOLOv5l as the teacher network and a smaller YOLOv5s as the student network, we found that with the increase of distillation temperature, the student's detection accuracy gradually improved, and finally achieved mAP50 and mAP50-95 indicators that were better than the original YOLOv5s model at a specific temperature. Experimental results show that appropriate knowledge distillation strategies can not only improve the accuracy of the model but also help improve the reliability and stability of the model in practical applications. This paper also records in detail the accuracy curve and loss function descent curve during the model training process and shows that the model converges to a stable state after 150 training cycles. These findings provide a theoretical basis and technical reference for further optimizing target detection algorithms.
☆ Understanding Expert Structures on Minimax Parameter Estimation in Contaminated Mixture of Experts
We conduct the convergence analysis of parameter estimation in the contaminated mixture of experts. This model is motivated from the prompt learning problem where ones utilize prompts, which can be formulated as experts, to fine-tune a large-scaled pre-trained model for learning downstream tasks. There are two fundamental challenges emerging from the analysis: (i) the proportion in the mixture of the pre-trained model and the prompt may converge to zero where the prompt vanishes during the training; (ii) the algebraic interaction among parameters of the pre-trained model and the prompt can occur via some partial differential equation and decelerate the prompt learning. In response, we introduce a distinguishability condition to control the previous parameter interaction. Additionally, we also consider various types of expert structures to understand their effects on the parameter estimation. In each scenario, we provide comprehensive convergence rates of parameter estimation along with the corresponding minimax lower bounds.
comment: Fanqi Yan, Huy Nguyen, Dung Le contributed equally to this work. 70 pages, 6 figures, 1 table
☆ Irregularity-Informed Time Series Analysis: Adaptive Modelling of Spatial and Temporal Dynamics
Irregular Time Series Data (IRTS) has shown increasing prevalence in real-world applications. We observed that IRTS can be divided into two specialized types: Natural Irregular Time Series (NIRTS) and Accidental Irregular Time Series (AIRTS). Various existing methods either ignore the impacts of irregular patterns or statically learn the irregular dynamics of NIRTS and AIRTS data and suffer from limited data availability due to the sparsity of IRTS. We proposed a novel transformer-based framework for general irregular time series data that treats IRTS from four views: Locality, Time, Spatio and Irregularity to motivate the data usage to the highest potential. Moreover, we design a sophisticated irregularity-gate mechanism to adaptively select task-relevant information from irregularity, which improves the generalization ability to various IRTS data. We implement extensive experiments to demonstrate the resistance of our work to three highly missing ratio datasets (88.4\%, 94.9\%, 60\% missing value) and investigate the significance of the irregularity information for both NIRTS and AIRTS by additional ablation study. We release our implementation in https://github.com/IcurasLW/MTSFormer-Irregular_Time_Series.git
☆ Dual Action Policy for Robust Sim-to-Real Reinforcement Learning
This paper presents Dual Action Policy (DAP), a novel approach to address the dynamics mismatch inherent in the sim-to-real gap of reinforcement learning. DAP uses a single policy to predict two sets of actions: one for maximizing task rewards in simulation and another specifically for domain adaptation via reward adjustments. This decoupling makes it easier to maximize the overall reward in the source domain during training. Additionally, DAP incorporates uncertainty-based exploration during training to enhance agent robustness. Experimental results demonstrate DAP's effectiveness in bridging the sim-to-real gap, outperforming baselines on challenging tasks in simulation, and further improvement is achieved by incorporating uncertainty estimation.
☆ Devil in the Tail: A Multi-Modal Framework for Drug-Drug Interaction Prediction in Long Tail Distinction
Drug-drug interaction (DDI) identification is a crucial aspect of pharmacology research. There are many DDI types (hundreds), and they are not evenly distributed with equal chance to occur. Some of the rarely occurred DDI types are often high risk and could be life-critical if overlooked, exemplifying the long-tailed distribution problem. Existing models falter against this distribution challenge and overlook the multi-faceted nature of drugs in DDI prediction. In this paper, a novel multi-modal deep learning-based framework, namely TFDM, is introduced to leverage multiple properties of a drug to achieve DDI classification. The proposed framework fuses multimodal features of drugs, including graph-based, molecular structure, Target and Enzyme, for DDI identification. To tackle the challenge posed by the distribution skewness across categories, a novel loss function called Tailed Focal Loss is introduced, aimed at further enhancing the model performance and address gradient vanishing problem of focal loss in extremely long-tailed dataset. Intensive experiments over 4 challenging long-tailed dataset demonstrate that the TFMD outperforms the most recent SOTA methods in long-tailed DDI classification tasks. The source code is released to reproduce our experiment results: https://github.com/IcurasLW/TFMD_Longtailed_DDI.git
☆ Transfer Learning on Multi-Dimensional Data: A Novel Approach to Neural Network-Based Surrogate Modeling
The development of efficient surrogates of partial differential equations (PDEs) is a critical step towards scalable modeling of complex, multiscale systems-of-systems. Convolutional neural networks (CNNs) have gained popularity as the basis for such surrogate models due to their success in capturing high-dimensional input-output mappings and the negligible cost of a forward pass. However, the high cost of generating training data -- typically via classical numerical solvers -- raises the question of whether these models are worth pursuing over more straightforward alternatives with well-established theoretical foundations, such as Monte Carlo methods. To reduce the cost of data generation, we propose training a CNN surrogate model on a mixture of numerical solutions to both the $d$-dimensional problem and its ($d-1$)-dimensional approximation, taking advantage of the efficiency savings guaranteed by the curse of dimensionality. We demonstrate our approach on a multiphase flow test problem, using transfer learning to train a dense fully-convolutional encoder-decoder CNN on the two classes of data. Numerical results from a sample uncertainty quantification task demonstrate that our surrogate model outperforms Monte Carlo with several times the data generation budget.
☆ Off-dynamics Conditional Diffusion Planners
Offline Reinforcement Learning (RL) offers an attractive alternative to interactive data acquisition by leveraging pre-existing datasets. However, its effectiveness hinges on the quantity and quality of the data samples. This work explores the use of more readily available, albeit off-dynamics datasets, to address the challenge of data scarcity in Offline RL. We propose a novel approach using conditional Diffusion Probabilistic Models (DPMs) to learn the joint distribution of the large-scale off-dynamics dataset and the limited target dataset. To enable the model to capture the underlying dynamics structure, we introduce two contexts for the conditional model: (1) a continuous dynamics score allows for partial overlap between trajectories from both datasets, providing the model with richer information; (2) an inverse-dynamics context guides the model to generate trajectories that adhere to the target environment's dynamic constraints. Empirical results demonstrate that our method significantly outperforms several strong baselines. Ablation studies further reveal the critical role of each dynamics context. Additionally, our model demonstrates that by modifying the context, we can interpolate between source and target dynamics, making it more robust to subtle shifts in the environment.
☆ Enhancing LLM Agents for Code Generation with Possibility and Pass-rate Prioritized Experience Replay
Nowadays transformer-based Large Language Models (LLM) for code generation tasks usually apply sampling and filtering pipelines. Due to the sparse reward problem in code generation tasks caused by one-token incorrectness, transformer-based models will sample redundant programs till they find a correct one, leading to low efficiency. To overcome the challenge, we incorporate Experience Replay (ER) in the fine-tuning phase, where codes and programs produced are stored and will be replayed to give the LLM agent a chance to learn from past experiences. Based on the spirit of ER, we introduce a novel approach called BTP pipeline which consists of three phases: beam search sampling, testing phase, and prioritized experience replay phase. The approach makes use of failed programs collected by code models and replays programs with high Possibility and Pass-rate Prioritized value (P2Value) from the replay buffer to improve efficiency. P2Value comprehensively considers the possibility of transformers' output and pass rate and can make use of the redundant resources caused by the problem that most programs collected by LLMs fail to pass any tests. We empirically apply our approach in several LLMs, demonstrating that it enhances their performance in code generation tasks and surpasses existing baselines.
☆ Improving the Generalization of Unseen Crowd Behaviors for Reinforcement Learning based Local Motion Planners
Deploying a safe mobile robot policy in scenarios with human pedestrians is challenging due to their unpredictable movements. Current Reinforcement Learning-based motion planners rely on a single policy to simulate pedestrian movements and could suffer from the over-fitting issue. Alternatively, framing the collision avoidance problem as a multi-agent framework, where agents generate dynamic movements while learning to reach their goals, can lead to conflicts with human pedestrians due to their homogeneity. To tackle this problem, we introduce an efficient method that enhances agent diversity within a single policy by maximizing an information-theoretic objective. This diversity enriches each agent's experiences, improving its adaptability to unseen crowd behaviors. In assessing an agent's robustness against unseen crowds, we propose diverse scenarios inspired by pedestrian crowd behaviors. Our behavior-conditioned policies outperform existing works in these challenging scenes, reducing potential collisions without additional time or travel.
☆ Causally-Aware Unsupervised Feature Selection Learning
Unsupervised feature selection (UFS) has recently gained attention for its effectiveness in processing unlabeled high-dimensional data. However, existing methods overlook the intrinsic causal mechanisms within the data, resulting in the selection of irrelevant features and poor interpretability. Additionally, previous graph-based methods fail to account for the differing impacts of non-causal and causal features in constructing the similarity graph, which leads to false links in the generated graph. To address these issues, a novel UFS method, called Causally-Aware UnSupErvised Feature Selection learning (CAUSE-FS), is proposed. CAUSE-FS introduces a novel causal regularizer that reweights samples to balance the confounding distribution of each treatment feature. This regularizer is subsequently integrated into a generalized unsupervised spectral regression model to mitigate spurious associations between features and clustering labels, thus achieving causal feature selection. Furthermore, CAUSE-FS employs causality-guided hierarchical clustering to partition features with varying causal contributions into multiple granularities. By integrating similarity graphs learned adaptively at different granularities, CAUSE-FS increases the importance of causal features when constructing the fused similarity graph to capture the reliable local structure of data. Extensive experimental results demonstrate the superiority of CAUSE-FS over state-of-the-art methods, with its interpretability further validated through feature visualization.
☆ Exploring the impact of virtual reality user engagement on tourist behavioral response integrated an environment concern of touristic travel perspective: A new hybrid machine learning approach
Due to the impact of the COVID-19 pandemic, new attractions ways are tended to be adapted by compelling sites to provide tours product and services, such as virtual reality (VR) to visitors. Based on a systematic human-computer interaction (HCI) user engagement and Narrative transportation theory, we develop and test a theoretical framework using a hybrid partial least squares structural equation model (PLS-SEM) and artificial neural network (ANN) machine learning approach that examines key user engagement drivers of visitors' imagery and in-person tour intentions (ITI) during COVID-19. Further, we proposed a novel and hybrid approach called Reflective and Formative PLS-SEM-ANN (FRPSA) with considering both reflective and second-order formative constructs in PLS-SEM giving scope to their different advantages in a complex model. According to a sample of visitors' responses, the results demonstrate that a) user engagement, including felt involvement, aesthetic appeal, perceived usability, focused attention, endurability, and novelty, all directly affect in-person tour intentions; b) environment concern of touristic travel (EC) positively moderates the relationships between user engagement and ITI; c) EC negatively moderates the relationships between imagery and ITI; d) imagery exerts the mediating effect between user engagement and ITI; e) the felt involvement and aesthetic appeal show both the linear significance impact and nonlinear importance. Finally, contributions to theories and practical implications are discussed accordingly.
☆ EdgeRL: Reinforcement Learning-driven Deep Learning Model Inference Optimization at Edge
Balancing mutually diverging performance metrics, such as, processing latency, outcome accuracy, and end device energy consumption is a challenging undertaking for deep learning model inference in ad-hoc edge environments. In this paper, we propose EdgeRL framework that seeks to strike such balance by using an Advantage Actor-Critic (A2C) Reinforcement Learning (RL) approach that can choose optimal run-time DNN inference parameters and aligns the performance metrics based on the application requirements. Using real world deep learning model and a hardware testbed, we evaluate the benefits of EdgeRL framework in terms of end device energy savings, inference accuracy improvement, and end-to-end inference latency reduction.
☆ Global Censored Quantile Random Forest
In recent years, censored quantile regression has enjoyed an increasing popularity for survival analysis while many existing works rely on linearity assumptions. In this work, we propose a Global Censored Quantile Random Forest (GCQRF) for predicting a conditional quantile process on data subject to right censoring, a forest-based flexible, competitive method able to capture complex nonlinear relationships. Taking into account the randomness in trees and connecting the proposed method to a randomized incomplete infinite degree U-process (IDUP), we quantify the prediction process' variation without assuming an infinite forest and establish its weak convergence. Moreover, feature importance ranking measures based on out-of-sample predictive accuracy are proposed. We demonstrate the superior predictive accuracy of the proposed method over a number of existing alternatives and illustrate the use of the proposed importance ranking measures on both simulated and real data.
☆ Divide-Verify-Refine: Aligning LLM Responses with Complex Instructions
Recent studies show that LLMs, particularly open-source models, struggle to follow complex instructions with multiple constraints. Despite the importance, methods to improve LLMs' adherence to such constraints remain unexplored, and current research focuses on evaluating this ability rather than developing solutions. While a few studies enhance constraint adherence through model tuning, this approach is computationally expensive and heavily reliant on training data quality. An alternative is to leverage LLMs' self-correction capabilities, allowing them to adjust responses to better meet specified constraints. However, this self-correction ability of LLMs is limited by the feedback quality, as LLMs cannot autonomously generate reliable feedback or detect errors. Moreover, the self-refinement process heavily depends on few-shot examples that illustrate how to modify responses to meet constraints. As constraints in complex instructions are diverse and vary widely, manually crafting few-shot examples for each constraint type can be labor-intensive and sub-optimal. To deal with these two challenges, we propose the Divide-Verify-Refine (DVR) framework with three steps: (1) Divide complex instructions into single constraints and prepare appropriate tools; (2) Verify: To address the feedback quality problem, these tools will rigorously verify responses and provide reliable feedback; (3) Refine: To address the constraint diversity challenge, we design a refinement repository that collects successful refinement processes and uses them as few-shot demonstrations for future cases, allowing LLMs to learn from the past experience during inference. Additionally, we develop a new dataset of complex instructions, each containing 1-6 constraints. Experiments show that the framework significantly improves performance, doubling LLama3.1-8B's constraint adherence on instructions with 6 constraints.
comment: Under review
☆ Abnormality Forecasting: Time Series Anomaly Prediction via Future Context Modeling KDD
Identifying anomalies from time series data plays an important role in various fields such as infrastructure security, intelligent operation and maintenance, and space exploration. Current research focuses on detecting the anomalies after they occur, which can lead to significant financial/reputation loss or infrastructure damage. In this work we instead study a more practical yet very challenging problem, time series anomaly prediction, aiming at providing early warnings for abnormal events before their occurrence. To tackle this problem, we introduce a novel principled approach, namely future context modeling (FCM). Its key insight is that the future abnormal events in a target window can be accurately predicted if their preceding observation window exhibits any subtle difference to normal data. To effectively capture such differences, FCM first leverages long-term forecasting models to generate a discriminative future context based on the observation data, aiming to amplify those subtle but unusual difference. It then models a normality correlation of the observation data with the forecasting future context to complement the normality modeling of the observation data in foreseeing possible abnormality in the target window. A joint variate-time attention learning is also introduced in FCM to leverage both temporal signals and features of the time series data for more discriminative normality modeling in the aforementioned two views. Comprehensive experiments on five datasets demonstrate that FCM gains good recall rate (70\%+) on multiple datasets and significantly outperforms all baselines in F1 score. Code is available at https://github.com/mala-lab/FCM.
comment: 11 pages, 5 figures, submitted to KDD conference
☆ SAT: Data-light Uncertainty Set Merging via Synthetics, Aggregation, and Test Inversion
The integration of uncertainty sets has diverse applications but also presents challenges, particularly when only initial sets and their control levels are available, along with potential dependencies. Examples include merging confidence sets from different distributed sites with communication constraints, as well as combining conformal prediction sets generated by different learning algorithms or data splits. In this article, we introduce an efficient and flexible Synthetic, Aggregation, and Test inversion (SAT) approach to merge various potentially dependent uncertainty sets into a single set. The proposed method constructs a novel class of synthetic test statistics, aggregates them, and then derives merged sets through test inversion. Our approach leverages the duality between set estimation and hypothesis testing, ensuring reliable coverage in dependent scenarios. The procedure is data-light, meaning it relies solely on initial sets and control levels without requiring raw data, and it adapts to any user-specified initial uncertainty sets, accommodating potentially varying coverage levels. Theoretical analyses and numerical experiments confirm that SAT provides finite-sample coverage guarantees and achieves small set sizes.
☆ Potential-Based Intrinsic Motivation: Preserving Optimality With Complex, Non-Markovian Shaping Rewards
Recently there has been a proliferation of intrinsic motivation (IM) reward-shaping methods to learn in complex and sparse-reward environments. These methods can often inadvertently change the set of optimal policies in an environment, leading to suboptimal behavior. Previous work on mitigating the risks of reward shaping, particularly through potential-based reward shaping (PBRS), has not been applicable to many IM methods, as they are often complex, trainable functions themselves, and therefore dependent on a wider set of variables than the traditional reward functions that PBRS was developed for. We present an extension to PBRS that we prove preserves the set of optimal policies under a more general set of functions than has been previously proven. We also present {\em Potential-Based Intrinsic Motivation} (PBIM) and {\em Generalized Reward Matching} (GRM), methods for converting IM rewards into a potential-based form that are useable without altering the set of optimal policies. Testing in the MiniGrid DoorKey and Cliff Walking environments, we demonstrate that PBIM and GRM successfully prevent the agent from converging to a suboptimal policy and can speed up training. Additionally, we prove that GRM is sufficiently general as to encompass all potential-based reward shaping functions. This paper expands on previous work introducing the PBIM method, and provides an extension to the more general method of GRM, as well as additional proofs, experimental results, and discussion.
comment: To be submit to joint AIJ-JAIR special track for award-winning papers. arXiv admin note: substantial text overlap with arXiv:2402.07411
☆ LPUF-AuthNet: A Lightweight PUF-Based IoT Authentication via Tandem Neural Networks and Split Learning
By 2025, the internet of things (IoT) is projected to connect over 75 billion devices globally, fundamentally altering how we interact with our environments in both urban and rural settings. However, IoT device security remains challenging, particularly in the authentication process. Traditional cryptographic methods often struggle with the constraints of IoT devices, such as limited computational power and storage. This paper considers physical unclonable functions (PUFs) as robust security solutions, utilizing their inherent physical uniqueness to authenticate devices securely. However, traditional PUF systems are vulnerable to machine learning (ML) attacks and burdened by large datasets. Our proposed solution introduces a lightweight PUF mechanism, called LPUF-AuthNet, combining tandem neural networks (TNN) with a split learning (SL) paradigm. The proposed approach provides scalability, supports mutual authentication, and enhances security by resisting various types of attacks, paving the way for secure integration into future 6G technologies.
comment: Accepted to Proc. IEEE Globecom 2024
☆ DAQ: Density-Aware Post-Training Weight-Only Quantization For LLMs
Large language models (LLMs) excel in various tasks but face deployment challenges due to hardware constraints. We propose density-aware post-training weight-only quantization (DAQ), which has two stages: 1) density-centric alignment, which identifies the center of high-density weights and centers the dynamic range on this point to align high-density weight regions with floating-point high-precision regions; 2) learnable dynamic range adjustment, which adjusts the dynamic range by optimizing quantization parameters (i.e., scale and zero-point) based on the impact of weights on the model output. Experiments on LLaMA and LLaMA-2 show that DAQ consistently outperforms the best baseline method, reducing perplexity loss by an average of 22.8% on LLaMA and 19.6% on LLaMA-2. Our code is available at https://anonymous.4open.science/r/DAQ-E747.
comment: 9 pages, 4 figures
☆ ExoTST: Exogenous-Aware Temporal Sequence Transformer for Time Series Prediction ICDM 2024
Accurate long-term predictions are the foundations for many machine learning applications and decision-making processes. Traditional time series approaches for prediction often focus on either autoregressive modeling, which relies solely on past observations of the target ``endogenous variables'', or forward modeling, which considers only current covariate drivers ``exogenous variables''. However, effectively integrating past endogenous and past exogenous with current exogenous variables remains a significant challenge. In this paper, we propose ExoTST, a novel transformer-based framework that effectively incorporates current exogenous variables alongside past context for improved time series prediction. To integrate exogenous information efficiently, ExoTST leverages the strengths of attention mechanisms and introduces a novel cross-temporal modality fusion module. This module enables the model to jointly learn from both past and current exogenous series, treating them as distinct modalities. By considering these series separately, ExoTST provides robustness and flexibility in handling data uncertainties that arise from the inherent distribution shift between historical and current exogenous variables. Extensive experiments on real-world carbon flux datasets and time series benchmarks demonstrate ExoTST's superior performance compared to state-of-the-art baselines, with improvements of up to 10\% in prediction accuracy. Moreover, ExoTST exhibits strong robustness against missing values and noise in exogenous drivers, maintaining consistent performance in real-world situations where these imperfections are common.
comment: Accepted at ICDM 2024
☆ Model Balancing Helps Low-data Training and Fine-tuning EMNLP 2024
Recent advances in foundation models have emphasized the need to align pre-trained models with specialized domains using small, curated datasets. Studies on these foundation models underscore the importance of low-data training and fine-tuning. This topic, well-known in natural language processing (NLP), has also gained increasing attention in the emerging field of scientific machine learning (SciML). To address the limitations of low-data training and fine-tuning, we draw inspiration from Heavy-Tailed Self-Regularization (HT-SR) theory, analyzing the shape of empirical spectral densities (ESDs) and revealing an imbalance in training quality across different model layers. To mitigate this issue, we adapt a recently proposed layer-wise learning rate scheduler, TempBalance, which effectively balances training quality across layers and enhances low-data training and fine-tuning for both NLP and SciML tasks. Notably, TempBalance demonstrates increasing performance gains as the amount of available tuning data decreases. Comparative analyses further highlight the effectiveness of TempBalance and its adaptability as an "add-on" method for improving model performance.
comment: EMNLP 2024 Oral. First two authors contributed equally
☆ Expected Sliced Transport Plans
The optimal transport (OT) problem has gained significant traction in modern machine learning for its ability to: (1) provide versatile metrics, such as Wasserstein distances and their variants, and (2) determine optimal couplings between probability measures. To reduce the computational complexity of OT solvers, methods like entropic regularization and sliced optimal transport have been proposed. The sliced OT framework improves efficiency by comparing one-dimensional projections (slices) of high-dimensional distributions. However, despite their computational efficiency, sliced-Wasserstein approaches lack a transportation plan between the input measures, limiting their use in scenarios requiring explicit coupling. In this paper, we address two key questions: Can a transportation plan be constructed between two probability measures using the sliced transport framework? If so, can this plan be used to define a metric between the measures? We propose a "lifting" operation to extend one-dimensional optimal transport plans back to the original space of the measures. By computing the expectation of these lifted plans, we derive a new transportation plan, termed expected sliced transport (EST) plans. We prove that using the EST plan to weight the sum of the individual Euclidean costs for moving from one point to another results in a valid metric between the input discrete probability measures. We demonstrate the connection between our approach and the recently proposed min-SWGG, along with illustrative numerical examples that support our theoretical findings.
☆ Reinforcement Learning with LTL and $ω$-Regular Objectives via Optimality-Preserving Translation to Average Rewards
Linear temporal logic (LTL) and, more generally, $\omega$-regular objectives are alternatives to the traditional discount sum and average reward objectives in reinforcement learning (RL), offering the advantage of greater comprehensibility and hence explainability. In this work, we study the relationship between these objectives. Our main result is that each RL problem for $\omega$-regular objectives can be reduced to a limit-average reward problem in an optimality-preserving fashion, via (finite-memory) reward machines. Furthermore, we demonstrate the efficacy of this approach by showing that optimal policies for limit-average problems can be found asymptotically by solving a sequence of discount-sum problems approximately. Consequently, we resolve an open problem: optimal policies for LTL and $\omega$-regular objectives can be learned asymptotically.
♻ ☆ Towards Scalable Exact Machine Unlearning Using Parameter-Efficient Fine-Tuning NeurIPS
Machine unlearning is the process of efficiently removing the influence of a training data instance from a trained machine learning model without retraining it from scratch. A popular subclass of unlearning approaches is exact machine unlearning, which focuses on techniques that explicitly guarantee the removal of the influence of a data instance from a model. Exact unlearning approaches use a machine learning model in which individual components are trained on disjoint subsets of the data. During deletion, exact unlearning approaches only retrain the affected components rather than the entire model. While existing approaches reduce retraining costs, it can still be expensive for an organization to retrain a model component as it requires halting a system in production, which leads to service failure and adversely impacts customers. To address these challenges, we introduce an exact unlearning framework -- Sequence-aware Sharded Sliced Training (S3T), which is designed to enhance the deletion capabilities of an exact unlearning system while minimizing the impact on model's performance. At the core of S3T, we utilize a lightweight parameter-efficient fine-tuning approach that enables parameter isolation by sequentially training layers with disjoint data slices. This enables efficient unlearning by simply deactivating the layers affected by data deletion. Furthermore, to reduce the retraining cost and improve model performance, we train the model on multiple data sequences, which allows S3T to handle an increased number of deletion requests. Both theoretically and empirically, we demonstrate that S3T attains superior deletion capabilities and enhanced performance compared to baselines across a wide range of settings.
comment: Preliminary version accepted at the SafeGenAi Workshop, NeurIPS, 2024
♻ ☆ Neural Algorithmic Reasoning with Multiple Correct Solutions
Neural Algorithmic Reasoning (NAR) aims to optimize classical algorithms. However, canonical implementations of NAR train neural networks to return only a single solution, even when there are multiple correct solutions to a problem, such as single-source shortest paths. For some applications, it is desirable to recover more than one correct solution. To that end, we give the first method for NAR with multiple solutions. We demonstrate our method on two classical algorithms: Bellman-Ford (BF) and Depth-First Search (DFS), favouring deeper insight into two algorithms over a broader survey of algorithms. This method involves generating appropriate training data as well as sampling and validating solutions from model output. Each step of our method, which can serve as a framework for neural algorithmic reasoning beyond the tasks presented in this paper, might be of independent interest to the field and our results represent the first attempt at this task in the NAR literature.
♻ ☆ cedar: Optimized and Unified Machine Learning Input Data Pipelines VLDB
The input data pipeline is an essential component of each machine learning (ML) training job. It is responsible for reading massive amounts of training data, processing batches of samples using complex transformations, and loading them onto training nodes at low latency and high throughput. Performant input data systems are becoming increasingly critical, driven by skyrocketing data volumes and training throughput demands. Unfortunately, current input data systems cannot fully leverage key performance optimizations, resulting in hugely inefficient infrastructures that require significant resources - or worse - underutilize expensive accelerators. To address these demands, we present cedar, an optimized and unified programming framework for ML input data pipelines. cedar allows users to define input data pipelines using composable operators that support arbitrary ML frameworks and libraries. cedar introduces an extensible optimizer that systematically applies a complex combination of optimizations (e.g., offloading, caching, prefetching, fusion, and reordering). It orchestrates processing across a customizable set of local and distributed compute resources in order to improve processing performance and efficiency, all without user input. Across eight pipelines, cedar improves performance by up to 1.87x to 10.65x compared to state-of-the-art input data systems.
comment: Accepted to PVLDB Volume 18
♻ ☆ Adversarial Training of Two-Layer Polynomial and ReLU Activation Networks via Convex Optimization
Training neural networks which are robust to adversarial attacks remains an important problem in deep learning, especially as heavily overparameterized models are adopted in safety-critical settings. Drawing from recent work which reformulates the training problems for two-layer ReLU and polynomial activation networks as convex programs, we devise a convex semidefinite program (SDP) for adversarial training of two-layer polynomial activation networks and prove that the convex SDP achieves the same globally optimal solution as its nonconvex counterpart. The convex SDP is observed to improve robust test accuracy against $\ell_\infty$ attacks relative to the original convex training formulation on multiple datasets. Additionally, we present scalable implementations of adversarial training for two-layer polynomial and ReLU networks which are compatible with standard machine learning libraries and GPU acceleration. Leveraging these implementations, we retrain the final two fully connected layers of a Pre-Activation ResNet-18 model on the CIFAR-10 dataset with both polynomial and ReLU activations. The two `robustified' models achieve significantly higher robust test accuracies against $\ell_\infty$ attacks than a Pre-Activation ResNet-18 model trained with sharpness-aware minimization, demonstrating the practical utility of convex adversarial training on large-scale problems.
comment: 17 pages, 2 figures. Added a proof of the main theorem in the appendix. Expanded numerical results section. Added references
♻ ☆ An exactly solvable model for emergence and scaling laws in the multitask sparse parity problem NeurIPS 2024
Deep learning models can exhibit what appears to be a sudden ability to solve a new problem as training time, training data, or model size increases, a phenomenon known as emergence. In this paper, we present a framework where each new ability (a skill) is represented as a basis function. We solve a simple multi-linear model in this skill-basis, finding analytic expressions for the emergence of new skills, as well as for scaling laws of the loss with training time, data size, model size, and optimal compute. We compare our detailed calculations to direct simulations of a two-layer neural network trained on multitask sparse parity, where the tasks in the dataset are distributed according to a power-law. Our simple model captures, using a single fit parameter, the sigmoidal emergence of multiple new skills as training time, data size or model size increases in the neural network.
comment: Accepted at NeurIPS 2024 Conference
♻ ☆ Energy and Carbon Considerations of Fine-Tuning BERT EMNLP 2023
Despite the popularity of the `pre-train then fine-tune' paradigm in the NLP community, existing work quantifying energy costs and associated carbon emissions has largely focused on language model pre-training. Although a single pre-training run draws substantially more energy than fine-tuning, fine-tuning is performed more frequently by many more individual actors, and thus must be accounted for when considering the energy and carbon footprint of NLP. In order to better characterize the role of fine-tuning in the landscape of energy and carbon emissions in NLP, we perform a careful empirical study of the computational costs of fine-tuning across tasks, datasets, hardware infrastructure and measurement modalities. Our experimental results allow us to place fine-tuning energy and carbon costs into perspective with respect to pre-training and inference, and outline recommendations to NLP researchers and practitioners who wish to improve their fine-tuning energy efficiency.
comment: EMNLP 2023 Findings; First two authors contributed equally; 12 pages
♻ ☆ Preferential Normalizing Flows NeurIPS2024
Eliciting a high-dimensional probability distribution from an expert via noisy judgments is notoriously challenging, yet useful for many applications, such as prior elicitation and reward modeling. We introduce a method for eliciting the expert's belief density as a normalizing flow based solely on preferential questions such as comparing or ranking alternatives. This allows eliciting in principle arbitrarily flexible densities, but flow estimation is susceptible to the challenge of collapsing or diverging probability mass that makes it difficult in practice. We tackle this problem by introducing a novel functional prior for the flow, motivated by a decision-theoretic argument, and show empirically that the belief density can be inferred as the function-space maximum a posteriori estimate. We demonstrate our method by eliciting multivariate belief densities of simulated experts, including the prior belief of a general-purpose large language model over a real-world dataset.
comment: 29 pages, 18 figures, Accepted at NeurIPS2024
♻ ☆ On the Effective Horizon of Inverse Reinforcement Learning
Inverse reinforcement learning (IRL) algorithms often rely on (forward) reinforcement learning or planning over a given time horizon to compute an approximately optimal policy for a hypothesized reward function and then match this policy with expert demonstrations. The time horizon plays a critical role in determining both the accuracy of reward estimates and the computational efficiency of IRL algorithms. Interestingly, an \emph{effective time horizon} shorter than the ground-truth value often produces better results faster. This work formally analyzes this phenomenon and provides an explanation: the time horizon controls the complexity of an induced policy class and mitigates overfitting with limited data. This analysis serves as a guide for the principled choice of the effective horizon for IRL. It also prompts us to re-examine the classic IRL formulation: it is more natural to learn jointly the reward and the effective horizon rather than the reward alone with a given horizon. To validate our findings, we implement a cross-validation extension and the experimental results confirm the theoretical analysis.
comment: 9 pages, under review
♻ ☆ Deep Optimal Experimental Design for Parameter Estimation Problems
Optimal experimental design is a well studied field in applied science and engineering. Techniques for estimating such a design are commonly used within the framework of parameter estimation. Nonetheless, in recent years parameter estimation techniques are changing rapidly with the introduction of deep learning techniques to replace traditional estimation methods. This in turn requires the adaptation of optimal experimental design that is associated with these new techniques. In this paper we investigate a new experimental design methodology that uses deep learning. We show that the training of a network as a Likelihood Free Estimator can be used to significantly simplify the design process and circumvent the need for the computationally expensive bi-level optimization problem that is inherent in optimal experimental design for non-linear systems. Furthermore, deep design improves the quality of the recovery process for parameter estimation problems. As proof of concept we apply our methodology to two different systems of Ordinary Differential Equations.
♻ ☆ Likelihood-based Differentiable Structure Learning NeurIPS 2024
Existing approaches to differentiable structure learning of directed acyclic graphs (DAGs) rely on strong identifiability assumptions in order to guarantee that global minimizers of the acyclicity-constrained optimization problem identifies the true DAG. Moreover, it has been observed empirically that the optimizer may exploit undesirable artifacts in the loss function. We explain and remedy these issues by studying the behavior of differentiable acyclicity-constrained programs under general likelihoods with multiple global minimizers. By carefully regularizing the likelihood, it is possible to identify the sparsest model in the Markov equivalence class, even in the absence of an identifiable parametrization. We first study the Gaussian case in detail, showing how proper regularization of the likelihood defines a score that identifies the sparsest model. Assuming faithfulness, it also recovers the Markov equivalence class. These results are then generalized to general models and likelihoods, where the same claims hold. These theoretical results are validated empirically, showing how this can be done using standard gradient-based optimizers, thus paving the way for differentiable structure learning under general models and losses.
comment: 38 pages, 14 figures, to appear at NeurIPS 2024
♻ ☆ SplitLLM: Collaborative Inference of LLMs for Model Placement and Throughput Optimization
Large language models (LLMs) have been a disruptive innovation in recent years, and they play a crucial role in our daily lives due to their ability to understand and generate human-like text. Their capabilities include natural language understanding, information retrieval and search, translation, chatbots, virtual assistance, and many more. However, it is well known that LLMs are massive in terms of the number of parameters. Additionally, the self-attention mechanism in the underlying architecture of LLMs, Transformers, has quadratic complexity in terms of both computation and memory with respect to the input sequence length. For these reasons, LLM inference is resource-intensive, and thus, the throughput of LLM inference is limited, especially for the longer sequences. In this report, we design a collaborative inference architecture between a server and its clients to alleviate the throughput limit. In this design, we consider the available resources on both sides, i.e., the computation and communication costs. We develop a dynamic programming-based algorithm to optimally allocate computation between the server and the client device to increase the server throughput, while not violating the service level agreement (SLA). We show in the experiments that we are able to efficiently distribute the workload allowing for roughly 1/3 reduction in the server workload, while achieving 19 percent improvement over a greedy method. As a result, we are able to demonstrate that, in an environment with different types of LLM inference requests, the throughput of the server is improved.
♻ ☆ Diffusion Language Models Are Versatile Protein Learners ICML 2024
This paper introduces diffusion protein language model (DPLM), a versatile protein language model that demonstrates strong generative and predictive capabilities for protein sequences. We first pre-train scalable DPLMs from evolutionary-scale protein sequences within a generative self-supervised discrete diffusion probabilistic framework, which generalizes language modeling for proteins in a principled way. After pre-training, DPLM exhibits the ability to generate structurally plausible, novel, and diverse protein sequences for unconditional generation. We further demonstrate the proposed diffusion generative pre-training makes DPLM possess a better understanding of proteins, making it a superior representation learner, which can be fine-tuned for various predictive tasks, comparing favorably to ESM2 (Lin et al., 2022). Moreover, DPLM can be tailored for various needs, which showcases its prowess of conditional generation in several ways: (1) conditioning on partial peptide sequences, e.g., generating scaffolds for functional motifs with high success rate; (2) incorporating other modalities as conditioner, e.g., structure-conditioned generation for inverse folding; and (3) steering sequence generation towards desired properties, e.g., satisfying specified secondary structures, through a plug-and-play classifier guidance. Code is released at \url{https://github.com/bytedance/dplm}.
comment: ICML 2024 camera-ready version
♻ ☆ Extreme time extrapolation capabilities and thermodynamic consistency of physics-inspired Neural Networks for the 3D microstructure evolution of materials via Cahn-Hilliard flow
A Convolutional Recurrent Neural Network (CRNN) is trained to reproduce the evolution of the spinodal decomposition process in three dimensions as described by the Cahn-Hilliard equation. A specialized, physics-inspired architecture is proven to provide close accordance between the predicted evolutions and the ground truth ones obtained via conventional integration schemes. The method can accurately reproduce the evolution of microstructures not represented in the training set at a fraction of the computational costs. Extremely long-time extrapolation capabilities are achieved, up to reaching the theoretically expected equilibrium state of the system, consisting of a layered, phase-separated morphology, despite the training set containing only relatively-short, initial phases of the evolution. Quantitative accordance with the decay rate of the Free energy is also demonstrated up to the late coarsening stages, proving that this class of Machine Learning approaches can become a new and powerful tool for the long timescale and high throughput simulation of materials, while retaining thermodynamic consistency and high-accuracy.
comment: 12 pages, 6 main text figures, 2 appendix figures, 1 supplementary material figure
♻ ☆ Open-Source Conversational AI with SpeechBrain 1.0
SpeechBrain is an open-source Conversational AI toolkit based on PyTorch, focused particularly on speech processing tasks such as speech recognition, speech enhancement, speaker recognition, text-to-speech, and much more. It promotes transparency and replicability by releasing both the pre-trained models and the complete "recipes" of code and algorithms required for training them. This paper presents SpeechBrain 1.0, a significant milestone in the evolution of the toolkit, which now has over 200 recipes for speech, audio, and language processing tasks, and more than 100 models available on Hugging Face. SpeechBrain 1.0 introduces new technologies to support diverse learning modalities, Large Language Model (LLM) integration, and advanced decoding strategies, along with novel models, tasks, and modalities. It also includes a new benchmark repository, offering researchers a unified platform for evaluating models across diverse tasks.
comment: Accepted to the Journal of Machine Learning research (JMLR), Machine Learning Open Source Software
♻ ☆ Nearly Tight Black-Box Auditing of Differentially Private Machine Learning NeurIPS 2024
This paper presents an auditing procedure for the Differentially Private Stochastic Gradient Descent (DP-SGD) algorithm in the black-box threat model that is substantially tighter than prior work. The main intuition is to craft worst-case initial model parameters, as DP-SGD's privacy analysis is agnostic to the choice of the initial model parameters. For models trained on MNIST and CIFAR-10 at theoretical $\varepsilon=10.0$, our auditing procedure yields empirical estimates of $\varepsilon_{emp} = 7.21$ and $6.95$, respectively, on a 1,000-record sample and $\varepsilon_{emp}= 6.48$ and $4.96$ on the full datasets. By contrast, previous audits were only (relatively) tight in stronger white-box models, where the adversary can access the model's inner parameters and insert arbitrary gradients. Overall, our auditing procedure can offer valuable insight into how the privacy analysis of DP-SGD could be improved and detect bugs and DP violations in real-world implementations. The source code needed to reproduce our experiments is available at https://github.com/spalabucr/bb-audit-dpsgd.
comment: To appear in the Proceedings of the Thirty-eighth Annual Conference on Neural Information Processing Systems (NeurIPS 2024). Please cite accordingly
♻ ☆ Task Aware Modulation using Representation Learning: An Approach for Few Shot Learning in Environmental Systems
We introduce TAM-RL (Task Aware Modulation using Representation Learning), a novel multimodal meta-learning framework for few-shot learning in heterogeneous systems, designed for science and engineering problems where entities share a common underlying forward model but exhibit heterogeneity due to entity-specific characteristics. TAM-RL leverages an amortized training process with a modulation network and a base network to learn task-specific modulation parameters, enabling efficient adaptation to new tasks with limited data. We evaluate TAM-RL on two real-world environmental datasets: Gross Primary Product (GPP) prediction and streamflow forecasting, demonstrating significant improvements over existing meta-learning methods. On the FLUXNET dataset, TAM-RL improves RMSE by 18.9\% over MMAML with just one month of few-shot data, while for streamflow prediction, it achieves an 8.21\% improvement with one year of data. Synthetic data experiments further validate TAM-RL's superior performance in heterogeneous task distributions, outperforming the baselines in the most heterogeneous setting. Notably, TAM-RL offers substantial computational efficiency, with at least 3x faster training times compared to gradient-based meta-learning approaches while being much simpler to train due to reduced complexity. Ablation studies highlight the importance of pretraining and adaptation mechanisms in TAM-RL's performance.
♻ ☆ Uncovering, Explaining, and Mitigating the Superficial Safety of Backdoor Defense NeurIPS 2024
Backdoor attacks pose a significant threat to Deep Neural Networks (DNNs) as they allow attackers to manipulate model predictions with backdoor triggers. To address these security vulnerabilities, various backdoor purification methods have been proposed to purify compromised models. Typically, these purified models exhibit low Attack Success Rates (ASR), rendering them resistant to backdoored inputs. However, Does achieving a low ASR through current safety purification methods truly eliminate learned backdoor features from the pretraining phase? In this paper, we provide an affirmative answer to this question by thoroughly investigating the Post-Purification Robustness of current backdoor purification methods. We find that current safety purification methods are vulnerable to the rapid re-learning of backdoor behavior, even when further fine-tuning of purified models is performed using a very small number of poisoned samples. Based on this, we further propose the practical Query-based Reactivation Attack (QRA) which could effectively reactivate the backdoor by merely querying purified models. We find the failure to achieve satisfactory post-purification robustness stems from the insufficient deviation of purified models from the backdoored model along the backdoor-connected path. To improve the post-purification robustness, we propose a straightforward tuning defense, Path-Aware Minimization (PAM), which promotes deviation along backdoor-connected paths with extra model updates. Extensive experiments demonstrate that PAM significantly improves post-purification robustness while maintaining a good clean accuracy and low ASR. Our work provides a new perspective on understanding the effectiveness of backdoor safety tuning and highlights the importance of faithfully assessing the model's safety.
comment: NeurIPS 2024 Spotlight paper. The first two authors contributed equally
♻ ☆ Pessimistic Backward Policy for GFlowNets
This paper studies Generative Flow Networks (GFlowNets), which learn to sample objects proportionally to a given reward function through the trajectory of state transitions. In this work, we observe that GFlowNets tend to under-exploit the high-reward objects due to training on insufficient number of trajectories, which may lead to a large gap between the estimated flow and the (known) reward value. In response to this challenge, we propose a pessimistic backward policy for GFlowNets (PBP-GFN), which maximizes the observed flow to align closely with the true reward for the object. We extensively evaluate PBP-GFN across eight benchmarks, including hyper-grid environment, bag generation, structured set generation, molecular generation, and four RNA sequence generation tasks. In particular, PBP-GFN enhances the discovery of high-reward objects, maintains the diversity of the objects, and consistently outperforms existing methods.
♻ ☆ ITINERA: Integrating Spatial Optimization with Large Language Models for Open-domain Urban Itinerary Planning
Citywalk, a recently popular form of urban travel, requires genuine personalization and understanding of fine-grained requests compared to traditional itinerary planning. In this paper, we introduce the novel task of Open-domain Urban Itinerary Planning (OUIP), which generates personalized urban itineraries from user requests in natural language. We then present ITINERA, an OUIP system that integrates spatial optimization with large language models to provide customized urban itineraries based on user needs. This involves decomposing user requests, selecting candidate points of interest (POIs), ordering the POIs based on cluster-aware spatial optimization, and generating the itinerary. Experiments on real-world datasets and the performance of the deployed system demonstrate our system's capacity to deliver personalized and spatially coherent itineraries compared to current solutions. Source codes of ITINERA are available at https://github.com/YihongT/ITINERA.
♻ ☆ CELL your Model: Contrastive Explanations for Large Language Models
The advent of black-box deep neural network classification models has sparked the need to explain their decisions. However, in the case of generative AI, such as large language models (LLMs), there is no class prediction to explain. Rather, one can ask why an LLM output a particular response to a given prompt. In this paper, we answer this question by proposing, to the best of our knowledge, the first contrastive explanation methods requiring simply black-box/query access. Our explanations suggest that an LLM outputs a reply to a given prompt because if the prompt was slightly modified, the LLM would have given a different response that is either less preferable or contradicts the original response. The key insight is that contrastive explanations simply require a scoring function that has meaning to the user and not necessarily a specific real valued quantity (viz. class label). We offer two algorithms for finding contrastive explanations: i) A myopic algorithm, which although effective in creating contrasts, requires many model calls and ii) A budgeted algorithm, our main algorithmic contribution, which intelligently creates contrasts adhering to a query budget, necessary for longer contexts. We show the efficacy of these methods on diverse natural language tasks such as open-text generation, automated red teaming, and explaining conversational degradation.
♻ ☆ Light-Weight Fault Tolerant Attention for Large Language Model Training
Large Language Models (LLMs) have demonstrated remarkable performance in various natural language processing tasks. However, the training of these models is computationally intensive and susceptible to faults, particularly in the attention mechanism, which is a critical component of transformer-based LLMs. In this paper, we investigate the impact of faults on LLM training, focusing on INF, NaN, and near-INF values in the computation results with systematic fault injection experiments. We observe the propagation patterns of these errors, which can trigger non-trainable states in the model and disrupt training, forcing the procedure to load from checkpoints. To mitigate the impact of these faults, we propose ATTNChecker, the first Algorithm-Based Fault Tolerance (ABFT) technique tailored for the attention mechanism in LLMs. ATTNChecker is designed based on fault propagation patterns of LLM and incorporates performance optimization to adapt to both system reliability and model vulnerability while providing lightweight protection for fast LLM training. Evaluations on four LLMs show that ATTNChecker on average incurs on average 7% overhead on training while detecting and correcting all extreme errors. Compared with the state-of-the-art checkpoint/restore approach, ATTNChecker reduces recovery overhead by up to 49x.
♻ ☆ CECILIA: Comprehensive Secure Machine Learning Framework
Since ML algorithms have proven their success in many different applications, there is also a big interest in privacy preserving (PP) ML methods for building models on sensitive data. Moreover, the increase in the number of data sources and the high computational power required by those algorithms force individuals to outsource the training and/or the inference of a ML model to the clouds providing such services. To address this, we propose a secure 3-party computation framework, CECILIA, offering PP building blocks to enable complex operations privately. In addition to the adapted and common operations like addition and multiplication, it offers multiplexer, most significant bit and modulus conversion. The first two are novel in terms of methodology and the last one is novel in terms of both functionality and methodology. CECILIA also has two complex novel methods, which are the exact exponential of a public base raised to the power of a secret value and the inverse square root of a secret Gram matrix. We use CECILIA to realize the private inference on pre-trained RKNs, which require more complex operations than most other DNNs, on the structural classification of proteins as the first study ever accomplishing the PP inference on RKNs. In addition to the successful private computation of basic building blocks, the results demonstrate that we perform the exact and fully private exponential computation, which is done by approximation in the literature so far. Moreover, they also show that we compute the exact inverse square root of a secret Gram matrix up to a certain privacy level, which has not been addressed in the literature at all. We also analyze the scalability of CECILIA to various settings on a synthetic dataset. The framework shows a great promise to make other ML algorithms as well as further computations privately computable by the building blocks of the framework.
comment: Preprint version of "A privacy-preserving approach for cloud-based protein fold recognition" paper published in Patterns, ~8 pages of the main paper, ~5 pages of Supplement
♻ ☆ Towards aerodynamic surrogate modeling based on $β$-variational autoencoders
Surrogate models that combine dimensionality reduction and regression techniques are essential to reduce the need for costly high-fidelity computational fluid dynamics data. New approaches using $\beta$-Variational Autoencoder ($\beta$-VAE) architectures have shown promise in obtaining high-quality low-dimensional representations of high-dimensional flow data while enabling physical interpretation of their latent spaces. We propose a surrogate model based on latent space regression to predict pressure distributions on a transonic wing given the flight conditions: Mach number and angle of attack. The $\beta$-VAE model, enhanced with Principal Component Analysis (PCA), maps high-dimensional data to a low-dimensional latent space, showing a direct correlation with flight conditions. Regularization through $\beta$ requires careful tuning to improve overall performance, while PCA preprocessing helps to construct an effective latent space, improving autoencoder training and performance. Gaussian Process Regression is used to predict latent space variables from flight conditions, showing robust behavior independent of $\beta$, and the decoder reconstructs the high-dimensional pressure field data. This pipeline provides insight into unexplored flight conditions. Furthermore, a fine-tuning process of the decoder further refines the model, reducing the dependence on $\beta$ and enhancing accuracy. Structured latent space, robust regression performance, and significant improvements in fine-tuning collectively create a highly accurate and efficient surrogate model. Our methodology demonstrates the effectiveness of $\beta$-VAEs for aerodynamic surrogate modeling, offering a rapid, cost-effective, and reliable alternative for aerodynamic data prediction.
comment: 19 pages, 12 figures
♻ ☆ Reward-Robust RLHF in LLMs
As Large Language Models (LLMs) continue to progress toward more advanced forms of intelligence, Reinforcement Learning from Human Feedback (RLHF) is increasingly seen as a key pathway toward achieving Artificial General Intelligence (AGI). However, the reliance on reward-model-based (RM-based) alignment methods introduces significant challenges due to the inherent instability and imperfections of Reward Models (RMs), which can lead to critical issues such as reward hacking and misalignment with human intentions. In this paper, we introduce a reward-robust RLHF framework aimed at addressing these fundamental challenges, paving the way for more reliable and resilient learning in LLMs. Our approach introduces a novel optimization objective that carefully balances performance and robustness by incorporating Bayesian Reward Model Ensembles (BRME) to model the uncertainty set of reward functions. This allows the framework to integrate both nominal performance and minimum reward signals, ensuring more stable learning even with imperfect RMs. Empirical results demonstrate that our framework consistently outperforms baselines across diverse benchmarks, showing improved accuracy and long-term stability. We also provide a theoretical analysis, demonstrating that reward-robust RLHF approaches the stability of constant reward settings, which proves to be acceptable even in a stochastic-case analysis. Together, these contributions highlight the framework potential to enhance both the performance and stability of LLM alignment.
♻ ☆ Time-Varying Gaussian Process Bandits with Unknown Prior
Bayesian optimisation requires fitting a Gaussian process model, which in turn requires specifying prior on the unknown black-box function -- most of the theoretical literature assumes this prior is known. However, it is common to have more than one possible prior for a given black-box function, for example suggested by domain experts with differing opinions. In some cases, the type-II maximum likelihood estimator for selecting prior enjoys the consistency guarantee, but it does not universally apply to all types of priors. If the problem is stationary, one could rely on the Regret Balancing scheme to conduct the optimisation, but in the case of time-varying problems, such a scheme cannot be used. To address this gap in existing research, we propose a novel algorithm, PE-GP-UCB, which is capable of solving time-varying Bayesian optimisation problems even without the exact knowledge of the function's prior. The algorithm relies on the fact that either the observed function values are consistent with some of the priors, in which case it is easy to reject the wrong priors, or the observations are consistent with all candidate priors, in which case it does not matter which prior our model relies on. We provide a regret bound on the proposed algorithm. Finally, we empirically evaluate our algorithm on toy and real-world time-varying problems and show that it outperforms the maximum likelihood estimator, fully Bayesian treatment of unknown prior and Regret Balancing.
♻ ☆ Energy-Efficient Computation with DVFS using Deep Reinforcement Learning for Multi-Task Systems in Edge Computing
Periodic soft real-time systems have broad applications in many areas, such as IoT. Finding an optimal energy-efficient policy that is adaptable to underlying edge devices while meeting deadlines for tasks has always been challenging. This research studies generalized systems with multi-task, multi-deadline scenarios with reinforcement learning-based DVFS for energy saving. This work addresses the limitation of previous work that models a periodic system as a single task and single-deadline scenario, which is too simplified to cope with complex situations. The method encodes time series information in the Linux kernel into information that is easy to use for reinforcement learning, allowing the system to generate DVFS policies to adapt system patterns based on the general workload. For encoding, we present two different methods for comparison. Both methods use only one performance counter: system utilization and the kernel only needs minimal information from the userspace. Our method is implemented on Jetson Nano Board (2GB) and is tested with three fixed multitask workloads, which are three, five, and eight tasks in the workload, respectively. For randomness and generalization, we also designed a random workload generator to build different multitask workloads to test. Based on the test results, our method could save 3%-10% power compared to Linux built-in governors.
♻ ☆ ReadMe++: Benchmarking Multilingual Language Models for Multi-Domain Readability Assessment EMNLP 2024
We present a comprehensive evaluation of large language models for multilingual readability assessment. Existing evaluation resources lack domain and language diversity, limiting the ability for cross-domain and cross-lingual analyses. This paper introduces ReadMe++, a multilingual multi-domain dataset with human annotations of 9757 sentences in Arabic, English, French, Hindi, and Russian, collected from 112 different data sources. This benchmark will encourage research on developing robust multilingual readability assessment methods. Using ReadMe++, we benchmark multilingual and monolingual language models in the supervised, unsupervised, and few-shot prompting settings. The domain and language diversity in ReadMe++ enable us to test more effective few-shot prompting, and identify shortcomings in state-of-the-art unsupervised methods. Our experiments also reveal exciting results of superior domain generalization and enhanced cross-lingual transfer capabilities by models trained on ReadMe++. We will make our data publicly available and release a python package tool for multilingual sentence readability prediction using our trained models at: https://github.com/tareknaous/readme
comment: Accepted to EMNLP 2024 Main Conference
♻ ☆ Increasing Both Batch Size and Learning Rate Accelerates Stochastic Gradient Descent
The performance of mini-batch stochastic gradient descent (SGD) strongly depends on setting the batch size and learning rate to minimize the empirical loss in training the deep neural network. In this paper, we present theoretical analyses of mini-batch SGD with four schedulers: (i) constant batch size and decaying learning rate scheduler, (ii) increasing batch size and decaying learning rate scheduler, (iii) increasing batch size and increasing learning rate scheduler, and (iv) increasing batch size and warm-up decaying learning rate scheduler. We show that mini-batch SGD using scheduler (i) does not always minimize the expectation of the full gradient norm of the empirical loss, whereas it does using any of schedulers (ii), (iii), and (iv). Furthermore, schedulers (iii) and (iv) accelerate mini-batch SGD. The paper also provides numerical results of supporting analyses showing that using scheduler (iii) or (iv) minimizes the full gradient norm of the empirical loss faster than using scheduler (i) or (ii).
comment: 28 pages, 18 figures
♻ ☆ From Explainable to Interpretable Deep Learning for Natural Language Processing in Healthcare: How Far from Reality?
Deep learning (DL) has substantially enhanced natural language processing (NLP) in healthcare research. However, the increasing complexity of DL-based NLP necessitates transparent model interpretability, or at least explainability, for reliable decision-making. This work presents a thorough scoping review of explainable and interpretable DL in healthcare NLP. The term "eXplainable and Interpretable Artificial Intelligence" (XIAI) is introduced to distinguish XAI from IAI. Different models are further categorized based on their functionality (model-, input-, output-based) and scope (local, global). Our analysis shows that attention mechanisms are the most prevalent emerging IAI technique. The use of IAI is growing, distinguishing it from XAI. The major challenges identified are that most XIAI does not explore "global" modelling processes, the lack of best practices, and the lack of systematic evaluation and benchmarks. One important opportunity is to use attention mechanisms to enhance multi-modal XIAI for personalized medicine. Additionally, combining DL with causal logic holds promise. Our discussion encourages the integration of XIAI in Large Language Models (LLMs) and domain-specific smaller models. In conclusion, XIAI adoption in healthcare requires dedicated in-house expertise. Collaboration with domain experts, end-users, and policymakers can lead to ready-to-use XIAI methods across NLP and medical tasks. While challenges exist, XIAI techniques offer a valuable foundation for interpretable NLP algorithms in healthcare.
comment: This paper has been accepted by Computational and Structural Biotechnology Journal
♻ ☆ Five Years of COVID-19 Discourse on Instagram: A Labeled Instagram Dataset of Over Half a Million Posts for Multilingual Sentiment Analysis
The work presented in this paper makes three scientific contributions with a specific focus on mining and analysis of COVID-19-related posts on Instagram. First, it presents a multilingual dataset of 500,153 Instagram posts about COVID-19 published between January 2020 and September 2024. This dataset, available at https://dx.doi.org/10.21227/d46p-v480, contains Instagram posts in 161 different languages as well as 535,021 distinct hashtags. After the development of this dataset, multilingual sentiment analysis was performed, which involved classifying each post as positive, negative, or neutral. The results of sentiment analysis are presented as a separate attribute in this dataset. Second, it presents the results of performing sentiment analysis per year from 2020 to 2024. The findings revealed the trends in sentiment related to COVID-19 on Instagram since the beginning of the pandemic. For instance, between 2020 and 2024, the sentiment trends show a notable shift, with positive sentiment decreasing from 38.35% to 28.69%, while neutral sentiment rising from 44.19% to 58.34%. Finally, the paper also presents findings of language-specific sentiment analysis. This analysis highlighted similar and contrasting trends of sentiment across posts published in different languages on Instagram. For instance, out of all English posts, 49.68% were positive, 14.84% were negative, and 35.48% were neutral. In contrast, among Hindi posts, 4.40% were positive, 57.04% were negative, and 38.56% were neutral, reflecting distinct differences in the sentiment distribution between these two languages.
♻ ☆ Semantic Token Reweighting for Interpretable and Controllable Text Embeddings in CLIP EMNLP 2024
A text encoder within Vision-Language Models (VLMs) like CLIP plays a crucial role in translating textual input into an embedding space shared with images, thereby facilitating the interpretative analysis of vision tasks through natural language. Despite the varying significance of different textual elements within a sentence depending on the context, efforts to account for variation of importance in constructing text embeddings have been lacking. We propose a framework of Semantic Token Reweighting to build Interpretable text embeddings (SToRI), which incorporates controllability as well. SToRI refines the text encoding process in CLIP by differentially weighting semantic elements based on contextual importance, enabling finer control over emphasis responsive to data-driven insights and user preferences. The efficacy of SToRI is demonstrated through comprehensive experiments on few-shot image classification and image retrieval tailored to user preferences.
comment: Accepted at EMNLP 2024 Findings
♻ ☆ NAR-*ICP: Neural Execution of Classical ICP-based Pointcloud Registration Algorithms
This study explores the intersection of neural networks and classical robotics algorithms through the Neural Algorithmic Reasoning (NAR) framework, allowing to train neural networks to effectively reason like classical robotics algorithms by learning to execute them. Algorithms are integral to robotics and safety-critical applications due to their predictable and consistent performance through logical and mathematical principles. In contrast, while neural networks are highly adaptable, handling complex, high-dimensional data and generalising across tasks, they often lack interpretability and transparency in their internal computations. We propose a Graph Neural Network (GNN)-based learning framework, NAR-*ICP, which learns the intermediate algorithmic steps of classical ICP-based pointcloud registration algorithms, and extend the CLRS Algorithmic Reasoning Benchmark with classical robotics perception algorithms. We evaluate our approach across diverse datasets, from real-world to synthetic, demonstrating its flexibility in handling complex and noisy inputs, along with its potential to be used as part of a larger learning system. Our results indicate that our method achieves superior performance across all benchmarks and datasets, consistently surpassing even the algorithms it has been trained on, further demonstrating its ability to generalise beyond the capabilities of traditional algorithms.
comment: 17 pages, 9 figures
♻ ☆ TorchQL: A Programming Framework for Integrity Constraints in Machine Learning
Finding errors in machine learning applications requires a thorough exploration of their behavior over data. Existing approaches used by practitioners are often ad-hoc and lack the abstractions needed to scale this process. We present TorchQL, a programming framework to evaluate and improve the correctness of machine learning applications. TorchQL allows users to write queries to specify and check integrity constraints over machine learning models and datasets. It seamlessly integrates relational algebra with functional programming to allow for highly expressive queries using only eight intuitive operators. We evaluate TorchQL on diverse use-cases including finding critical temporal inconsistencies in objects detected across video frames in autonomous driving, finding data imputation errors in time-series medical records, finding data labeling errors in real-world images, and evaluating biases and constraining outputs of language models. Our experiments show that TorchQL enables up to 13x faster query executions than baselines like Pandas and MongoDB, and up to 40% shorter queries than native Python. We also conduct a user study and find that TorchQL is natural enough for developers familiar with Python to specify complex integrity constraints.
♻ ☆ PromptDSI: Prompt-based Rehearsal-free Instance-wise Incremental Learning for Document Retrieval
Differentiable Search Index (DSI) utilizes Pre-trained Language Models (PLMs) for efficient document retrieval without relying on external indexes. However, DSI needs full re-training to handle updates in dynamic corpora, causing significant computational inefficiencies. We introduce PromptDSI, a prompt-based rehearsal-free approach for instance-wise incremental learning document retrieval. PromptDSI attaches prompts to the frozen PLM's encoder of DSI, leveraging its powerful representation to efficiently index new corpora while maintaining a balance between stability and plasticity. We eliminate the initial forward pass of prompt-based continual learning methods that doubles training and inference time. Moreover, we propose a topic-aware prompt pool that employs neural topic embeddings as fixed keys. This strategy ensures diverse and effective prompt usage, addressing the challenge of parameter underutilization caused by the collapse of the query-key matching mechanism. Our empirical evaluations demonstrate that BERT-based PromptDSI matches IncDSI in managing forgetting while improving new corpora performance by more than 4% Hits@10 on NQ320k and upto 3% MRR@10 on MS MARCO 300k.
comment: 20 pages
♻ ☆ Data Augmentation for Continual RL via Adversarial Gradient Episodic Memory
Data efficiency of learning, which plays a key role in the Reinforcement Learning (RL) training process, becomes even more important in continual RL with sequential environments. In continual RL, the learner interacts with non-stationary, sequential tasks and is required to learn new tasks without forgetting previous knowledge. However, there is little work on implementing data augmentation for continual RL. In this paper, we investigate the efficacy of data augmentation for continual RL. Specifically, we provide benchmarking data augmentations for continual RL, by (1) summarising existing data augmentation methods and (2) including a new augmentation method for continual RL: Adversarial Augmentation with Gradient Episodic Memory (Adv-GEM). Extensive experiments show that data augmentations, such as random amplitude scaling, state-switch, mixup, adversarial augmentation, and Adv-GEM, can improve existing continual RL algorithms in terms of their average performance, catastrophic forgetting, and forward transfer, on robot control tasks. All data augmentation methods are implemented as plug-in modules for trivial integration into continual RL methods.
♻ ☆ Scalable Structure Learning for Sparse Context-Specific Systems
Several approaches to graphically representing context-specific relations among jointly distributed categorical variables have been proposed, along with structure learning algorithms. While existing optimization-based methods have limited scalability due to the large number of context-specific models, the constraint-based methods are more prone to error than even constraint-based directed acyclic graph learning algorithms since more relations must be tested. We present an algorithm for learning context-specific models that scales to hundreds of variables. Scalable learning is achieved through a combination of an order-based Markov chain Monte-Carlo search and a novel, context-specific sparsity assumption that is analogous to those typically invoked for directed acyclic graphical models. Unlike previous Markov chain Monte-Carlo search methods, our Markov chain is guaranteed to have the true posterior of the variable orderings as the stationary distribution. To implement the method, we solve a first case of an open problem recently posed by Alon and Balogh. Future work solving increasingly general instances of this problem would allow our methods to learn increasingly dense models. The method is shown to perform well on synthetic data and real world examples, in terms of both accuracy and scalability.
comment: 34 pages, 6 figures; for associated code, see https://cstrees.readthedocs.io
♻ ☆ FLEX: Expert-level False-Less EXecution Metric for Reliable Text-to-SQL Benchmark
Text-to-SQL systems have become crucial for translating natural language into SQL queries in various industries, enabling non-technical users to perform complex data operations. The need for accurate evaluation methods has increased as these systems have grown more sophisticated. However, the Execution Accuracy (EX), the most prevalent evaluation metric, still shows many false positives and negatives. Thus, this paper introduces FLEX (False-Less EXecution), a novel approach to evaluating text-to-SQL systems using large language models (LLMs) to emulate human expert-level evaluation of SQL queries. Our metric improves agreement with human experts (from 62 to 87.04 in Cohen's kappa) with comprehensive context and sophisticated criteria. Our extensive experiments yield several key insights: (1) Models' performance increases by over 2.6 points on average, substantially affecting rankings on Spider and BIRD benchmarks; (2) The underestimation of models in EX primarily stems from annotation quality issues; and (3) Model performance on particularly challenging questions tends to be overestimated. This work contributes to a more accurate and nuanced evaluation of text-to-SQL systems, potentially reshaping our understanding of state-of-the-art performance in this field.
comment: preprint, under review
♻ ☆ Mixture of Experts Made Personalized: Federated Prompt Learning for Vision-Language Models
Prompt learning for pre-trained Vision-Language Models (VLMs) like CLIP has demonstrated potent applicability across diverse downstream tasks. This lightweight approach has quickly gained traction from federated learning (FL) researchers who seek to efficiently adapt VLMs to heterogeneous scenarios. However, current federated prompt learning methods are habitually restricted to the traditional FL paradigm, where the participating clients are generally only allowed to download a single globally aggregated model from the server. While justifiable for training full-sized models under federated settings, in this work, we argue that this paradigm is ill-suited for lightweight prompts. By facilitating the clients to download multiple pre-aggregated prompts as fixed non-local experts, we propose Personalized Federated Mixture of Adaptive Prompts (pFedMoAP), a novel FL framework that personalizes the prompt learning process through the lens of Mixture of Experts (MoE). pFedMoAP implements a local attention-based gating network that learns to generate enhanced text features for better alignment with local image data on the client, benefiting from both local and downloaded non-local adaptive prompt experts. The non-local experts are sparsely selected from a server-maintained pool, fostering collaborative learning across clients. To evaluate the proposed algorithm, we conduct extensive experiments across 9 datasets under various heterogeneous federated settings. The results show that pFedMoAP consistently outperforms the state-of-the-art alternatives, underscoring its efficacy in personalizing prompt learning for CLIP within the federated learning paradigm.
comment: 16 pages, 4 figures
♻ ☆ Causal Discovery under Latent Class Confounding
An acyclic causal structure can be described with directed acyclic graph (DAG), where arrows indicate the possibility of direct causation. The task of learning this structure from data is known as "causal discovery." Diverse populations or changing environments can sometimes give rise to data that is heterogeneous in the following sense: each population/environment is a "source" which idiosyncratically determines the forms of those direct causal effects. From this perspective, the source is a latent common cause for every observed variable. While some methods for causal discovery are able to work around latent confounding in special cases, especially when only few observables are confounded, a global confounder is a difficult challenge. The only known ways to deal with latent global confounding involve assumptions that limit the structural equations and/or noise functions. We demonstrate that globally confounded causal structures can still be identifiable with arbitrary structural equations and noise functions, so long as the number of latent classes remains small relative to the size and sparsity of the underlying DAG.
♻ ☆ Parsimony or Capability? Decomposition Delivers Both in Long-term Time Series Forecasting
Long-term time series forecasting (LTSF) represents a critical frontier in time series analysis, characterized by extensive input sequences, as opposed to the shorter spans typical of traditional approaches. While longer sequences inherently offer richer information for enhanced predictive precision, prevailing studies often respond by escalating model complexity. These intricate models can inflate into millions of parameters, resulting in prohibitive parameter scales. Our study demonstrates, through both analytical and empirical evidence, that decomposition is key to containing excessive model inflation while achieving uniformly superior and robust results across various datasets. Remarkably, by tailoring decomposition to the intrinsic dynamics of time series data, our proposed model outperforms existing benchmarks, using over 99 \% fewer parameters than the majority of competing methods. Through this work, we aim to unleash the power of a restricted set of parameters by capitalizing on domain characteristics--a timely reminder that in the realm of LTSF, bigger is not invariably better.
♻ ☆ Interpret Your Decision: Logical Reasoning Regularization for Generalization in Visual Classification NeurIPS2024
Vision models excel in image classification but struggle to generalize to unseen data, such as classifying images from unseen domains or discovering novel categories. In this paper, we explore the relationship between logical reasoning and deep learning generalization in visual classification. A logical regularization termed L-Reg is derived which bridges a logical analysis framework to image classification. Our work reveals that L-Reg reduces the complexity of the model in terms of the feature distribution and classifier weights. Specifically, we unveil the interpretability brought by L-Reg, as it enables the model to extract the salient features, such as faces to persons, for classification. Theoretical analysis and experiments demonstrate that L-Reg enhances generalization across various scenarios, including multi-domain generalization and generalized category discovery. In complex real-world scenarios where images span unknown classes and unseen domains, L-Reg consistently improves generalization, highlighting its practical efficacy.
comment: Accepted by NeurIPS2024 as Spotlight
♻ ☆ Token-based Decision Criteria Are Suboptimal in In-context Learning
In-Context Learning (ICL) typically utilizes classification criteria from output probabilities of manually selected label tokens. However, we argue that such token-based classification criteria lead to suboptimal decision boundaries, despite delicate calibrations through translation and constrained rotation applied. To address this problem, we propose Hidden Calibration, which renounces token probabilities and uses the nearest centroid classifier on the LM's last hidden states. In detail, we assign the label of the nearest centroid previously estimated from a calibration set to the test sample as the predicted label. Our experiments on 6 models and 10 classification datasets indicate that Hidden Calibration consistently outperforms current token-based baselines by about 20%~50%, achieving a strong state-of-the-art in ICL. Our further analysis demonstrates that Hidden Calibration finds better classification criteria with less inter-class overlap, and LMs provide linearly separable intra-class clusters with the help of demonstrations, which supports Hidden Calibration and gives new insights into the principle of ICL.
comment: 24 pages, 15 figures, 13 tables
♻ ☆ DeltaDock: A Unified Framework for Accurate, Efficient, and Physically Reliable Molecular Docking NeurIPS'24
Molecular docking, a technique for predicting ligand binding poses, is crucial in structure-based drug design for understanding protein-ligand interactions. Recent advancements in docking methods, particularly those leveraging geometric deep learning (GDL), have demonstrated significant efficiency and accuracy advantages over traditional sampling methods. Despite these advancements, current methods are often tailored for specific docking settings, and limitations such as the neglect of protein side-chain structures, difficulties in handling large binding pockets, and challenges in predicting physically valid structures exist. To accommodate various docking settings and achieve accurate, efficient, and physically reliable docking, we propose a novel two-stage docking framework, DeltaDock, consisting of pocket prediction and site-specific docking. We innovatively reframe the pocket prediction task as a pocket-ligand alignment problem rather than direct prediction in the first stage. Then we follow a bi-level coarse-to-fine iterative refinement process to perform site-specific docking. Comprehensive experiments demonstrate the superior performance of DeltaDock. Notably, in the blind docking setting, DeltaDock achieves a 31\% relative improvement over the docking success rate compared with the previous state-of-the-art GDL model. With the consideration of physical validity, this improvement increases to about 300\%.
comment: Accepted by NeurIPS'24
♻ ☆ Sharing to learn and learning to share; Fitting together Meta-Learning, Multi-Task Learning, and Transfer Learning: A meta review
Integrating knowledge across different domains is an essential feature of human learning. Learning paradigms such as transfer learning, meta-learning, and multi-task learning reflect the human learning process by exploiting the prior knowledge for new tasks, encouraging faster learning and good generalization for new tasks. This article gives a detailed view of these learning paradigms and their comparative analysis. The weakness of one learning algorithm turns out to be a strength of another, and thus, merging them is a prevalent trait in the literature. Numerous research papers focus on each of these learning paradigms separately and provide a comprehensive overview of them. However, this article reviews research studies that combine (two of) these learning algorithms. This survey describes how these techniques are combined to solve problems in many different fields of research, including computer vision, natural language processing, hyper-spectral imaging, and many more, in a supervised setting only. Based on the knowledge accumulated from the literature, we hypothesize a generic task-agnostic and model-agnostic learning network - an ensemble of meta-learning, transfer learning, and multi-task learning, termed Multi-modal Multi-task Meta Transfer Learning. We also present some open research questions, limitations, and future research directions for this proposed network. The aim of this article is to spark interest among scholars in effectively merging existing learning algorithms with the intention of advancing research in this field. Instead of presenting experimental results, we invite readers to explore and contemplate techniques for merging algorithms while navigating through their limitations.
comment: This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may slightly change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3478805
♻ ☆ Mini-Omni2: Towards Open-source GPT-4o with Vision, Speech and Duplex Capabilities
GPT-4o, an all-encompassing model, represents a milestone in the development of large multi-modal language models. It can understand visual, auditory, and textual modalities, directly output audio, and support flexible duplex interaction. Models from the open-source community often achieve some functionalities of GPT-4o, such as visual understanding and voice chat. Nevertheless, training a unified model that incorporates all modalities is challenging due to the complexities of multi-modal data, intricate model architectures, and training processes. In this paper, we introduce Mini-Omni2, a visual-audio assistant capable of providing real-time, end-to-end voice responses to visoin and audio queries. By integrating pretrained visual and auditory encoders, Mini-Omni2 maintains performance in individual modalities. We propose a three-stage training process to align modalities, allowing the language model to handle multi-modal inputs and outputs after training on a limited dataset. For interaction, we introduce a command-based interruption mechanism, enabling more flexible interaction with users. To the best of our knowledge, Mini-Omni2 is one of the closest reproductions of GPT-4o, which have similar form of functionality, and we hope it can offer valuable insights for subsequent research.
comment: 13 pages, 6 figures
♻ ☆ FedCCRL: Federated Domain Generalization with Cross-Client Representation Learning
Domain Generalization (DG) aims to train models that can effectively generalize to unseen domains. However, in the context of Federated Learning (FL), where clients collaboratively train a model without directly sharing their data, most existing DG algorithms are not directly applicable to the FL setting due to privacy constraints, as well as the limited data quantity and domain diversity at each client. To tackle these challenges, we propose FedCCRL, a novel federated domain generalization method that significantly improves the model's ability to generalize to unseen domains without compromising privacy or incurring excessive computational and communication costs. Specifically, we adapt MixStyle to the federated setting to transfer domain-specific features while AugMix is employed to perturb domain-invariant features. Furthermore, we leverage supervised contrastive loss for representation alignment and utilize Jensen-Shannon divergence to ensure consistent predictions between original and augmented samples. Extensive experimental results demonstrate that FedCCRL achieves the state-of-the-art performances on the PACS, OfficeHome and miniDomainNet datasets across varying numbers of clients. Code is available at https://github.com/SanphouWang/FedCCRL.
♻ ☆ Mixed-Precision Federated Learning via Multi-Precision Over-The-Air Aggregation
Over-the-Air Federated Learning (OTA-FL) is a privacy-preserving distributed learning mechanism, by aggregating updates in the electromagnetic channel rather than at the server. A critical research gap in existing OTA-FL research is the assumption of homogeneous client computational bit precision. While in real world application, clients with varying hardware resources may exploit approximate computing (AxC) to operate at different bit precisions optimized for energy and computational efficiency. And model updates of various precisions amongst clients poses an open challenge for OTA-FL, as it is incompatible in the wireless modulation superposition. Here, we propose an mixed-precision OTA-FL framework of clients with multiple bit precisions, demonstrating the following innovations: (i) the superior trade-off for both server and clients within the constraints of varying edge computing capabilities, energy efficiency, and learning accuracy requirements comparing to homogeneous client bit precision, and (ii) a multi-precision gradient modulation scheme to ensure compatibility with OTA aggregation and eliminate the overheads of precision conversion. Through case study with real world data, we validate our modulation scheme that enables AxC based mixed-precision OTA-FL. In comparison to homogeneous standard precision of 32-bit and 16-bit, our framework presents more than 10% in 4-bit ultra low precision client performance and over 65%and 13% of energy savings respectively. This demonstrates the great potential of our mixed-precision OTA-FL approach in heterogeneous edge computing environments.
comment: Submitted to WCNC 2025
♻ ☆ Parallel Momentum Methods Under Biased Gradient Estimations
Parallel stochastic gradient methods are gaining prominence in solving large-scale machine learning problems that involve data distributed across multiple nodes. However, obtaining unbiased stochastic gradients, which have been the focus of most theoretical research, is challenging in many distributed machine learning applications. The gradient estimations easily become biased, for example, when gradients are compressed or clipped, when data is shuffled, and in meta-learning and reinforcement learning. In this work, we establish worst-case bounds on parallel momentum methods under biased gradient estimation on both general non-convex and $\mu$-PL problems. Our analysis covers general distributed optimization problems, and we work out the implications for special cases where gradient estimates are biased, i.e. in meta-learning and when the gradients are compressed or clipped. Our numerical experiments verify our theoretical findings and show faster convergence performance of momentum methods than traditional biased gradient descent.
comment: 12 pages
♻ ☆ Reverse Stable Diffusion: What prompt was used to generate this image?
Text-to-image diffusion models have recently attracted the interest of many researchers, and inverting the diffusion process can play an important role in better understanding the generative process and how to engineer prompts in order to obtain the desired images. To this end, we study the task of predicting the prompt embedding given an image generated by a generative diffusion model. We consider a series of white-box and black-box models (with and without access to the weights of the diffusion network) to deal with the proposed task. We propose a novel learning framework comprising a joint prompt regression and multi-label vocabulary classification objective that generates improved prompts. To further improve our method, we employ a curriculum learning procedure that promotes the learning of image-prompt pairs with lower labeling noise (i.e. that are better aligned). We conduct experiments on the DiffusionDB data set, predicting text prompts from images generated by Stable Diffusion. In addition, we make an interesting discovery: training a diffusion model on the prompt generation task can make the model generate images that are much better aligned with the input prompts, when the model is directly reused for text-to-image generation. Our code is publicly available for download at https://github.com/CroitoruAlin/Reverse-Stable-Diffusion.
comment: Accepted for publication in Computer Vision and Image Understanding
♻ ☆ Instruction-Guided Visual Masking NeurIPS 2024
Instruction following is crucial in contemporary LLM. However, when extended to multimodal setting, it often suffers from misalignment between specific textual instruction and targeted local region of an image. To achieve more accurate and nuanced multimodal instruction following, we introduce Instruction-guided Visual Masking (IVM), a new versatile visual grounding model that is compatible with diverse multimodal models, such as LMM and robot model. By constructing visual masks for instruction-irrelevant regions, IVM-enhanced multimodal models can effectively focus on task-relevant image regions to better align with complex instructions. Specifically, we design a visual masking data generation pipeline and create an IVM-Mix-1M dataset with 1 million image-instruction pairs. We further introduce a new learning technique, Discriminator Weighted Supervised Learning (DWSL) for preferential IVM training that prioritizes high-quality data samples. Experimental results on generic multimodal tasks such as VQA and embodied robotic control demonstrate the versatility of IVM, which as a plug-and-play tool, significantly boosts the performance of diverse multimodal models, yielding new state-of-the-art results across challenging multimodal benchmarks. Code, model and data are available at https://github.com/2toinf/IVM.
comment: NeurIPS 2024
♻ ☆ Collocation-based Robust Variational Physics-Informed Neural Networks (CRVPINN)
Physics-Informed Neural Networks (PINNs) have been successfully applied to solve Partial Differential Equations (PDEs). Their loss function is founded on a strong residual minimization scheme. Variational Physics-Informed Neural Networks (VPINNs) are their natural extension to weak variational settings. In this context, the recent work of Robust Variational Physics-Informed Neural Networks (RVPINNs) highlights the importance of conveniently translating the norms of the underlying continuum-level spaces to the discrete level. Otherwise, VPINNs might become unrobust, implying that residual minimization might be highly uncorrelated with a desired minimization of the error in the energy norm. However, applying this robustness to VPINNs typically entails dealing with the inverse of a Gram matrix, usually producing slow convergence speeds during training. In this work, we accelerate the implementation of RVPINN, establishing a LU factorization of sparse Gram matrix in a kind of point-collocation scheme with the same spirit as original PINNs. We call out method the Collocation-based Robust Variational Physics Informed Neural Networks (CRVPINN). We test our efficient CRVPINN algorithm on Laplace, advection-diffusion, and Stokes problems in two spatial dimensions.
comment: 39 pages, 16 figures
♻ ☆ Manifolds, Random Matrices and Spectral Gaps: The geometric phases of generative diffusion
In this paper, we investigate the latent geometry of generative diffusion models under the manifold hypothesis. To this purpose, we analyze the spectrum of eigenvalues (and singular values) of the Jacobian of the score function, whose discontinuities (gaps) reveal the presence and dimensionality of distinct sub-manifolds. Using a statistical physics approach, we derive the spectral distributions and formulas for the spectral gaps under several distributional assumptions and we compare these theoretical predictions with the spectra estimated from trained networks. Our analysis reveals the existence of three distinct qualitative phases during the generative process: a trivial phase; a manifold coverage phase where the diffusion process fits the distribution internal to the manifold; a consolidation phase where the score becomes orthogonal to the manifold and all particles are projected on the support of the data. This `division of labor' between different timescales provides an elegant explanation on why generative diffusion models are not affected by the manifold overfitting phenomenon that plagues likelihood-based models, since the internal distribution and the manifold geometry are produced at different time points during generation.
♻ ☆ Can Search-Based Testing with Pareto Optimization Effectively Cover Failure-Revealing Test Inputs?
Search-based software testing (SBST) is a widely adopted technique for testing complex systems with large input spaces, such as Deep Learning-enabled (DL-enabled) systems. Many SBST techniques focus on Pareto-based optimization, where multiple objectives are optimized in parallel to reveal failures. However, it is important to ensure that identified failures are spread throughout the entire failure-inducing area of a search domain and not clustered in a sub-region. This ensures that identified failures are semantically diverse and reveal a wide range of underlying causes. In this paper, we present a theoretical argument explaining why testing based on Pareto optimization is inadequate for covering failure-inducing areas within a search domain. We support our argument with empirical results obtained by applying two widely used types of Pareto-based optimization techniques, namely NSGA-II (an evolutionary algorithm) and OMOPSO (a swarm-based Pareto-optimization algorithm), to two DL-enabled systems: an industrial Automated Valet Parking (AVP) system and a system for classifying handwritten digits. We measure the coverage of failure-revealing test inputs in the input space using a metric that we refer to as the Coverage Inverted Distance quality indicator. Our results show that NSGA-II-based search and OMOPSO are not more effective than a na\"ive random search baseline in covering test inputs that reveal failures. The replication package for this study is available in a GitHub repository.
comment: Accepted for publication by Empirical Software Engineering Journal (EMSE) (in October 2024)
♻ ☆ A Survey of Out-of-distribution Generalization for Graph Machine Learning from a Causal View
Graph machine learning (GML) has been successfully applied across a wide range of tasks. Nonetheless, GML faces significant challenges in generalizing over out-of-distribution (OOD) data, which raises concerns about its wider applicability. Recent advancements have underscored the crucial role of causality-driven approaches in overcoming these generalization challenges. Distinct from traditional GML methods that primarily rely on statistical dependencies, causality-focused strategies delve into the underlying causal mechanisms of data generation and model prediction, thus significantly improving the generalization of GML across different environments. This paper offers a thorough review of recent progress in causality-involved GML generalization. We elucidate the fundamental concepts of employing causality to enhance graph model generalization and categorize the various approaches, providing detailed descriptions of their methodologies and the connections among them. Furthermore, we explore the incorporation of causality in other related important areas of trustworthy GML, such as explanation, fairness, and robustness. Concluding with a discussion on potential future research directions, this review seeks to articulate the continuing development and future potential of causality in enhancing the trustworthiness of graph machine learning.
comment: 15 pages, 2 figures, 1 table
♻ ☆ PIVOT-R: Primitive-Driven Waypoint-Aware World Model for Robotic Manipulation NeurIPS 2024
Language-guided robotic manipulation is a challenging task that requires an embodied agent to follow abstract user instructions to accomplish various complex manipulation tasks. Previous work trivially fitting the data without revealing the relation between instruction and low-level executable actions, these models are prone to memorizing the surficial pattern of the data instead of acquiring the transferable knowledge, and thus are fragile to dynamic environment changes. To address this issue, we propose a PrIrmitive-driVen waypOinT-aware world model for Robotic manipulation (PIVOT-R) that focuses solely on the prediction of task-relevant waypoints. Specifically, PIVOT-R consists of a Waypoint-aware World Model (WAWM) and a lightweight action prediction module. The former performs primitive action parsing and primitive-driven waypoint prediction, while the latter focuses on decoding low-level actions. Additionally, we also design an asynchronous hierarchical executor (AHE), which can use different execution frequencies for different modules of the model, thereby helping the model reduce computational redundancy and improve model execution efficiency. Our PIVOT-R outperforms state-of-the-art (SoTA) open-source models on the SeaWave benchmark, achieving an average relative improvement of 19.45% across four levels of instruction tasks. Moreover, compared to the synchronously executed PIVOT-R, the execution efficiency of PIVOT-R with AHE is increased by 28-fold, with only a 2.9% drop in performance. These results provide compelling evidence that our PIVOT-R can significantly improve both the performance and efficiency of robotic manipulation.
comment: Accepted to NeurIPS 2024
♻ ☆ Understanding What Affects the Generalization Gap in Visual Reinforcement Learning: Theory and Empirical Evidence
Recently, there are many efforts attempting to learn useful policies for continuous control in visual reinforcement learning (RL). In this scenario, it is important to learn a generalizable policy, as the testing environment may differ from the training environment, e.g., there exist distractors during deployment. Many practical algorithms are proposed to handle this problem. However, to the best of our knowledge, none of them provide a theoretical understanding of what affects the generalization gap and why their proposed methods work. In this paper, we bridge this issue by theoretically answering the key factors that contribute to the generalization gap when the testing environment has distractors. Our theories indicate that minimizing the representation distance between training and testing environments, which aligns with human intuition, is the most critical for the benefit of reducing the generalization gap. Our theoretical results are supported by the empirical evidence in the DMControl Generalization Benchmark (DMC-GB).
comment: Accepted by Journal of Artificial Intelligence Research (JAIR)
♻ ☆ Nonconvex Stochastic Bregman Proximal Gradient Method with Application to Deep Learning
Stochastic gradient methods for minimizing nonconvex composite objective functions typically rely on the Lipschitz smoothness of the differentiable part, but this assumption fails in many important problem classes like quadratic inverse problems and neural network training, leading to instability of the algorithms in both theory and practice. To address this, we propose a family of stochastic Bregman proximal gradient (SBPG) methods that only require smooth adaptivity. SBPG replaces the quadratic approximation in SGD with a Bregman proximity measure, offering a better approximation model that handles non-Lipschitz gradients in nonconvex objectives. We establish the convergence properties of vanilla SBPG and show it achieves optimal sample complexity in the nonconvex setting. Experimental results on quadratic inverse problems demonstrate SBPG's robustness in terms of stepsize selection and sensitivity to the initial point. Furthermore, we introduce a momentum-based variant, MSBPG, which enhances convergence by relaxing the mini-batch size requirement while preserving the optimal oracle complexity. We apply MSBPG to the training of deep neural networks, utilizing a polynomial kernel function to ensure smooth adaptivity of the loss function. Experimental results on benchmark datasets confirm the effectiveness and robustness of MSBPG in training neural networks. Given its negligible additional computational cost compared to SGD in large-scale optimization, MSBPG shows promise as a universal open-source optimizer for future applications.
comment: 44 pages
♻ ☆ A SARS-CoV-2 Interaction Dataset and VHH Sequence Corpus for Antibody Language Models
Antibodies are crucial proteins produced by the immune system to eliminate harmful foreign substances and have become pivotal therapeutic agents for treating human diseases. To accelerate the discovery of antibody therapeutics, there is growing interest in constructing language models using antibody sequences. However, the applicability of pre-trained language models for antibody discovery has not been thoroughly evaluated due to the scarcity of labeled datasets. To overcome these limitations, we introduce AVIDa-SARS-CoV-2, a dataset featuring the antigen-variable domain of heavy chain of heavy chain antibody (VHH) interactions obtained from two alpacas immunized with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike proteins. AVIDa-SARS-CoV-2 includes binary labels indicating the binding or non-binding of diverse VHH sequences to 12 SARS-CoV-2 mutants, such as the Delta and Omicron variants. Furthermore, we release VHHCorpus-2M, a pre-training dataset for antibody language models, containing over two million VHH sequences. We report benchmark results for predicting SARS-CoV-2-VHH binding using VHHBERT pre-trained on VHHCorpus-2M and existing general protein and antibody-specific pre-trained language models. These results confirm that AVIDa-SARS-CoV-2 provides valuable benchmarks for evaluating the representation capabilities of antibody language models for binding prediction, thereby facilitating the development of AI-driven antibody discovery. The datasets are available at https://datasets.cognanous.com.
♻ ☆ Benign Overfitting under Learning Rate Conditions for $α$ Sub-exponential Input
This paper investigates the phenomenon of benign overfitting in binary classification problems with heavy-tailed input distributions, extending the analysis of maximum margin classifiers to $\alpha$ sub-exponential distributions ($\alpha \in (0, 2]$). This generalizes previous work focused on sub-gaussian inputs. We provide generalization error bounds for linear classifiers trained using gradient descent on unregularized logistic loss in this heavy-tailed setting. Our results show that, under certain conditions on the dimensionality $p$ and the distance between the centers of the distributions, the misclassification error of the maximum margin classifier asymptotically approaches the noise level, the theoretical optimal value. Moreover, we derive an upper bound on the learning rate $\beta$ for benign overfitting to occur and show that as the tail heaviness of the input distribution $\alpha$ increases, the upper bound on the learning rate decreases. These results demonstrate that benign overfitting persists even in settings with heavier-tailed inputs than previously studied, contributing to a deeper understanding of the phenomenon in more realistic data environments.
♻ ☆ Exploring Changes in Nation Perception with Nationality-Assigned Personas in LLMs
Persona assignment has become a common strategy for customizing LLM use to particular tasks and contexts. In this study, we explore how evaluation of different nations change when LLMs are assigned specific nationality personas. We assign 193 different nationality personas (e.g., an American person) to four LLMs and examine how the LLM evaluations (or ''perceptions'')of countries change. We find that all LLM-persona combinations tend to favor Western European nations, though nation-personas push LLM behaviors to focus more on and treat the nation-persona's own region more favorably. Eastern European, Latin American, and African nations are treated more negatively by different nationality personas. We additionally find that evaluations by nation-persona LLMs of other nations correlate with human survey responses but fail to match the values closely. Our study provides insight into how biases and stereotypes are realized within LLMs when adopting different national personas. In line with the ''Blueprint for an AI Bill of Rights'', our findings underscore the critical need for developing mechanisms to ensure that LLM outputs promote fairness and avoid over-generalization.
comment: Pre-print, Under review
♻ ☆ Generative Models: What Do They Know? Do They Know Things? Let's Find Out!
Generative models excel at mimicking real scenes, suggesting they might inherently encode important intrinsic scene properties. In this paper, we aim to explore the following key questions: (1) What intrinsic knowledge do generative models like GANs, Autoregressive models, and Diffusion models encode? (2) Can we establish a general framework to recover intrinsic representations from these models, regardless of their architecture or model type? (3) How minimal can the required learnable parameters and labeled data be to successfully recover this knowledge? (4) Is there a direct link between the quality of a generative model and the accuracy of the recovered scene intrinsics? Our findings indicate that a small Low-Rank Adaptators (LoRA) can recover intrinsic images-depth, normals, albedo and shading-across different generators (Autoregressive, GANs and Diffusion) while using the same decoder head that generates the image. As LoRA is lightweight, we introduce very few learnable parameters (as few as 0.04% of Stable Diffusion model weights for a rank of 2), and we find that as few as 250 labeled images are enough to generate intrinsic images with these LoRA modules. Finally, we also show a positive correlation between the generative model's quality and the accuracy of the recovered intrinsics through control experiments.
comment: https://intrinsic-lora.github.io/
♻ ☆ Adaptation Odyssey in LLMs: Why Does Additional Pretraining Sometimes Fail to Improve? EMNLP 2024
In the last decade, the generalization and adaptation abilities of deep learning models were typically evaluated on fixed training and test distributions. Contrary to traditional deep learning, large language models (LLMs) are (i) even more overparameterized, (ii) trained on unlabeled text corpora curated from the Internet with minimal human intervention, and (iii) trained in an online fashion. These stark contrasts prevent researchers from transferring lessons learned on model generalization and adaptation in deep learning contexts to LLMs. To this end, our short paper introduces empirical observations that aim to shed light on further training of already pretrained language models. Specifically, we demonstrate that training a model on a text domain could degrade its perplexity on the test portion of the same domain. We observe with our subsequent analysis that the performance degradation is positively correlated with the similarity between the additional and the original pretraining dataset of the LLM. Our further token-level perplexity observations reveals that the perplexity degradation is due to a handful of tokens that are not informative about the domain. We hope these findings will guide us in determining when to adapt a model vs when to rely on its foundational capabilities.
comment: Accepted to EMNLP 2024 Main Conference
♻ ☆ Conv-Basis: A New Paradigm for Efficient Attention Inference and Gradient Computation in Transformers
The self-attention mechanism is the key to the success of transformers in recent Large Language Models (LLMs). However, the quadratic computational cost $O(n^2)$ in the input sequence length $n$ is a notorious obstacle for further improvement and scalability in longer contexts. In this work, we leverage the convolution-like structure of attention matrices to develop an efficient approximation method for attention computation using convolution matrices. We propose a $\mathsf{conv}$ basis system, analogous to the rank basis, and show that any lower triangular matrix can always be decomposed as a sum of structured convolution matrices in this basis. We then design a fast algorithm to approximate the attention matrix via a sum of such $k$ convolution matrices. This allows us to compute the attention {\it inference} via Fast Fourier Transforms (FFT) in $O(knd \log n)$ time, where $d$ is the hidden dimension, and thus achieve almost linear time $n^{1+o(1)}$ in the practical scenario where $kd = n^{o(1)}$. Furthermore, the attention {\it training forward} and {\it backward gradient} can be computed in $n^{1+o(1)}$ as well. We provide theoretical guarantees on the run time and approximation error and conduct preliminary experiments to evaluate its effectiveness. We hope our new paradigm for accelerating attention computation in transformer models can help their application to longer contexts.
♻ ☆ Towards Physically Consistent Deep Learning For Climate Model Parameterizations ICML
Climate models play a critical role in understanding and projecting climate change. Due to their complexity, their horizontal resolution of about 40-100 km remains too coarse to resolve processes such as clouds and convection, which need to be approximated via parameterizations. These parameterizations are a major source of systematic errors and large uncertainties in climate projections. Deep learning (DL)-based parameterizations, trained on data from computationally expensive short, high-resolution simulations, have shown great promise for improving climate models in that regard. However, their lack of interpretability and tendency to learn spurious non-physical correlations result in reduced trust in the climate simulation. We propose an efficient supervised learning framework for DL-based parameterizations that leads to physically consistent models with improved interpretability and negligible computational overhead compared to standard supervised training. First, key features determining the target physical processes are uncovered. Subsequently, the neural network is fine-tuned using only those relevant features. We show empirically that our method robustly identifies a small subset of the inputs as actual physical drivers, therefore removing spurious non-physical relationships. This results in by design physically consistent and interpretable neural networks while maintaining the predictive performance of unconstrained black-box DL-based parameterizations.
comment: Accepted at ICMLA 2024
♻ ☆ Fourier Circuits in Neural Networks and Transformers: A Case Study of Modular Arithmetic with Multiple Inputs
In the evolving landscape of machine learning, a pivotal challenge lies in deciphering the internal representations harnessed by neural networks and Transformers. Building on recent progress toward comprehending how networks execute distinct target functions, our study embarks on an exploration of the underlying reasons behind networks adopting specific computational strategies. We direct our focus to the complex algebraic learning task of modular addition involving $k$ inputs. Our research presents a thorough analytical characterization of the features learned by stylized one-hidden layer neural networks and one-layer Transformers in addressing this task. A cornerstone of our theoretical framework is the elucidation of how the principle of margin maximization shapes the features adopted by one-hidden layer neural networks. Let $p$ denote the modulus, $D_p$ denote the dataset of modular arithmetic with $k$ inputs and $m$ denote the network width. We demonstrate that a neuron count of $ m \geq 2^{2k-2} \cdot (p-1) $, these networks attain a maximum $ L_{2,k+1} $-margin on the dataset $ D_p $. Furthermore, we establish that each hidden-layer neuron aligns with a specific Fourier spectrum, integral to solving modular addition problems. By correlating our findings with the empirical observations of similar studies, we contribute to a deeper comprehension of the intrinsic computational mechanisms of neural networks. Furthermore, we observe similar computational mechanisms in attention matrices of one-layer Transformers. Our work stands as a significant stride in unraveling their operation complexities, particularly in the realm of complex algebraic tasks.
♻ ☆ Why Go Full? Elevating Federated Learning Through Partial Network Updates NeurIPS 2024
Federated learning is a distributed machine learning paradigm designed to protect user data privacy, which has been successfully implemented across various scenarios. In traditional federated learning, the entire parameter set of local models is updated and averaged in each training round. Although this full network update method maximizes knowledge acquisition and sharing for each model layer, it prevents the layers of the global model from cooperating effectively to complete the tasks of each client, a challenge we refer to as layer mismatch. This mismatch problem recurs after every parameter averaging, consequently slowing down model convergence and degrading overall performance. To address the layer mismatch issue, we introduce the FedPart method, which restricts model updates to either a single layer or a few layers during each communication round. Furthermore, to maintain the efficiency of knowledge acquisition and sharing, we develop several strategies to select trainable layers in each round, including sequential updating and multi-round cycle training. Through both theoretical analysis and experiments, our findings demonstrate that the FedPart method significantly surpasses conventional full network update strategies in terms of convergence speed and accuracy, while also reducing communication and computational overheads.
comment: 27 pages, 8 figures, accepted by NeurIPS 2024
♻ ☆ Towards Infinite-Long Prefix in Transformer
Prompting and context-based fine-tuning methods, which we call Prefix Learning, have been proposed to enhance the performance of language models on various downstream tasks. They are empirically efficient and effective, matching the performance of full parameter fine-tuning, but the theoretical understandings are limited. In this paper, we aim to address this limitation by studying their ability from the perspective of prefix length. In particular, we provide a convergence guarantee for training an ultra-long prefix in a stylized setting using the Neural Tangent Kernel (NTK) framework. Based on this strong theoretical guarantee, we design and implement an algorithm that only needs to introduce and fine-tune a few extra trainable parameters instead of an infinite-long prefix in each layer of a transformer, and can approximate the prefix attention to a guaranteed polynomial-small error. Preliminary experimental results on vision, natural language, and math data show that our method achieves superior or competitive performance compared to existing methods like full parameters fine-tuning, P-Tuning V2, and LoRA. This demonstrates our method is promising for parameter-efficient fine-tuning. Our code can be found at \url{https://github.com/ChristianYang37/chiwun/tree/main/src/NTK-Attention}.
♻ ☆ $α$-DPO: Adaptive Reward Margin is What Direct Preference Optimization Needs
Aligning large language models (LLMs) with human values and intentions is crucial for their utility, honesty, and safety. Reinforcement learning from human feedback (RLHF) is a popular approach to achieve this alignment, but it faces challenges in computational efficiency and training stability. Recent methods like Direct Preference Optimization (DPO) and Simple Preference Optimization (SimPO) have proposed offline alternatives to RLHF, simplifying the process by reparameterizing the reward function. However, DPO depends on a potentially suboptimal reference model, and SimPO's assumption of a fixed target reward margin may lead to suboptimal decisions in diverse data settings. In this work, we propose $\alpha$-DPO, an adaptive preference optimization algorithm designed to address these limitations by introducing a dynamic reward margin. Specifically, $\alpha$-DPO employs an adaptive preference distribution, balancing the policy model and the reference model to achieve personalized reward margins. We provide theoretical guarantees for $\alpha$-DPO, demonstrating its effectiveness as a surrogate optimization objective and its ability to balance alignment and diversity through KL divergence control. Empirical evaluations on AlpacaEval 2 and Arena-Hard show that $\alpha$-DPO consistently outperforms DPO and SimPO across various model settings, establishing it as a robust approach for fine-tuning LLMs. Our method achieves significant improvements in win rates, highlighting its potential as a powerful tool for LLM alignment. The code is available at https://github.com/junkangwu/alpha-DPO
♻ ☆ How does Your RL Agent Explore? An Optimal Transport Analysis of Occupancy Measure Trajectories
The rising successes of RL are propelled by combining smart algorithmic strategies and deep architectures to optimize the distribution of returns and visitations over the state-action space. A quantitative framework to compare the learning processes of these eclectic RL algorithms is currently absent but desired in practice. We address this gap by representing the learning process of an RL algorithm as a sequence of policies generated during training, and then studying the policy trajectory induced in the manifold of state-action occupancy measures. Using an optimal transport-based metric, we measure the length of the paths induced by the policy sequence yielded by an RL algorithm between an initial policy and a final optimal policy. Hence, we first define the 'Effort of Sequential Learning' (ESL). ESL quantifies the relative distance that an RL algorithm travels compared to the shortest path from the initial to the optimal policy. Further, we connect the dynamics of policies in the occupancy measure space and regret (another metric to understand the suboptimality of an RL algorithm), by defining the 'Optimal Movement Ratio' (OMR). OMR assesses the fraction of movements in the occupancy measure space that effectively reduce an analogue of regret. Finally, we derive approximation guarantees to estimate ESL and OMR with finite number of samples and without access to an optimal policy. Through empirical analyses across various environments and algorithms, we demonstrate that ESL and OMR provide insights into the exploration processes of RL algorithms and hardness of different tasks in discrete and continuous MDPs.
♻ ☆ Can LLMs Patch Security Issues?
Large Language Models (LLMs) have shown impressive proficiency in code generation. Unfortunately, these models share a weakness with their human counterparts: producing code that inadvertently has security vulnerabilities. These vulnerabilities could allow unauthorized attackers to access sensitive data or systems, which is unacceptable for safety-critical applications. In this work, we propose Feedback-Driven Security Patching (FDSP), where LLMs automatically refine generated, vulnerable code. Our approach leverages automatic static code analysis to empower the LLM to generate and implement potential solutions to address vulnerabilities. We address the research communitys needs for safe code generation by introducing a large-scale dataset, PythonSecurityEval, covering the diversity of real-world applications, including databases, websites and operating systems. We empirically validate that FDSP outperforms prior work that uses self-feedback from LLMs by up to 17.6% through our procedure that injects targeted, external feedback. Code and data are available at \url{https://github.com/Kamel773/LLM-code-refine}
♻ ☆ Instruction Tuning for Large Language Models: A Survey
This paper surveys research works in the quickly advancing field of instruction tuning (IT), a crucial technique to enhance the capabilities and controllability of large language models (LLMs). Instruction tuning refers to the process of further training LLMs on a dataset consisting of \textsc{(instruction, output)} pairs in a supervised fashion, which bridges the gap between the next-word prediction objective of LLMs and the users' objective of having LLMs adhere to human instructions. In this work, we make a systematic review of the literature, including the general methodology of IT, the construction of IT datasets, the training of IT models, and applications to different modalities, domains and applications, along with an analysis on aspects that influence the outcome of IT (e.g., generation of instruction outputs, size of the instruction dataset, etc). We also review the potential pitfalls of IT along with criticism against it, along with efforts pointing out current deficiencies of existing strategies and suggest some avenues for fruitful research. Project page: github.com/xiaoya-li/Instruction-Tuning-Survey
comment: V3; Last update: Oct 16, 2024
♻ ☆ Investigating the Transferability of Code Repair for Low-Resource Programming Languages
Large language models (LLMs) have shown remarkable performance on code generation tasks. A recent use case is iterative code repair, where an LLM fixes an incorrect program by rationalizing about errors and generating new code. Recent works augment the code repair process by integrating modern techniques such as chain-of-thought reasoning or distillation, but only study their benefits on high-resource languages like Python, and ignore low-resource languages like Perl. To address this gap of knowledge, we investigate the benefits of distilling code repair for both high and low resource languages to determine if the techniques that are effective in a high resource setting are also applicable in a low resource setting. Our evaluation shows that distilling the ability to repair code has language dependent benefits. To explain this behavior, we perform a further analysis and find that contrary to preexisting beliefs, the correlation between reasoning ability and code correction ability is weak. We hypothesize this weak correlation is magnified in low-resource settings where base models lack deep knowledge of a programming language, leading to wavering benefits of code repair.
♻ ☆ AnimateLCM: Computation-Efficient Personalized Style Video Generation without Personalized Video Data SIGGRAPH
This paper introduces an effective method for computation-efficient personalized style video generation without requiring access to any personalized video data. It reduces the necessary generation time of similarly sized video diffusion models from 25 seconds to around 1 second while maintaining the same level of performance. The method's effectiveness lies in its dual-level decoupling learning approach: 1) separating the learning of video style from video generation acceleration, which allows for personalized style video generation without any personalized style video data, and 2) separating the acceleration of image generation from the acceleration of video motion generation, enhancing training efficiency and mitigating the negative effects of low-quality video data.
comment: Accepted as a Short Paper by SIGGRAPH ASIA 2024 Technical Communications. This is a short version of the original work. Project Page: https://animatelcm.github.io/
♻ ☆ FredNormer: Frequency Domain Normalization for Non-stationary Time Series Forecasting
Recent normalization-based methods have shown great success in tackling the distribution shift issue, facilitating non-stationary time series forecasting. Since these methods operate in the time domain, they may fail to fully capture the dynamic patterns that are more apparent in the frequency domain, leading to suboptimal results. This paper first theoretically analyzes how normalization methods affect frequency components. We prove that the current normalization methods that operate in the time domain uniformly scale non-zero frequencies, and thus, they struggle to determine components that contribute to more robust forecasting. Therefore, we propose FredNormer, which observes datasets from a frequency perspective and adaptively up-weights the key frequency components. To this end, FredNormer consists of two components: a statistical metric that normalizes the input samples based on their frequency stability and a learnable weighting layer that adjusts stability and introduces sample-specific variations. Notably, FredNormer is a plug-and-play module, which does not compromise the efficiency compared to existing normalization methods. Extensive experiments show that FredNormer improves the averaged MSE of backbone forecasting models by 33.3% and 55.3% on the ETTm2 dataset. Compared to the baseline normalization methods, FredNormer achieves 18 top-1 results and 6 top-2 results out of 28 settings.
♻ ☆ Sample what you cant compress
For learned image representations, basic autoencoders often produce blurry results. Reconstruction quality can be improved by incorporating additional penalties such as adversarial (GAN) and perceptual losses. Arguably, these approaches lack a principled interpretation. Concurrently, in generative settings diffusion has demonstrated a remarkable ability to create crisp, high quality results and has solid theoretical underpinnings (from variational inference to direct study as the Fisher Divergence). Our work combines autoencoder representation learning with diffusion and is, to our knowledge, the first to demonstrate the efficacy of jointly learning a continuous encoder and decoder under a diffusion-based loss. We demonstrate that this approach yields better reconstruction quality as compared to GAN-based autoencoders while being easier to tune. We also show that the resulting representation is easier to model with a latent diffusion model as compared to the representation obtained from a state-of-the-art GAN-based loss. Since our decoder is stochastic, it can generate details not encoded in the otherwise deterministic latent representation; we therefore name our approach "Sample what you can't compress", or SWYCC for short.
♻ ☆ How Transformers Implement Induction Heads: Approximation and Optimization Analysis
Transformers have demonstrated exceptional in-context learning capabilities, yet the theoretical understanding of the underlying mechanisms remain limited. A recent work (Elhage et al., 2021) identified a "rich" in-context mechanism known as induction head, contrasting with "lazy" $n$-gram models that overlook long-range dependencies. In this work, we provide both approximation and optimization analyses of how transformers implement induction heads. In the approximation analysis, we formalize both standard and generalized induction head mechanisms, and examine how transformers can efficiently implement them, with an emphasis on the distinct role of each transformer submodule. For the optimization analysis, we study the training dynamics on a synthetic mixed target, composed of a 4-gram and an in-context 2-gram component. This setting enables us to precisely characterize the entire training process and uncover an {\em abrupt transition} from lazy (4-gram) to rich (induction head) mechanisms as training progresses.
comment: 39 pages
♻ ☆ Knowledge Circuits in Pretrained Transformers NeurIPS 2024
The remarkable capabilities of modern large language models are rooted in their vast repositories of knowledge encoded within their parameters, enabling them to perceive the world and engage in reasoning. The inner workings of how these models store knowledge have long been a subject of intense interest and investigation among researchers. To date, most studies have concentrated on isolated components within these models, such as the Multilayer Perceptrons and attention head. In this paper, we delve into the computation graph of the language model to uncover the knowledge circuits that are instrumental in articulating specific knowledge. The experiments, conducted with GPT2 and TinyLLAMA, have allowed us to observe how certain information heads, relation heads, and Multilayer Perceptrons collaboratively encode knowledge within the model. Moreover, we evaluate the impact of current knowledge editing techniques on these knowledge circuits, providing deeper insights into the functioning and constraints of these editing methodologies. Finally, we utilize knowledge circuits to analyze and interpret language model behaviors such as hallucinations and in-context learning. We believe the knowledge circuits hold potential for advancing our understanding of Transformers and guiding the improved design of knowledge editing. Code and data are available in https://github.com/zjunlp/KnowledgeCircuits.
comment: NeurIPS 2024, 32 pages
♻ ☆ Which Spaces can be Embedded in $L_p$-type Reproducing Kernel Banach Space? A Characterization via Metric Entropy
In this paper, we establish a novel connection between the metric entropy growth and the embeddability of function spaces into reproducing kernel Hilbert/Banach spaces. Metric entropy characterizes the information complexity of function spaces and has implications for their approximability and learnability. Classical results show that embedding a function space into a reproducing kernel Hilbert space (RKHS) implies a bound on its metric entropy growth. Surprisingly, we prove a \textbf{converse}: a bound on the metric entropy growth of a function space allows its embedding to a $L_p-$type Reproducing Kernel Banach Space (RKBS). This shows that the ${L}_p-$type RKBS provides a broad modeling framework for learnable function classes with controlled metric entropies. Our results shed new light on the power and limitations of kernel methods for learning complex function spaces.
♻ ☆ Implicit Counterfactual Data Augmentation for Robust Learning
Machine learning models are prone to capturing the spurious correlations between non-causal attributes and classes, with counterfactual data augmentation being a promising direction for breaking these spurious associations. However, generating counterfactual data explicitly poses a challenge, and incorporating augmented data into the training process decreases training efficiency. This study proposes an Implicit Counterfactual Data Augmentation (ICDA) method to remove spurious correlations and make stable predictions. Specifically, first, a novel sample-wise augmentation strategy is developed that generates semantically and counterfactually meaningful deep features with distinct augmentation strength for each sample. Second, we derive an easy-to-compute surrogate loss on the augmented feature set when the number of augmented samples becomes infinite. Third, two concrete schemes are proposed, including direct quantification and meta-learning, to derive the key parameters for the robust loss. In addition, ICDA is explained from a regularization perspective, revealing its capacity to improve intra-class compactness and augment margins at both class and sample levels. Extensive experiments have been conducted across various biased learning scenarios covering both image and text datasets, demonstrating that ICDA consistently enhances the generalization and robustness performance of popular networks.
comment: 12 pages, 11 figures
♻ ☆ Rewards-in-Context: Multi-objective Alignment of Foundation Models with Dynamic Preference Adjustment ICML 2024
We consider the problem of multi-objective alignment of foundation models with human preferences, which is a critical step towards helpful and harmless AI systems. However, it is generally costly and unstable to fine-tune large foundation models using reinforcement learning (RL), and the multi-dimensionality, heterogeneity, and conflicting nature of human preferences further complicate the alignment process. In this paper, we introduce Rewards-in-Context (RiC), which conditions the response of a foundation model on multiple rewards in its prompt context and applies supervised fine-tuning for alignment. The salient features of RiC are simplicity and adaptivity, as it only requires supervised fine-tuning of a single foundation model and supports dynamic adjustment for user preferences during inference time. Inspired by the analytical solution of an abstracted convex optimization problem, our dynamic inference-time adjustment method approaches the Pareto-optimal solution for multiple objectives. Empirical evidence demonstrates the efficacy of our method in aligning both Large Language Models (LLMs) and diffusion models to accommodate diverse rewards with only around 10% GPU hours compared with multi-objective RL baseline.
comment: Accepted by ICML 2024
♻ ☆ $\textit{lucie}$: An Improved Python Package for Loading Datasets from the UCI Machine Learning Repository
The University of California--Irvine (UCI) Machine Learning (ML) Repository (UCIMLR) is consistently cited as one of the most popular dataset repositories, hosting hundreds of high-impact datasets. However, a significant portion, including 28.4% of the top 250, cannot be imported via the $\textit{ucimlrepo}$ package that is provided and recommended by the UCIMLR website. Instead, they are hosted as .zip files, containing nonstandard formats that are difficult to import without additional ad hoc processing. To address this issue, here we present $\textit{lucie}$ -- $\underline{l}oad$ $\underline{U}niversity$ $\underline{C}alifornia$ $\underline{I}rvine$ $\underline{e}xamples$ -- a utility that automatically determines the data format and imports many of these previously non-importable datasets, while preserving as much of a tabular data structure as possible. $\textit{lucie}$ was designed using the top 100 most popular datasets and benchmarked on the next 130, where it resulted in a success rate of 95.4% vs. 73.1% for $\textit{ucimlrepo}$. $\textit{lucie}$ is available as a Python package on PyPI with 98% code coverage.
comment: 5 pages, 3 figures
♻ ☆ CVCP-Fusion: On Implicit Depth Estimation for 3D Bounding Box Prediction
Combining LiDAR and Camera-view data has become a common approach for 3D Object Detection. However, previous approaches combine the two input streams at a point-level, throwing away semantic information derived from camera features. In this paper we propose Cross-View Center Point-Fusion, a state-of-the-art model to perform 3D object detection by combining camera and LiDAR-derived features in the BEV space to preserve semantic density from the camera stream while incorporating spacial data from the LiDAR stream. Our architecture utilizes aspects from previously established algorithms, Cross-View Transformers and CenterPoint, and runs their backbones in parallel, allowing efficient computation for real-time processing and application. In this paper we find that while an implicitly calculated depth-estimate may be sufficiently accurate in a 2D map-view representation, explicitly calculated geometric and spacial information is needed for precise bounding box prediction in the 3D world-view space.
comment: 7 pages, 5 figures. arXiv admin note: text overlap with arXiv:2205.02833 by other authors
♻ ☆ ECoDe: A Sample-Efficient Method for Co-Design of Robotic Agents
Co-designing autonomous robotic agents involves simultaneously optimizing the controller and physical design of the agent. Its inherent bi-level optimization formulation necessitates an outer loop design optimization driven by an inner loop control optimization. This can be challenging when the design space is large and each design evaluation involves a data-intensive reinforcement learning process for control optimization. To improve the sample efficiency of co-design, we propose a multi-fidelity-based exploration strategy in which we tie the controllers learned across the design spaces through a universal policy learner for warm-starting subsequent controller learning problems. Experiments performed on a wide range of agent design problems demonstrate the superiority of our method compared to baselines. Additionally, analysis of the optimized designs shows interesting design alterations, including design simplifications and non-intuitive alterations.
comment: 17 pages, 10 figures
Multimedia 8
☆ Embedding an Ethical Mind: Aligning Text-to-Image Synthesis via Lightweight Value Optimization
Recent advancements in diffusion models trained on large-scale data have enabled the generation of indistinguishable human-level images, yet they often produce harmful content misaligned with human values, e.g., social bias, and offensive content. Despite extensive research on Large Language Models (LLMs), the challenge of Text-to-Image (T2I) model alignment remains largely unexplored. Addressing this problem, we propose LiVO (Lightweight Value Optimization), a novel lightweight method for aligning T2I models with human values. LiVO only optimizes a plug-and-play value encoder to integrate a specified value principle with the input prompt, allowing the control of generated images over both semantics and values. Specifically, we design a diffusion model-tailored preference optimization loss, which theoretically approximates the Bradley-Terry model used in LLM alignment but provides a more flexible trade-off between image quality and value conformity. To optimize the value encoder, we also develop a framework to automatically construct a text-image preference dataset of 86k (prompt, aligned image, violating image, value principle) samples. Without updating most model parameters and through adaptive value selection from the input prompt, LiVO significantly reduces harmful outputs and achieves faster convergence, surpassing several strong baselines and taking an initial step towards ethically aligned T2I models.
comment: Accepted by ACM Multimedia 2024. The dataset and code can be found at https://github.com/achernarwang/LiVO
☆ Beyond Coarse-Grained Matching in Video-Text Retrieval ACCV 2024
Video-text retrieval has seen significant advancements, yet the ability of models to discern subtle differences in captions still requires verification. In this paper, we introduce a new approach for fine-grained evaluation. Our approach can be applied to existing datasets by automatically generating hard negative test captions with subtle single-word variations across nouns, verbs, adjectives, adverbs, and prepositions. We perform comprehensive experiments using four state-of-the-art models across two standard benchmarks (MSR-VTT and VATEX) and two specially curated datasets enriched with detailed descriptions (VLN-UVO and VLN-OOPS), resulting in a number of novel insights: 1) our analyses show that the current evaluation benchmarks fall short in detecting a model's ability to perceive subtle single-word differences, 2) our fine-grained evaluation highlights the difficulty models face in distinguishing such subtle variations. To enhance fine-grained understanding, we propose a new baseline that can be easily combined with current methods. Experiments on our fine-grained evaluations demonstrate that this approach enhances a model's ability to understand fine-grained differences.
comment: Accepted to ACCV 2024
☆ Rethinking Bjøntegaard Delta for Compression Efficiency Evaluation: Are We Calculating It Precisely and Reliably?
For decades, the Bj{\o}ntegaard Delta (BD) has been the metric for evaluating codec Rate-Distortion (R-D) performance. Yet, in most studies, BD is determined using just 4-5 R-D data points, could this be sufficient? As codecs and quality metrics advance, does the conventional BD estimation still hold up? Crucially, are the performance improvements of new codecs and tools genuine, or merely artifacts of estimation flaws? This paper addresses these concerns by reevaluating BD estimation. We present a novel approach employing a parameterized deep neural network to model R-D curves with high precision across various metrics, accompanied by a comprehensive R-D dataset. This approach both assesses the reliability of BD calculations and serves as a precise BD estimator. Our findings advocate for the adoption of rigorous R-D sampling and reliability metrics in future compression research to ensure the validity and reliability of results.
☆ OmnixR: Evaluating Omni-modality Language Models on Reasoning across Modalities
We introduce OmnixR, an evaluation suite designed to benchmark SoTA Omni-modality Language Models, such as GPT-4o and Gemini. Evaluating OLMs, which integrate multiple modalities such as text, vision, and audio, presents unique challenges. Particularly, the user message might often consist of multiple modalities, such that OLMs have to establish holistic understanding and reasoning across modalities to accomplish the task. Existing benchmarks are limited to single modality or dual-modality tasks, overlooking comprehensive multi-modal assessments of model reasoning. To address this, OmnixR offers two evaluation variants: (1)synthetic subset: a synthetic dataset generated automatically by translating text into multiple modalities--audio, images, video, and hybrids (Omnify). (2)realistic subset: a real-world dataset, manually curated and annotated by experts, for evaluating cross-modal reasoning in natural settings. OmnixR presents a unique evaluation towards assessing OLMs over a diverse mix of modalities, such as a question that involves video, audio, and text, providing a rigorous cross-modal reasoning testbed unlike any existing benchmarks. Our experiments find that all state-of-the-art OLMs struggle with OmnixR questions that require integrating information from multiple modalities to answer. Further analysis highlights differences in reasoning behavior, underscoring the challenges of omni-modal AI alignment.
comment: 19 pages, 6 figures, 12 tables
☆ Test-time adaptation for image compression with distribution regularization
Current test- or compression-time adaptation image compression (TTA-IC) approaches, which leverage both latent and decoder refinements as a two-step adaptation scheme, have potentially enhanced the rate-distortion (R-D) performance of learned image compression models on cross-domain compression tasks, \textit{e.g.,} from natural to screen content images. However, compared with the emergence of various decoder refinement variants, the latent refinement, as an inseparable ingredient, is barely tailored to cross-domain scenarios. To this end, we aim to develop an advanced latent refinement method by extending the effective hybrid latent refinement (HLR) method, which is designed for \textit{in-domain} inference improvement but shows noticeable degradation of the rate cost in \textit{cross-domain} tasks. Specifically, we first provide theoretical analyses, in a cue of marginalization approximation from in- to cross-domain scenarios, to uncover that the vanilla HLR suffers from an underlying mismatch between refined Gaussian conditional and hyperprior distributions, leading to deteriorated joint probability approximation of marginal distribution with increased rate consumption. To remedy this issue, we introduce a simple Bayesian approximation-endowed \textit{distribution regularization} to encourage learning a better joint probability approximation in a plug-and-play manner. Extensive experiments on six in- and cross-domain datasets demonstrate that our proposed method not only improves the R-D performance compared with other latent refinement counterparts, but also can be flexibly integrated into existing TTA-IC methods with incremental benefits.
☆ Self-Comparison for Dataset-Level Membership Inference in Large (Vision-)Language Models
Large Language Models (LLMs) and Vision-Language Models (VLMs) have made significant advancements in a wide range of natural language processing and vision-language tasks. Access to large web-scale datasets has been a key factor in their success. However, concerns have been raised about the unauthorized use of copyrighted materials and potential copyright infringement. Existing methods, such as sample-level Membership Inference Attacks (MIA) and distribution-based dataset inference, distinguish member data (data used for training) and non-member data by leveraging the common observation that models tend to memorize and show greater confidence in member data. Nevertheless, these methods face challenges when applied to LLMs and VLMs, such as the requirement for ground-truth member data or non-member data that shares the same distribution as the test data. In this paper, we propose a novel dataset-level membership inference method based on Self-Comparison. We find that a member prefix followed by a non-member suffix (paraphrased from a member suffix) can further trigger the model's memorization on training data. Instead of directly comparing member and non-member data, we introduce paraphrasing to the second half of the sequence and evaluate how the likelihood changes before and after paraphrasing. Unlike prior approaches, our method does not require access to ground-truth member data or non-member data in identical distribution, making it more practical. Extensive experiments demonstrate that our proposed method outperforms traditional MIA and dataset inference techniques across various datasets and models, including including public models, fine-tuned models, and API-based commercial models.
☆ MuVi: Video-to-Music Generation with Semantic Alignment and Rhythmic Synchronization
Generating music that aligns with the visual content of a video has been a challenging task, as it requires a deep understanding of visual semantics and involves generating music whose melody, rhythm, and dynamics harmonize with the visual narratives. This paper presents MuVi, a novel framework that effectively addresses these challenges to enhance the cohesion and immersive experience of audio-visual content. MuVi analyzes video content through a specially designed visual adaptor to extract contextually and temporally relevant features. These features are used to generate music that not only matches the video's mood and theme but also its rhythm and pacing. We also introduce a contrastive music-visual pre-training scheme to ensure synchronization, based on the periodicity nature of music phrases. In addition, we demonstrate that our flow-matching-based music generator has in-context learning ability, allowing us to control the style and genre of the generated music. Experimental results show that MuVi demonstrates superior performance in both audio quality and temporal synchronization. The generated music video samples are available at https://muvi-v2m.github.io.
comment: Working in progress
♻ ☆ Video-to-Audio Generation with Hidden Alignment
Generating semantically and temporally aligned audio content in accordance with video input has become a focal point for researchers, particularly following the remarkable breakthrough in text-to-video generation. In this work, we aim to offer insights into the video-to-audio generation paradigm, focusing on three crucial aspects: vision encoders, auxiliary embeddings, and data augmentation techniques. Beginning with a foundational model built on a simple yet surprisingly effective intuition, we explore various vision encoders and auxiliary embeddings through ablation studies. Employing a comprehensive evaluation pipeline that emphasizes generation quality and video-audio synchronization alignment, we demonstrate that our model exhibits state-of-the-art video-to-audio generation capabilities. Furthermore, we provide critical insights into the impact of different data augmentation methods on enhancing the generation framework's overall capacity. We showcase possibilities to advance the challenge of generating synchronized audio from semantic and temporal perspectives. We hope these insights will serve as a stepping stone toward developing more realistic and accurate audio-visual generation models.
comment: https://sites.google.com/view/vta-ldm
Computation and Language 183
☆ A Hitchhiker's Guide to Scaling Law Estimation
Scaling laws predict the loss of a target machine learning model by extrapolating from easier-to-train models with fewer parameters or smaller training sets. This provides an efficient way for practitioners and researchers alike to compare pretraining decisions involving optimizers, datasets, and model architectures. Despite the widespread use of scaling laws to model the dynamics of language model training, there has been little work on understanding how to best estimate and interpret them. We collect (and release) a large-scale dataset containing losses and downstream evaluations for 485 previously published pretrained models. We use these to estimate more than 1000 scaling laws, then derive a set of best practices for estimating scaling laws in new model families. We find that fitting scaling laws to intermediate checkpoints of training runs (and not just their final losses) substantially improves accuracy, and that -- all else equal -- estimates of performance are generally most accurate when derived from other models of similar sizes. However, because there is a significant degree of variability across model seeds, training multiple small models is sometimes more useful than training a single large one. Moreover, while different model families differ scaling behavior, they are often similar enough that a target model's behavior can be predicted from a single model with the same architecture, along with scaling parameter estimates derived from other model families.
☆ NesTools: A Dataset for Evaluating Nested Tool Learning Abilities of Large Language Models
Large language models (LLMs) combined with tool learning have gained impressive results in real-world applications. During tool learning, LLMs may call multiple tools in nested orders, where the latter tool call may take the former response as its input parameters. However, current research on the nested tool learning capabilities is still under-explored, since the existing benchmarks lack of relevant data instances. To address this problem, we introduce NesTools to bridge the current gap in comprehensive nested tool learning evaluations. NesTools comprises a novel automatic data generation method to construct large-scale nested tool calls with different nesting structures. With manual review and refinement, the dataset is in high quality and closely aligned with real-world scenarios. Therefore, NesTools can serve as a new benchmark to evaluate the nested tool learning abilities of LLMs. We conduct extensive experiments on 22 LLMs, and provide in-depth analyses with NesTools, which shows that current LLMs still suffer from the complex nested tool learning task.
☆ Selection-p: Self-Supervised Task-Agnostic Prompt Compression for Faithfulness and Transferability EMNLP 2024
Large Language Models (LLMs) have demonstrated impressive capabilities in a wide range of natural language processing tasks when leveraging in-context learning. To mitigate the additional computational and financial costs associated with in-context learning, several prompt compression methods have been proposed to compress the in-context learning prompts. Despite their success, these methods face challenges with transferability due to model-specific compression, or rely on external training data, such as GPT-4. In this paper, we investigate the ability of LLMs to develop a unified compression method that discretizes uninformative tokens, utilizing a self-supervised pre-training technique. By introducing a small number of parameters during the continual pre-training, the proposed Selection-p produces a probability for each input token, indicating whether to preserve or discard it. Experiments show Selection-p achieves state-of-the-art performance across numerous classification tasks, achieving compression rates of up to 10 times while experiencing only a marginal 0.8% decrease in performance. Moreover, it exhibits superior transferability to different models compared to prior work. Additionally, we further analyze how Selection-p helps maintain performance on in-context learning with long contexts.
comment: 14 pages, 5 figures, 10 tables, EMNLP 2024 Findings
☆ MLLM can see? Dynamic Correction Decoding for Hallucination Mitigation
Multimodal Large Language Models (MLLMs) frequently exhibit hallucination phenomena, but the underlying reasons remain poorly understood. In this paper, we present an empirical analysis and find that, although MLLMs incorrectly generate the objects in the final output, they are actually able to recognize visual objects in the preceding layers. We speculate that this may be due to the strong knowledge priors of the language model suppressing the visual information, leading to hallucinations. Motivated by this, we propose a novel dynamic correction decoding method for MLLMs (DeCo), which adaptively selects the appropriate preceding layers and proportionally integrates knowledge into the final layer to adjust the output logits. Note that DeCo is model agnostic and can be seamlessly incorporated with various classic decoding strategies and applied to different MLLMs. We evaluate DeCo on widely-used benchmarks, demonstrating that it can reduce hallucination rates by a large margin compared to baselines, highlighting its potential to mitigate hallucinations. Code is available at https://github.com/zjunlp/DeCo.
comment: Ongoing work
☆ Layer-wise Importance Matters: Less Memory for Better Performance in Parameter-efficient Fine-tuning of Large Language Models EMNLP 2024
Parameter-Efficient Fine-Tuning (PEFT) methods have gained significant popularity for adapting pre-trained Large Language Models (LLMs) to downstream tasks, primarily due to their potential to significantly reduce memory and computational overheads. However, a common limitation in most PEFT approaches is their application of a uniform architectural design across all layers. This uniformity involves identical trainable modules and ignores the varying importance of each layer, leading to sub-optimal fine-tuning results. To overcome the above limitation and obtain better performance, we develop a novel approach, Importance-aware Sparse Tuning (IST), to fully utilize the inherent sparsity and select the most important subset of full layers with effective layer-wise importance scoring. The proposed IST is a versatile and plug-and-play technique compatible with various PEFT methods that operate on a per-layer basis. By leveraging the estimated importance scores, IST dynamically updates these selected layers in PEFT modules, leading to reduced memory demands. We further provide theoretical proof of convergence and empirical evidence of superior performance to demonstrate the advantages of IST over uniform updating strategies. Extensive experiments on a range of LLMs, PEFTs, and downstream tasks substantiate the effectiveness of our proposed method, showcasing IST's capacity to enhance existing layer-based PEFT methods. Our code is available at https://github.com/Kaiseem/IST.
comment: EMNLP 2024
☆ Latent Action Pretraining from Videos
We introduce Latent Action Pretraining for general Action models (LAPA), an unsupervised method for pretraining Vision-Language-Action (VLA) models without ground-truth robot action labels. Existing Vision-Language-Action models require action labels typically collected by human teleoperators during pretraining, which significantly limits possible data sources and scale. In this work, we propose a method to learn from internet-scale videos that do not have robot action labels. We first train an action quantization model leveraging VQ-VAE-based objective to learn discrete latent actions between image frames, then pretrain a latent VLA model to predict these latent actions from observations and task descriptions, and finally finetune the VLA on small-scale robot manipulation data to map from latent to robot actions. Experimental results demonstrate that our method significantly outperforms existing techniques that train robot manipulation policies from large-scale videos. Furthermore, it outperforms the state-of-the-art VLA model trained with robotic action labels on real-world manipulation tasks that require language conditioning, generalization to unseen objects, and semantic generalization to unseen instructions. Training only on human manipulation videos also shows positive transfer, opening up the potential for leveraging web-scale data for robotics foundation model.
comment: Website: https://latentactionpretraining.github.io
☆ Personas with Attitudes: Controlling LLMs for Diverse Data Annotation
We present a novel approach for enhancing diversity and control in data annotation tasks by personalizing large language models (LLMs). We investigate the impact of injecting diverse persona descriptions into LLM prompts across two studies, exploring whether personas increase annotation diversity and whether the impacts of individual personas on the resulting annotations are consistent and controllable. Our results show that persona-prompted LLMs produce more diverse annotations than LLMs prompted without personas and that these effects are both controllable and repeatable, making our approach a suitable tool for improving data annotation in subjective NLP tasks like toxicity detection.
comment: 21 pages, 13 figures
☆ Converging to a Lingua Franca: Evolution of Linguistic Regions and Semantics Alignment in Multilingual Large Language Models
Large language models (LLMs) have demonstrated remarkable performance, particularly in multilingual contexts. While recent studies suggest that LLMs can transfer skills learned in one language to others, the internal mechanisms behind this ability remain unclear. We observed that the neuron activation patterns of LLMs exhibit similarities when processing the same language, revealing the existence and location of key linguistic regions. Additionally, we found that neuron activation patterns are similar when processing sentences with the same semantic meaning in different languages. This indicates that LLMs map semantically identical inputs from different languages into a "Lingua Franca", a common semantic latent space that allows for consistent processing across languages. This semantic alignment becomes more pronounced with training and increased model size, resulting in a more language-agnostic activation pattern. Moreover, we found that key linguistic neurons are concentrated in the first and last layers of LLMs, becoming denser in the first layers as training progresses. Experiments on BLOOM and LLaMA2 support these findings, highlighting the structural evolution of multilingual LLMs during training and scaling up. This paper provides insights into the internal workings of LLMs, offering a foundation for future improvements in their cross-lingual capabilities.
comment: 16 pages, 11 figures, 4 tables
☆ MTU-Bench: A Multi-granularity Tool-Use Benchmark for Large Language Models
Large Language Models (LLMs) have displayed massive improvements in reasoning and decision-making skills and can hold natural conversations with users. Recently, many tool-use benchmark datasets have been proposed. However, existing datasets have the following limitations: (1). Insufficient evaluation scenarios (e.g., only cover limited tool-use scenes). (2). Extensive evaluation costs (e.g., GPT API costs). To address these limitations, in this work, we propose a multi-granularity tool-use benchmark for large language models called MTU-Bench. For the "multi-granularity" property, our MTU-Bench covers five tool usage scenes (i.e., single-turn and single-tool, single-turn and multiple-tool, multiple-turn and single-tool, multiple-turn and multiple-tool, and out-of-distribution tasks). Besides, all evaluation metrics of our MTU-Bench are based on the prediction results and the ground truth without using any GPT or human evaluation metrics. Moreover, our MTU-Bench is collected by transforming existing high-quality datasets to simulate real-world tool usage scenarios, and we also propose an instruction dataset called MTU-Instruct data to enhance the tool-use abilities of existing LLMs. Comprehensive experimental results demonstrate the effectiveness of our MTU-Bench. Code and data will be released at https: //github.com/MTU-Bench-Team/MTU-Bench.git.
☆ Magnifier Prompt: Tackling Multimodal Hallucination via Extremely Simple Instructions
Hallucinations in multimodal large language models (MLLMs) hinder their practical applications. To address this, we propose a Magnifier Prompt (MagPrompt), a simple yet effective method to tackle hallucinations in MLLMs via extremely simple instructions. MagPrompt is based on the following two key principles, which guide the design of various effective prompts, demonstrating robustness: (1) MLLMs should focus more on the image. (2) When there are conflicts between the image and the model's inner knowledge, MLLMs should prioritize the image. MagPrompt is training-free and can be applied to open-source and closed-source models, such as GPT-4o and Gemini-pro. It performs well across many datasets and its effectiveness is comparable or even better than more complex methods like VCD. Furthermore, our prompt design principles and experimental analyses provide valuable insights into multimodal hallucination.
comment: 9 pages, 13 tables, 4 figures
☆ IntGrad MT: Eliciting LLMs' Machine Translation Capabilities with Sentence Interpolation and Gradual MT
Recent Large Language Models (LLMs) have demonstrated strong performance in translation without needing to be finetuned on additional parallel corpora. However, they still underperform for low-resource language pairs. Previous works have focused on mitigating this issue by leveraging relevant few-shot examples or external resources such as dictionaries or grammar books, making models heavily reliant on these nonparametric sources of information. In this paper, we propose a novel method named IntGrad MT that focuses on fully exploiting an LLM's inherent translation capability. IntGrad MT achieves this by constructing a chain of few-shot examples, each consisting of a source sentence and the model's own translation, that rise incrementally in difficulty. IntGrad MT employs two techniques: Sentence Interpolation, which generates a sequence of sentences that gradually change from an easy sentence to translate to a difficult one, and Gradual MT, which sequentially translates this chain using translations of earlier sentences as few-shot examples for the translation of subsequent ones. With this approach, we observe a substantial enhancement in the xCOMET scores of various LLMs for multiple languages, especially in low-resource languages such as Hindi(8.26), Swahili(7.10), Bengali(6.97) and Marathi(13.03). Our approach presents a practical way of enhancing LLMs' performance without extra training.
☆ Are UFOs Driving Innovation? The Illusion of Causality in Large Language Models
Illusions of causality occur when people develop the belief that there is a causal connection between two variables with no supporting evidence. This cognitive bias has been proposed to underlie many societal problems including social prejudice, stereotype formation, misinformation and superstitious thinking. In this research we investigate whether large language models develop the illusion of causality in real-world settings. We evaluated and compared news headlines generated by GPT-4o-Mini, Claude-3.5-Sonnet, and Gemini-1.5-Pro to determine whether the models incorrectly framed correlations as causal relationships. In order to also measure sycophantic behavior, which occurs when a model aligns with a user's beliefs in order to look favorable even if it is not objectively correct, we additionally incorporated the bias into the prompts, observing if this manipulation increases the likelihood of the models exhibiting the illusion of causality. We found that Claude-3.5-Sonnet is the model that presents the lowest degree of causal illusion aligned with experiments on Correlation-to-Causation Exaggeration in human-written press releases. On the other hand, our findings suggest that while mimicry sycophancy increases the likelihood of causal illusions in these models, especially in GPT-4o-Mini, Claude-3.5-Sonnet remains the most robust against this cognitive bias.
☆ Understanding Likelihood Over-optimisation in Direct Alignment Algorithms
Direct Alignment Algorithms (DAAs), such as Direct Preference Optimisation (DPO) and Identity Preference Optimisation (IPO), have emerged as alternatives to online Reinforcement Learning from Human Feedback (RLHF) algorithms such as Proximal Policy Optimisation (PPO) for aligning language models to human preferences, without the need for explicit reward modelling. These methods generally aim to increase the likelihood of generating better (preferred) completions while discouraging worse (non-preferred) ones, while staying close to the original model's behaviour. In this work, we explore the relationship between completion likelihood and model performance in state-of-the-art DAAs, and identify a critical issue of likelihood over-optimisation. Contrary to expectations, we find that higher likelihood of better completions and larger margins between better and worse completion likelihoods do not necessarily lead to better performance, and may even degrade it. Our analysis reveals that while higher likelihood correlates with better memorisation of factual knowledge patterns, a slightly lower completion likelihood tends to improve output diversity, thus leading to better generalisation to unseen scenarios. Moreover, we identify two key indicators that signal when over-optimised output diversity begins to harm performance: Decreasing Entropy over Top-k Tokens and Diminishing Top-k Probability Mass. Our experimental results validate that these indicators are reliable signs of declining performance under different regularisations, helping prevent over-optimisation and improve alignment with human preferences.
comment: Preprint Version
☆ LLM-Mixer: Multiscale Mixing in LLMs for Time Series Forecasting
Time series forecasting remains a challenging task, particularly in the context of complex multiscale temporal patterns. This study presents LLM-Mixer, a framework that improves forecasting accuracy through the combination of multiscale time-series decomposition with pre-trained LLMs (Large Language Models). LLM-Mixer captures both short-term fluctuations and long-term trends by decomposing the data into multiple temporal resolutions and processing them with a frozen LLM, guided by a textual prompt specifically designed for time-series data. Extensive experiments conducted on multivariate and univariate datasets demonstrate that LLM-Mixer achieves competitive performance, outperforming recent state-of-the-art models across various forecasting horizons. This work highlights the potential of combining multiscale analysis and LLMs for effective and scalable time-series forecasting.
comment: Time series forecasting using LLMs
☆ Leaving the barn door open for Clever Hans: Simple features predict LLM benchmark answers
The integrity of AI benchmarks is fundamental to accurately assess the capabilities of AI systems. The internal validity of these benchmarks - i.e., making sure they are free from confounding factors - is crucial for ensuring that they are measuring what they are designed to measure. In this paper, we explore a key issue related to internal validity: the possibility that AI systems can solve benchmarks in unintended ways, bypassing the capability being tested. This phenomenon, widely known in human and animal experiments, is often referred to as the 'Clever Hans' effect, where tasks are solved using spurious cues, often involving much simpler processes than those putatively assessed. Previous research suggests that language models can exhibit this behaviour as well. In several older Natural Language Processing (NLP) benchmarks, individual $n$-grams like "not" have been found to be highly predictive of the correct labels, and supervised NLP models have been shown to exploit these patterns. In this work, we investigate the extent to which simple $n$-grams extracted from benchmark instances can be combined to predict labels in modern multiple-choice benchmarks designed for LLMs, and whether LLMs might be using such $n$-gram patterns to solve these benchmarks. We show how simple classifiers trained on these $n$-grams can achieve high scores on several benchmarks, despite lacking the capabilities being tested. Additionally, we provide evidence that modern LLMs might be using these superficial patterns to solve benchmarks. This suggests that the internal validity of these benchmarks may be compromised and caution should be exercised when interpreting LLM performance results on them.
☆ VisualRWKV-HD and UHD: Advancing High-Resolution Processing for Visual Language Models
Accurately understanding complex visual information is crucial for visual language models (VLMs). Enhancing image resolution can improve visual perception capabilities, not only reducing hallucinations but also boosting performance in tasks that demand high resolution, such as text-rich or document analysis. In this paper, we present VisualRWKV-HD and VisualRWKV-UHD, two advancements in the VisualRWKV model family, specifically designed to process high-resolution visual inputs. For VisualRWKV-HD, we developed a lossless downsampling method to effectively integrate a high-resolution vision encoder with low-resolution encoders, without extending the input sequence length. For the VisualRWKV-UHD model, we enhanced image representation by dividing the image into four segments, which are then recombined with the original image. This technique allows the model to incorporate both high-resolution and low-resolution features, effectively balancing coarse and fine-grained information. As a result, the model supports resolutions up to 4096 x 4096 pixels, offering a more detailed and comprehensive visual processing capability. Both VisualRWKV-HD and VisualRWKV-UHD not only achieve strong results on VLM benchmarks but also show marked improvements in performance for text-rich tasks.
☆ Eliciting Textual Descriptions from Representations of Continuous Prompts
Continuous prompts, or "soft prompts", are a widely-adopted parameter-efficient tuning strategy for large language models, but are often less favorable due to their opaque nature. Prior attempts to interpret continuous prompts relied on projecting individual prompt tokens onto the vocabulary space. However, this approach is problematic as performant prompts can yield arbitrary or contradictory text, and it interprets prompt tokens individually. In this work, we propose a new approach to interpret continuous prompts that elicits textual descriptions from their representations during model inference. Using a Patchscopes variant (Ghandeharioun et al., 2024) called InSPEcT over various tasks, we show our method often yields accurate task descriptions which become more faithful as task performance increases. Moreover, an elaborated version of InSPEcT reveals biased features in continuous prompts, whose presence correlates with biased model predictions. Providing an effective interpretability solution, InSPEcT can be leveraged to debug unwanted properties in continuous prompts and inform developers on ways to mitigate them.
☆ Unveiling the Mystery of Visual Attributes of Concrete and Abstract Concepts: Variability, Nearest Neighbors, and Challenging Categories
The visual representation of a concept varies significantly depending on its meaning and the context where it occurs; this poses multiple challenges both for vision and multimodal models. Our study focuses on concreteness, a well-researched lexical-semantic variable, using it as a case study to examine the variability in visual representations. We rely on images associated with approximately 1,000 abstract and concrete concepts extracted from two different datasets: Bing and YFCC. Our goals are: (i) evaluate whether visual diversity in the depiction of concepts can reliably distinguish between concrete and abstract concepts; (ii) analyze the variability of visual features across multiple images of the same concept through a nearest neighbor analysis; and (iii) identify challenging factors contributing to this variability by categorizing and annotating images. Our findings indicate that for classifying images of abstract versus concrete concepts, a combination of basic visual features such as color and texture is more effective than features extracted by more complex models like Vision Transformer (ViT). However, ViTs show better performances in the nearest neighbor analysis, emphasizing the need for a careful selection of visual features when analyzing conceptual variables through modalities other than text.
☆ Retrieval Augmented Spelling Correction for E-Commerce Applications
The rapid introduction of new brand names into everyday language poses a unique challenge for e-commerce spelling correction services, which must distinguish genuine misspellings from novel brand names that use unconventional spelling. We seek to address this challenge via Retrieval Augmented Generation (RAG). On this approach, product names are retrieved from a catalog and incorporated into the context used by a large language model (LLM) that has been fine-tuned to do contextual spelling correction. Through quantitative evaluation and qualitative error analyses, we find improvements in spelling correction utilizing the RAG framework beyond a stand-alone LLM. We also demonstrate the value of additional finetuning of the LLM to incorporate retrieved context.
Transformer Layer Injection: A Novel Approach for Efficient Upscaling of Large Language Models
In this paper, we propose Transformer Layer Injection (TLI), a novel method for efficiently upscaling large language models (LLMs) while minimizing computational costs and maintaining model performance. Model scale is a key factor in enhancing the quality of machine learning models, and TLI addresses the challenge of scaling by reducing initial loss, minimizing fine-tuning requirements, and preserving model complexity. Our approach improves upon the conventional Depth Up-Scaling (DUS) technique by injecting new layers into every set of K layers, enabling hidden representations to pass through transformer blocks with minimal disruption. We compare TLI with existing approaches, including Mixture of Experts (MoE) and DUS, and validate its efficiency through experiments on small LLMs (LLama3 1B, 3B, and 8B). Results show that TLI achieves better initialization, requires fewer training steps, and delivers superior accuracy on tasks such as KoBEST and KMCQA, with models performing effectively even without additional training. TLI is demonstrated to be both data-efficient and cost-effective, significantly outperforming existing methods. Its scalability and simplicity make it a promising solution for upscaling transformer-based models, with potential applications in scaling models from 10B to 405B parameters.
☆ Measuring Spiritual Values and Bias of Large Language Models
Large language models (LLMs) have become integral tool for users from various backgrounds. LLMs, trained on vast corpora, reflect the linguistic and cultural nuances embedded in their pre-training data. However, the values and perspectives inherent in this data can influence the behavior of LLMs, leading to potential biases. As a result, the use of LLMs in contexts involving spiritual or moral values necessitates careful consideration of these underlying biases. Our work starts with verification of our hypothesis by testing the spiritual values of popular LLMs. Experimental results show that LLMs' spiritual values are quite diverse, as opposed to the stereotype of atheists or secularists. We then investigate how different spiritual values affect LLMs in social-fairness scenarios e.g., hate speech identification). Our findings reveal that different spiritual values indeed lead to different sensitivity to different hate target groups. Furthermore, we propose to continue pre-training LLMs on spiritual texts, and empirical results demonstrate the effectiveness of this approach in mitigating spiritual bias.
comment: 9 pages including appendix; 5 figures; 5 tables; submitted to ARR - Octobor 2024
☆ Tokenization and Morphology in Multilingual Language Models: A~Comparative Analysis of mT5 and ByT5
Morphology is a crucial factor for multilingual language modeling as it poses direct challenges for tokenization. Here, we seek to understand how tokenization influences the morphological knowledge encoded in multilingual language models. Specifically, we capture the impact of tokenization by contrasting two multilingual language models: mT5 and ByT5. The two models share the same architecture, training objective, and training data and only differ in their tokenization strategies: subword tokenization vs. character-level tokenization. Probing the morphological knowledge encoded in these models on four tasks and 17 languages, our analyses show that multilingual language models learn the morphological systems of some languages better than others despite similar average performance and that morphological information is encoded in the middle and late layers, where characted-based models need a few more layers to yield commensurate probing accuracy. Finally, we show that languages with more irregularities benefit more from having a higher share of the pre-training data.
☆ Findings of the WMT 2024 Shared Task on Chat Translation
This paper presents the findings from the third edition of the Chat Translation Shared Task. As with previous editions, the task involved translating bilingual customer support conversations, specifically focusing on the impact of conversation context in translation quality and evaluation. We also include two new language pairs: English-Korean and English-Dutch, in addition to the set of language pairs from previous editions: English-German, English-French, and English-Brazilian Portuguese. We received 22 primary submissions and 32 contrastive submissions from eight teams, with each language pair having participation from at least three teams. We evaluated the systems comprehensively using both automatic metrics and human judgments via a direct assessment framework. The official rankings for each language pair were determined based on human evaluation scores, considering performance in both translation directions--agent and customer. Our analysis shows that while the systems excelled at translating individual turns, there is room for improvement in overall conversation-level translation quality.
comment: 12 pages, 5 figures, 13 tables
☆ VidEgoThink: Assessing Egocentric Video Understanding Capabilities for Embodied AI
Recent advancements in Multi-modal Large Language Models (MLLMs) have opened new avenues for applications in Embodied AI. Building on previous work, EgoThink, we introduce VidEgoThink, a comprehensive benchmark for evaluating egocentric video understanding capabilities. To bridge the gap between MLLMs and low-level control in Embodied AI, we design four key interrelated tasks: video question-answering, hierarchy planning, visual grounding and reward modeling. To minimize manual annotation costs, we develop an automatic data generation pipeline based on the Ego4D dataset, leveraging the prior knowledge and multimodal capabilities of GPT-4o. Three human annotators then filter the generated data to ensure diversity and quality, resulting in the VidEgoThink benchmark. We conduct extensive experiments with three types of models: API-based MLLMs, open-source image-based MLLMs, and open-source video-based MLLMs. Experimental results indicate that all MLLMs, including GPT-4o, perform poorly across all tasks related to egocentric video understanding. These findings suggest that foundation models still require significant advancements to be effectively applied to first-person scenarios in Embodied AI. In conclusion, VidEgoThink reflects a research trend towards employing MLLMs for egocentric vision, akin to human capabilities, enabling active observation and interaction in the complex real-world environments.
☆ MultiVENT 2.0: A Massive Multilingual Benchmark for Event-Centric Video Retrieval
Efficiently retrieving and synthesizing information from large-scale multimodal collections has become a critical challenge. However, existing video retrieval datasets suffer from scope limitations, primarily focusing on matching descriptive but vague queries with small collections of professionally edited, English-centric videos. To address this gap, we introduce $\textbf{MultiVENT 2.0}$, a large-scale, multilingual event-centric video retrieval benchmark featuring a collection of more than 218,000 news videos and 3,906 queries targeting specific world events. These queries specifically target information found in the visual content, audio, embedded text, and text metadata of the videos, requiring systems leverage all these sources to succeed at the task. Preliminary results show that state-of-the-art vision-language models struggle significantly with this task, and while alternative approaches show promise, they are still insufficient to adequately address this problem. These findings underscore the need for more robust multimodal retrieval systems, as effective video retrieval is a crucial step towards multimodal content understanding and generation tasks.
☆ Causal Reasoning in Large Language Models: A Knowledge Graph Approach NeurIPS 2024
Large language models (LLMs) typically improve performance by either retrieving semantically similar information, or enhancing reasoning abilities through structured prompts like chain-of-thought. While both strategies are considered crucial, it remains unclear which has a greater impact on model performance or whether a combination of both is necessary. This paper answers this question by proposing a knowledge graph (KG)-based random-walk reasoning approach that leverages causal relationships. We conduct experiments on the commonsense question answering task that is based on a KG. The KG inherently provides both relevant information, such as related entity keywords, and a reasoning structure through the connections between nodes. Experimental results show that the proposed KG-based random-walk reasoning method improves the reasoning ability and performance of LLMs. Interestingly, incorporating three seemingly irrelevant sentences into the query using KG-based random-walk reasoning enhances LLM performance, contrary to conventional wisdom. These findings suggest that integrating causal structures into prompts can significantly improve reasoning capabilities, providing new insights into the role of causality in optimizing LLM performance.
comment: Accepted at NeurIPS 2024 Workshop on Causality and Large Models (CaLM)
☆ Y-Mol: A Multiscale Biomedical Knowledge-Guided Large Language Model for Drug Development
Large Language Models (LLMs) have recently demonstrated remarkable performance in general tasks across various fields. However, their effectiveness within specific domains such as drug development remains challenges. To solve these challenges, we introduce \textbf{Y-Mol}, forming a well-established LLM paradigm for the flow of drug development. Y-Mol is a multiscale biomedical knowledge-guided LLM designed to accomplish tasks across lead compound discovery, pre-clinic, and clinic prediction. By integrating millions of multiscale biomedical knowledge and using LLaMA2 as the base LLM, Y-Mol augments the reasoning capability in the biomedical domain by learning from a corpus of publications, knowledge graphs, and expert-designed synthetic data. The capability is further enriched with three types of drug-oriented instructions: description-based prompts from processed publications, semantic-based prompts for extracting associations from knowledge graphs, and template-based prompts for understanding expert knowledge from biomedical tools. Besides, Y-Mol offers a set of LLM paradigms that can autonomously execute the downstream tasks across the entire process of drug development, including virtual screening, drug design, pharmacological properties prediction, and drug-related interaction prediction. Our extensive evaluations of various biomedical sources demonstrate that Y-Mol significantly outperforms general-purpose LLMs in discovering lead compounds, predicting molecular properties, and identifying drug interaction events.
comment: 12 pages, Under Review
☆ Multi-round jailbreak attack on large language models
Ensuring the safety and alignment of large language models (LLMs) with human values is crucial for generating responses that are beneficial to humanity. While LLMs have the capability to identify and avoid harmful queries, they remain vulnerable to "jailbreak" attacks, where carefully crafted prompts can induce the generation of toxic content. Traditional single-round jailbreak attacks, such as GCG and AutoDAN, do not alter the sensitive words in the dangerous prompts. Although they can temporarily bypass the model's safeguards through prompt engineering, their success rate drops significantly as the LLM is further fine-tuned, and they cannot effectively circumvent static rule-based filters that remove the hazardous vocabulary. In this study, to better understand jailbreak attacks, we introduce a multi-round jailbreak approach. This method can rewrite the dangerous prompts, decomposing them into a series of less harmful sub-questions to bypass the LLM's safety checks. We first use the LLM to perform a decomposition task, breaking down a set of natural language questions into a sequence of progressive sub-questions, which are then used to fine-tune the Llama3-8B model, enabling it to decompose hazardous prompts. The fine-tuned model is then used to break down the problematic prompt, and the resulting sub-questions are sequentially asked to the victim model. If the victim model rejects a sub-question, a new decomposition is generated, and the process is repeated until the final objective is achieved. Our experimental results show a 94\% success rate on the llama2-7B and demonstrate the effectiveness of this approach in circumventing static rule-based filters.
☆ Human-LLM Collaborative Construction of a Cantonese Emotion Lexicon
Large Language Models (LLMs) have demonstrated remarkable capabilities in language understanding and generation. Advanced utilization of the knowledge embedded in LLMs for automated annotation has consistently been explored. This study proposed to develop an emotion lexicon for Cantonese, a low-resource language, through collaborative efforts between LLM and human annotators. By integrating emotion labels provided by LLM and human annotators, the study leveraged existing linguistic resources including lexicons in other languages and local forums to construct a Cantonese emotion lexicon enriched with colloquial expressions. The consistency of the proposed emotion lexicon in emotion extraction was assessed through modification and utilization of three distinct emotion text datasets. This study not only validates the efficacy of the constructed lexicon but also emphasizes that collaborative annotation between human and artificial intelligence can significantly enhance the quality of emotion labels, highlighting the potential of such partnerships in facilitating natural language processing tasks for low-resource languages.
comment: 13 pages
☆ TopoLM: brain-like spatio-functional organization in a topographic language model
Neurons in the brain are spatially organized such that neighbors on tissue often exhibit similar response profiles. In the human language system, experimental studies have observed clusters for syntactic and semantic categories, but the mechanisms underlying this functional organization remain unclear. Here, building on work from the vision literature, we develop TopoLM, a transformer language model with an explicit two-dimensional spatial representation of model units. By combining a next-token prediction objective with a spatial smoothness loss, representations in this model assemble into clusters that correspond to semantically interpretable groupings of text and closely match the functional organization in the brain's language system. TopoLM successfully predicts the emergence of the spatio-functional organization of a cortical language system as well as the organization of functional clusters selective for fine-grained linguistic features empirically observed in human cortex. Our results suggest that the functional organization of the human language system is driven by a unified spatial objective, and provide a functionally and spatially aligned model of language processing in the brain.
☆ Revisiting Benchmark and Assessment: An Agent-based Exploratory Dynamic Evaluation Framework for LLMs
While various vertical domain large language models (LLMs) have been developed, the challenge of automatically evaluating their performance across different domains remains significant in addressing real-world user needs. Current benchmark-based evaluation methods exhibit rigid, purposeless interactions and rely on pre-collected static datasets that are costly to build, inflexible across domains, and misaligned with practical user needs. To address this, we revisit the evaluation components and introduce two definitions: **Benchmark+**, which extends traditional QA benchmarks into a more flexible ``strategy-criterion'' format; and **Assessment+**, which enhances the interaction process for greater exploration and enables both quantitative metrics and qualitative insights that capture nuanced target LLM behaviors from richer multi-turn interactions. We propose an agent-based evaluation framework called *TestAgent*, which implements these two concepts through retrieval augmented generation and reinforcement learning. Experiments on tasks ranging from building vertical domain evaluation from scratch to activating existing benchmarks demonstrate the effectiveness of *TestAgent* across various scenarios. We believe this work offers an interesting perspective on automatic evaluation for LLMs.
☆ DynamicER: Resolving Emerging Mentions to Dynamic Entities for RAG EMNLP 2024
In the rapidly evolving landscape of language, resolving new linguistic expressions in continuously updating knowledge bases remains a formidable challenge. This challenge becomes critical in retrieval-augmented generation (RAG) with knowledge bases, as emerging expressions hinder the retrieval of relevant documents, leading to generator hallucinations. To address this issue, we introduce a novel task aimed at resolving emerging mentions to dynamic entities and present DynamicER benchmark. Our benchmark includes dynamic entity mention resolution and entity-centric knowledge-intensive QA task, evaluating entity linking and RAG model's adaptability to new expressions, respectively. We discovered that current entity linking models struggle to link these new expressions to entities. Therefore, we propose a temporal segmented clustering method with continual adaptation, effectively managing the temporal dynamics of evolving entities and emerging mentions. Extensive experiments demonstrate that our method outperforms existing baselines, enhancing RAG model performance on QA task with resolved mentions.
comment: EMNLP 2024 Main
☆ O-Edit: Orthogonal Subspace Editing for Language Model Sequential Editing
Large language models (LLMs) acquire knowledge during pre-training, but over time, this knowledge may become incorrect or outdated, necessitating updates after training. Knowledge editing techniques address this issue without the need for costly re-training. However, most existing methods are designed for single edits, and as the number of edits increases, they often cause a decline in the model's overall performance, posing significant challenges for sequential editing. To overcome this, we propose Orthogonal Subspace Editing, O-Edit. This algorithm orthogonalizes the direction of each knowledge update, minimizing interference between successive updates and reducing the impact of new updates on unrelated knowledge. Our approach does not require replaying previously edited data and processes each edit knowledge on time. It can perform thousands of edits on mainstream LLMs, achieving an average performance improvement that is 4.2 times better than existing methods while effectively preserving the model's performance on downstream tasks, all with minimal additional parameter overhead.
☆ Mitigating Frequency Bias and Anisotropy in Language Model Pre-Training with Syntactic Smoothing
Language models strongly rely on frequency information because they maximize the likelihood of tokens during pre-training. As a consequence, language models tend to not generalize well to tokens that are seldom seen during training. Moreover, maximum likelihood training has been discovered to give rise to anisotropy: representations of tokens in a model tend to cluster tightly in a high-dimensional cone, rather than spreading out over their representational capacity. Our work introduces a method for quantifying the frequency bias of a language model by assessing sentence-level perplexity with respect to token-level frequency. We then present a method for reducing the frequency bias of a language model by inducing a syntactic prior over token representations during pre-training. Our Syntactic Smoothing method adjusts the maximum likelihood objective function to distribute the learning signal to syntactically similar tokens. This approach results in better performance on infrequent English tokens and a decrease in anisotropy. We empirically show that the degree of anisotropy in a model correlates with its frequency bias.
☆ Jigsaw Puzzles: Splitting Harmful Questions to Jailbreak Large Language Models
Large language models (LLMs) have exhibited outstanding performance in engaging with humans and addressing complex questions by leveraging their vast implicit knowledge and robust reasoning capabilities. However, such models are vulnerable to jailbreak attacks, leading to the generation of harmful responses. Despite recent research on single-turn jailbreak strategies to facilitate the development of defence mechanisms, the challenge of revealing vulnerabilities under multi-turn setting remains relatively under-explored. In this work, we propose Jigsaw Puzzles (JSP), a straightforward yet effective multi-turn jailbreak strategy against the advanced LLMs. JSP splits questions into harmless fractions as the input of each turn, and requests LLMs to reconstruct and respond to questions under multi-turn interaction. Our experimental results demonstrate that the proposed JSP jailbreak bypasses original safeguards against explicitly harmful content, achieving an average attack success rate of 93.76% on 189 harmful queries across 5 advanced LLMs (Gemini-1.5-Pro, Llama-3.1-70B, GPT-4, GPT-4o, GPT-4o-mini). Moreover, JSP achieves a state-of-the-art attack success rate of 92% on GPT-4 on the harmful query benchmark, and exhibits strong resistant to defence strategies. Warning: this paper contains offensive examples.
☆ LR-SQL: A Supervised Fine-Tuning Method for Text2SQL Tasks under Low-Resource Scenarios
Large language models revolutionize Text2SQL through supervised fine-tuning, yet a crucial limitation is overlooked: the complexity of databases leads to an increased context length, consequently resulting in higher GPU memory demands for model fine-tuning. To address this issue, we propose LR-SQL. LR-SQL comprises two supervised fine-tuning models: the schema\_link model and the SQL\_generation model, with the schema\_link model serving as the focal point for streamlining the overall process. During the fine-tuning of the schema\_link model, LR-SQL breaks down the complete database into flexible combinations of tables with adjustable quantities, enabling the model to learn the relationships within the entire database from these dispersed slices. Furthermore, to enhance the model's ability to perceive the relationships among various discrete slices during inference, LR-SQL trains the model's Chain-of-Thought capability for this task. Experimental results demonstrate that LR-SQL can reduce the total GPU memory usage by 40\% compared to existing fine-tuning methods, while only losing 2\% of table prediction accuracy in schema\_link task. For the overall Text2SQL task, the Execution Accuracy decrease by 0.6\%.Our project is now available on https://github.com/hongWin/LR-SQL
comment: 12pages, 4 figures,submitting to a journal
☆ Tending Towards Stability: Convergence Challenges in Small Language Models
Increasing the number of parameters in language models is a common strategy to enhance their performance. However, smaller language models remain valuable due to their lower operational costs. Despite their advantages, smaller models frequently underperform compared to their larger counterparts, even when provided with equivalent data and computational resources. Specifically, their performance tends to degrade in the late pretraining phase. This is anecdotally attributed to their reduced representational capacity. Yet, the exact causes of this performance degradation remain unclear. We use the Pythia model suite to analyse the training dynamics that underlie this phenomenon. Across different model sizes, we investigate the convergence of the Attention and MLP activations to their final state and examine how the effective rank of their parameters influences this process. We find that nearly all layers in larger models stabilise early in training - within the first 20% - whereas layers in smaller models exhibit slower and less stable convergence, especially when their parameters have lower effective rank. By linking the convergence of layers' activations to their parameters' effective rank, our analyses can guide future work to address inefficiencies in the learning dynamics of small models.
☆ A Cross-Lingual Statutory Article Retrieval Dataset for Taiwan Legal Studies
This paper introduces a cross-lingual statutory article retrieval (SAR) dataset designed to enhance legal information retrieval in multilingual settings. Our dataset features spoken-language-style legal inquiries in English, paired with corresponding Chinese versions and relevant statutes, covering all Taiwanese civil, criminal, and administrative laws. This dataset aims to improve access to legal information for non-native speakers, particularly for foreign nationals in Taiwan. We propose several LLM-based methods as baselines for evaluating retrieval effectiveness, focusing on mitigating translation errors and improving cross-lingual retrieval performance. Our work provides a valuable resource for developing inclusive legal information retrieval systems.
☆ AIC CTU system at AVeriTeC: Re-framing automated fact-checking as a simple RAG task
This paper describes our $3^{rd}$ place submission in the AVeriTeC shared task in which we attempted to address the challenge of fact-checking with evidence retrieved in the wild using a simple scheme of Retrieval-Augmented Generation (RAG) designed for the task, leveraging the predictive power of Large Language Models. We release our codebase and explain its two modules - the Retriever and the Evidence & Label generator - in detail, justifying their features such as MMR-reranking and Likert-scale confidence estimation. We evaluate our solution on AVeriTeC dev and test set and interpret the results, picking the GPT-4o as the most appropriate model for our pipeline at the time of our publication, with Llama 3.1 70B being a promising open-source alternative. We perform an empirical error analysis to see that faults in our predictions often coincide with noise in the data or ambiguous fact-checks, provoking further research and data augmentation.
☆ Difficult Task Yes but Simple Task No: Unveiling the Laziness in Multimodal LLMs EMNLP 2024
Multimodal Large Language Models (MLLMs) demonstrate a strong understanding of the real world and can even handle complex tasks. However, they still fail on some straightforward visual question-answering (VQA) problems. This paper dives deeper into this issue, revealing that models tend to err when answering easy questions (e.g. Yes/No questions) about an image, even though they can correctly describe it. We refer to this model behavior discrepancy between difficult and simple questions as model laziness. To systematically investigate model laziness, we manually construct LazyBench, a benchmark that includes Yes/No, multiple choice, short answer questions, and image description tasks that are related to the same subjects in the images. Based on LazyBench, we observe that laziness widely exists in current advanced MLLMs (e.g. GPT-4o, Gemini-1.5-pro, Claude 3 and LLaVA-v1.5-13B), and it is more pronounced on stronger models. We also analyze the VQA v2 (LLaVA-v1.5-13B) benchmark and find that about half of its failure cases are caused by model laziness, which further highlights the importance of ensuring that the model fully utilizes its capability. To this end, we conduct preliminary exploration on how to mitigate laziness and find that chain of thought (CoT) can effectively address this issue.
comment: EMNLP 2024 Findings
☆ Titanic Calling: Low Bandwidth Video Conference from the Titanic Wreck
In this paper, we report on communication experiments conducted in the summer of 2022 during a deep dive to the wreck of the Titanic. Radio transmission is not possible in deep sea water, and communication links rely on sonar signals. Due to the low bandwidth of sonar signals and the need to communicate readable data, text messaging is used in deep-sea missions. In this paper, we report results and experiences from a messaging system that converts speech to text in a submarine, sends text messages to the surface, and reconstructs those messages as synthetic lip-synchronous videos of the speakers. The resulting system was tested during an actual dive to Titanic in the summer of 2022. We achieved an acceptable latency for a system of such complexity as well as good quality. The system demonstration video can be found at the following link: https://youtu.be/C4lyM86-5Ig
☆ ReDeEP: Detecting Hallucination in Retrieval-Augmented Generation via Mechanistic Interpretability
Retrieval-Augmented Generation (RAG) models are designed to incorporate external knowledge, reducing hallucinations caused by insufficient parametric (internal) knowledge. However, even with accurate and relevant retrieved content, RAG models can still produce hallucinations by generating outputs that conflict with the retrieved information. Detecting such hallucinations requires disentangling how Large Language Models (LLMs) utilize external and parametric knowledge. Current detection methods often focus on one of these mechanisms or without decoupling their intertwined effects, making accurate detection difficult. In this paper, we investigate the internal mechanisms behind hallucinations in RAG scenarios. We discover hallucinations occur when the Knowledge FFNs in LLMs overemphasize parametric knowledge in the residual stream, while Copying Heads fail to effectively retain or integrate external knowledge from retrieved content. Based on these findings, we propose ReDeEP, a novel method that detects hallucinations by decoupling LLM's utilization of external context and parametric knowledge. Our experiments show that ReDeEP significantly improves RAG hallucination detection accuracy. Additionally, we introduce AARF, which mitigates hallucinations by modulating the contributions of Knowledge FFNs and Copying Heads.
comment: 23pages
☆ PMMT: Preference Alignment in Multilingual Machine Translation via LLM Distillation
Translation is important for cross-language communication, and many efforts have been made to improve its accuracy. However, less investment is conducted in aligning translations with human preferences, such as translation tones or styles. In this paper, a new method is proposed to effectively generate large-scale multilingual parallel corpora with specific translation preferences using Large Language Models (LLMs). Meanwhile, an automatic pipeline is designed to distill human preferences into smaller Machine Translation (MT) models for efficiently and economically supporting large-scale calls in online services. Experiments indicate that the proposed method takes the lead in translation tasks with aligned human preferences by a large margin. Meanwhile, on popular public benchmarks like WMT and Flores, on which our models were not trained, the proposed method also shows a competitive performance compared to SOTA works.
☆ Do LLMs Have the Generalization Ability in Conducting Causal Inference?
In causal inference, generalization capability refers to the ability to conduct causal inference methods on new data to estimate the causal-effect between unknown phenomenon, which is crucial for expanding the boundaries of knowledge. Studies have evaluated the causal inference capabilities of Large Language Models (LLMs) concerning known phenomena, yet the generalization capabilities of LLMs concerning unseen phenomena remain unexplored. In this paper, we selected four tasks: Causal Path Discovery (CP), Backdoor Adjustment (BA), Factual Inference (FI), and Counterfactual Inference (CI) as representatives of causal inference tasks. To generate evaluation questions about previously unseen phenomena in new data on the four tasks, we propose a benchmark generation framework, which employs randomly generated graphs and node names to formulate questions within hypothetical new causal scenarios. Based on this framework, we compile a benchmark dataset of varying levels of question complexity. We extensively tested the generalization capabilities of five leading LLMs across four tasks. Experiment results reveal that while LLMs exhibit good generalization performance in solving simple CP, FI, and complex CI questions, they encounter difficulties when tackling BA questions and face obvious performance fluctuations as the problem complexity changes. Furthermore, when the names of phenomena incorporate existing terms, even if these names are entirely novel, their generalization performance can still be hindered by interference from familiar terms.
☆ A Framework for Adapting Human-Robot Interaction to Diverse User Groups
To facilitate natural and intuitive interactions with diverse user groups in real-world settings, social robots must be capable of addressing the varying requirements and expectations of these groups while adapting their behavior based on user feedback. While previous research often focuses on specific demographics, we present a novel framework for adaptive Human-Robot Interaction (HRI) that tailors interactions to different user groups and enables individual users to modulate interactions through both minor and major interruptions. Our primary contributions include the development of an adaptive, ROS-based HRI framework with an open-source code base. This framework supports natural interactions through advanced speech recognition and voice activity detection, and leverages a large language model (LLM) as a dialogue bridge. We validate the efficiency of our framework through module tests and system trials, demonstrating its high accuracy in age recognition and its robustness to repeated user inputs and plan changes.
comment: Accepted at the 16th International Conference on Social Robotics (ICSR) 2024
☆ Learning from Imperfect Data: Towards Efficient Knowledge Distillation of Autoregressive Language Models for Text-to-SQL EMNLP2024
Large Language Models (LLMs) have shown promising performance in text-to-SQL, which involves translating natural language questions into SQL queries. However, current text-to-SQL LLMs are computationally expensive and challenging to deploy in real-world applications, highlighting the importance of compressing them. To achieve this goal, knowledge distillation (KD) is a common approach, which aims to distill the larger teacher model into a smaller student model. While numerous KD methods for autoregressive LLMs have emerged recently, it is still under-explored whether they work well in complex text-to-SQL scenarios. To this end, we conduct a series of analyses and reveal that these KD methods generally fall short in balancing performance and efficiency. In response to this problem, we propose to improve the KD with Imperfect Data, namely KID, which effectively boosts the performance without introducing much training budget. The core of KID is to efficiently mitigate the training-inference mismatch by simulating the cascading effect of inference in the imperfect training data. Extensive experiments on 5 text-to-SQL benchmarks show that, KID can not only achieve consistent and significant performance gains (up to +5.83% average score) across all model types and sizes, but also effectively improve the training efficiency.
comment: Accepted to EMNLP2024 Findings
☆ Enhance Graph Alignment for Large Language Models
Graph-structured data is prevalent in the real world. Recently, due to the powerful emergent capabilities, Large Language Models (LLMs) have shown promising performance in modeling graphs. The key to effectively applying LLMs on graphs is converting graph data into a format LLMs can comprehend. Graph-to-token approaches are popular in enabling LLMs to process graph information. They transform graphs into sequences of tokens and align them with text tokens through instruction tuning, where self-supervised instruction tuning helps LLMs acquire general knowledge about graphs, and supervised fine-tuning specializes LLMs for the downstream tasks on graphs. Despite their initial success, we find that existing methods have a misalignment between self-supervised tasks and supervised downstream tasks, resulting in negative transfer from self-supervised fine-tuning to downstream tasks. To address these issues, we propose Graph Alignment Large Language Models (GALLM) to benefit from aligned task templates. In the self-supervised tuning stage, we introduce a novel text matching task using templates aligned with downstream tasks. In the task-specific tuning stage, we propose two category prompt methods that learn supervision information from additional explanation with further aligned templates. Experimental evaluations on four datasets demonstrate substantial improvements in supervised learning, multi-dataset generalizability, and particularly in zero-shot capability, highlighting the model's potential as a graph foundation model.
comment: Under review
☆ LargePiG: Your Large Language Model is Secretly a Pointer Generator
Recent research on query generation has focused on using Large Language Models (LLMs), which despite bringing state-of-the-art performance, also introduce issues with hallucinations in the generated queries. In this work, we introduce relevance hallucination and factuality hallucination as a new typology for hallucination problems brought by query generation based on LLMs. We propose an effective way to separate content from form in LLM-generated queries, which preserves the factual knowledge extracted and integrated from the inputs and compiles the syntactic structure, including function words, using the powerful linguistic capabilities of the LLM. Specifically, we introduce a model-agnostic and training-free method that turns the Large Language Model into a Pointer-Generator (LargePiG), where the pointer attention distribution leverages the LLM's inherent attention weights, and the copy probability is derived from the difference between the vocabulary distribution of the model's high layers and the last layer. To validate the effectiveness of LargePiG, we constructed two datasets for assessing the hallucination problems in query generation, covering both document and video scenarios. Empirical studies on various LLMs demonstrated the superiority of LargePiG on both datasets. Additional experiments also verified that LargePiG could reduce hallucination in large vision language models and improve the accuracy of document-based question-answering and factuality evaluation tasks.
comment: 24 pages
☆ Reducing Labeling Costs in Sentiment Analysis via Semi-Supervised Learning
Labeling datasets is a noteworthy challenge in machine learning, both in terms of cost and time. This research, however, leverages an efficient answer. By exploring label propagation in semi-supervised learning, we can significantly reduce the number of labels required compared to traditional methods. We employ a transductive label propagation method based on the manifold assumption for text classification. Our approach utilizes a graph-based method to generate pseudo-labels for unlabeled data for the text classification task, which are then used to train deep neural networks. By extending labels based on cosine proximity within a nearest neighbor graph from network embeddings, we combine unlabeled data into supervised learning, thereby reducing labeling costs. Based on previous successes in other domains, this study builds and evaluates this approach's effectiveness in sentiment analysis, presenting insights into semi-supervised learning.
comment: 12 pages, 7 figures, accepted at the 2024 8th International Conference on Natural Language Processing and Information Retrieval (NLPIR 2024), Okayama, Japan, 2024
☆ RATE: Score Reward Models with Imperfect Rewrites of Rewrites ICLR 2025
This paper concerns the evaluation of reward models used in language modeling. A reward model is a function that takes a prompt and a response and assigns a score indicating how good that response is for the prompt. A key challenge is that reward models are usually imperfect proxies for actual preferences. For example, we may worry that a model trained to reward helpfulness learns to instead prefer longer responses. In this paper, we develop an evaluation method, RATE (Rewrite-based Attribute Treatment Estimators), that allows us to measure the causal effect of a given attribute of a response (e.g., length) on the reward assigned to that response. The core idea is to use large language models to rewrite responses to produce imperfect counterfactuals, and to adjust for rewriting error by rewriting twice. We show that the RATE estimator is consistent under reasonable assumptions. We demonstrate the effectiveness of RATE on synthetic and real-world data, showing that it can accurately estimate the effect of a given attribute on the reward model.
comment: Submitted as a conference paper to ICLR 2025. Code is available at https://github.com/toddnief/RATE
☆ SHAKTI: A 2.5 Billion Parameter Small Language Model Optimized for Edge AI and Low-Resource Environments
We introduce Shakti, a 2.5 billion parameter language model specifically optimized for resource-constrained environments such as edge devices, including smartphones, wearables, and IoT systems. Shakti combines high-performance NLP with optimized efficiency and precision, making it ideal for real-time AI applications where computational resources and memory are limited. With support for vernacular languages and domain-specific tasks, Shakti excels in industries such as healthcare, finance, and customer service. Benchmark evaluations demonstrate that Shakti performs competitively against larger models while maintaining low latency and on-device efficiency, positioning it as a leading solution for edge AI.
comment: Paper in pdf format is 11 pages and contains 4 tables
☆ Sequential LLM Framework for Fashion Recommendation
The fashion industry is one of the leading domains in the global e-commerce sector, prompting major online retailers to employ recommendation systems for product suggestions and customer convenience. While recommendation systems have been widely studied, most are designed for general e-commerce problems and struggle with the unique challenges of the fashion domain. To address these issues, we propose a sequential fashion recommendation framework that leverages a pre-trained large language model (LLM) enhanced with recommendation-specific prompts. Our framework employs parameter-efficient fine-tuning with extensive fashion data and introduces a novel mix-up-based retrieval technique for translating text into relevant product suggestions. Extensive experiments show our proposed framework significantly enhances fashion recommendation performance.
☆ Speculative Knowledge Distillation: Bridging the Teacher-Student Gap Through Interleaved Sampling
Recent advances in knowledge distillation (KD) have enabled smaller student models to approach the performance of larger teacher models. However, popular methods such as supervised KD and on-policy KD, are adversely impacted by the knowledge gaps between teacher-student in practical scenarios. Supervised KD suffers from a distribution mismatch between training with a static dataset and inference over final student-generated outputs. Conversely, on-policy KD, which uses student-generated samples for training, can suffer from low-quality training examples with which teacher models are not familiar, resulting in inaccurate teacher feedback. To address these limitations, we introduce Speculative Knowledge Distillation (SKD), a novel approach that leverages cooperation between student and teacher models to generate high-quality training data on-the-fly while aligning with the student's inference-time distribution. In SKD, the student proposes tokens, and the teacher replaces poorly ranked ones based on its own distribution, transferring high-quality knowledge adaptively. We evaluate SKD on various text generation tasks, including translation, summarization, math, and instruction following, and show that SKD consistently outperforms existing KD methods across different domains, data sizes, and model initialization strategies.
☆ Self-adaptive Multimodal Retrieval-Augmented Generation
Traditional Retrieval-Augmented Generation (RAG) methods are limited by their reliance on a fixed number of retrieved documents, often resulting in incomplete or noisy information that undermines task performance. Although recent adaptive approaches alleviated these problems, their application in intricate and real-world multimodal tasks remains limited. To address these, we propose a new approach called Self-adaptive Multimodal Retrieval-Augmented Generation (SAM-RAG), tailored specifically for multimodal contexts. SAM-RAG not only dynamically filters relevant documents based on the input query, including image captions when needed, but also verifies the quality of both the retrieved documents and the output. Extensive experimental results show that SAM-RAG surpasses existing state-of-the-art methods in both retrieval accuracy and response generation. By further ablation experiments and effectiveness analysis, SAM-RAG maintains high recall quality while improving overall task performance in multimodal RAG task. Our codes are available at https://github.com/SAM-RAG/SAM_RAG.
☆ Deciphering the Chaos: Enhancing Jailbreak Attacks via Adversarial Prompt Translation
Automatic adversarial prompt generation provides remarkable success in jailbreaking safely-aligned large language models (LLMs). Existing gradient-based attacks, while demonstrating outstanding performance in jailbreaking white-box LLMs, often generate garbled adversarial prompts with chaotic appearance. These adversarial prompts are difficult to transfer to other LLMs, hindering their performance in attacking unknown victim models. In this paper, for the first time, we delve into the semantic meaning embedded in garbled adversarial prompts and propose a novel method that "translates" them into coherent and human-readable natural language adversarial prompts. In this way, we can effectively uncover the semantic information that triggers vulnerabilities of the model and unambiguously transfer it to the victim model, without overlooking the adversarial information hidden in the garbled text, to enhance jailbreak attacks. It also offers a new approach to discovering effective designs for jailbreak prompts, advancing the understanding of jailbreak attacks. Experimental results demonstrate that our method significantly improves the success rate of jailbreak attacks against various safety-aligned LLMs and outperforms state-of-the-arts by large margins. With at most 10 queries, our method achieves an average attack success rate of 81.8% in attacking 7 commercial closed-source LLMs, including GPT and Claude-3 series, on HarmBench. Our method also achieves over 90% attack success rates against Llama-2-Chat models on AdvBench, despite their outstanding resistance to jailbreak attacks. Code at: https://github.com/qizhangli/Adversarial-Prompt-Translator.
☆ SEER: Self-Aligned Evidence Extraction for Retrieval-Augmented Generation EMNLP 2024
Recent studies in Retrieval-Augmented Generation (RAG) have investigated extracting evidence from retrieved passages to reduce computational costs and enhance the final RAG performance, yet it remains challenging. Existing methods heavily rely on heuristic-based augmentation, encountering several issues: (1) Poor generalization due to hand-crafted context filtering; (2) Semantics deficiency due to rule-based context chunking; (3) Skewed length due to sentence-wise filter learning. To address these issues, we propose a model-based evidence extraction learning framework, SEER, optimizing a vanilla model as an evidence extractor with desired properties through self-aligned learning. Extensive experiments show that our method largely improves the final RAG performance, enhances the faithfulness, helpfulness, and conciseness of the extracted evidence, and reduces the evidence length by 9.25 times. The code will be available at https://github.com/HITsz-TMG/SEER.
comment: 15 pages, 6 figures, 5 tables. Accepted by EMNLP 2024 (main)
☆ Data Selection for Task-Specific Model Finetuning
Finetuning foundation models for specific tasks is an emerging paradigm in modern machine learning. The efficacy of task-specific finetuning largely depends on the selection of appropriate training data. We present a framework to select data for task-specific model finetuning, guided by a small but representative set of examples from the target task. To do so, we formulate data selection for task-specific finetuning as an optimization problem with a distribution alignment loss based on optimal transport to capture the discrepancy between the selected data and the target distribution. In addition, we add a regularizer to encourage the diversity of the selected data and incorporate kernel density estimation into the regularizer to reduce the negative effects of near-duplicates among the candidate data. We connect our optimization problem to nearest neighbor search and design efficient algorithms to compute the optimal solution based on approximate nearest neighbor search techniques. We evaluate our method on data selection for both continued pretraining and instruction tuning of language models. We show that instruction tuning using data selected by our method with a 1% selection ratio often outperforms using the full dataset and beats the baseline selection methods by 1.5 points in F1 score on average.
comment: 31 pages, 1 figure
☆ Have the VLMs Lost Confidence? A Study of Sycophancy in VLMs
In the study of LLMs, sycophancy represents a prevalent hallucination that poses significant challenges to these models. Specifically, LLMs often fail to adhere to original correct responses, instead blindly agreeing with users' opinions, even when those opinions are incorrect or malicious. However, research on sycophancy in visual language models (VLMs) has been scarce. In this work, we extend the exploration of sycophancy from LLMs to VLMs, introducing the MM-SY benchmark to evaluate this phenomenon. We present evaluation results from multiple representative models, addressing the gap in sycophancy research for VLMs. To mitigate sycophancy, we propose a synthetic dataset for training and employ methods based on prompts, supervised fine-tuning, and DPO. Our experiments demonstrate that these methods effectively alleviate sycophancy in VLMs. Additionally, we probe VLMs to assess the semantic impact of sycophancy and analyze the attention distribution of visual tokens. Our findings indicate that the ability to prevent sycophancy is predominantly observed in higher layers of the model. The lack of attention to image knowledge in these higher layers may contribute to sycophancy, and enhancing image attention at high layers proves beneficial in mitigating this issue.
☆ Enhancing Assamese NLP Capabilities: Introducing a Centralized Dataset Repository
This paper introduces a centralized, open-source dataset repository designed to advance NLP and NMT for Assamese, a low-resource language. The repository supports various tasks like sentiment analysis, named entity recognition, and machine translation by providing both pre-training and fine-tuning corpora. We review existing datasets, highlighting the need for standardized resources in Assamese NLP, and discuss potential applications in AI-driven research, such as LLMs, OCR, and chatbots. While promising, challenges like data scarcity and linguistic diversity remain. The repository aims to foster collaboration and innovation, promoting Assamese language research in the digital age.
comment: 6 pages, 1 table, 1 figure
☆ Process Reward Model with Q-Value Rankings
Process Reward Modeling (PRM) is critical for complex reasoning and decision-making tasks where the accuracy of intermediate steps significantly influences the overall outcome. Existing PRM approaches, primarily framed as classification problems, employ cross-entropy loss to independently evaluate each step's correctness. This method can lead to suboptimal reward distribution and does not adequately address the interdependencies among steps. To address these limitations, we introduce the Process Q-value Model (PQM), a novel framework that redefines PRM in the context of a Markov Decision Process. PQM optimizes Q-value rankings based on a novel comparative loss function, enhancing the model's ability to capture the intricate dynamics among sequential decisions. This approach provides a more granular and theoretically grounded methodology for process rewards. Our extensive empirical evaluations across various sampling policies, language model backbones, and multi-step reasoning benchmarks show that PQM outperforms classification-based PRMs. The effectiveness of the comparative loss function is highlighted in our comprehensive ablation studies, confirming PQM's practical efficacy and theoretical advantage.
☆ Cognitive Overload Attack:Prompt Injection for Long Context
Large Language Models (LLMs) have demonstrated remarkable capabilities in performing tasks across various domains without needing explicit retraining. This capability, known as In-Context Learning (ICL), while impressive, exposes LLMs to a variety of adversarial prompts and jailbreaks that manipulate safety-trained LLMs into generating undesired or harmful output. In this paper, we propose a novel interpretation of ICL in LLMs through the lens of cognitive neuroscience, by drawing parallels between learning in human cognition with ICL. We applied the principles of Cognitive Load Theory in LLMs and empirically validate that similar to human cognition, LLMs also suffer from cognitive overload a state where the demand on cognitive processing exceeds the available capacity of the model, leading to potential errors. Furthermore, we demonstrated how an attacker can exploit ICL to jailbreak LLMs through deliberately designed prompts that induce cognitive overload on LLMs, thereby compromising the safety mechanisms of LLMs. We empirically validate this threat model by crafting various cognitive overload prompts and show that advanced models such as GPT-4, Claude-3.5 Sonnet, Claude-3 OPUS, Llama-3-70B-Instruct, Gemini-1.0-Pro, and Gemini-1.5-Pro can be successfully jailbroken, with attack success rates of up to 99.99%. Our findings highlight critical vulnerabilities in LLMs and underscore the urgency of developing robust safeguards. We propose integrating insights from cognitive load theory into the design and evaluation of LLMs to better anticipate and mitigate the risks of adversarial attacks. By expanding our experiments to encompass a broader range of models and by highlighting vulnerabilities in LLMs' ICL, we aim to ensure the development of safer and more reliable AI systems.
comment: 40 pages, 31 Figures
☆ In-Context Learning for Long-Context Sentiment Analysis on Infrastructure Project Opinions
Large language models (LLMs) have achieved impressive results across various tasks. However, they still struggle with long-context documents. This study evaluates the performance of three leading LLMs: GPT-4o, Claude 3.5 Sonnet, and Gemini 1.5 Pro on lengthy, complex, and opinion-varying documents concerning infrastructure projects, under both zero-shot and few-shot scenarios. Our results indicate that GPT-4o excels in zero-shot scenarios for simpler, shorter documents, while Claude 3.5 Sonnet surpasses GPT-4o in handling more complex, sentiment-fluctuating opinions. In few-shot scenarios, Claude 3.5 Sonnet outperforms overall, while GPT-4o shows greater stability as the number of demonstrations increases.
☆ Beyond Linear Approximations: A Novel Pruning Approach for Attention Matrix
Large Language Models (LLMs) have shown immense potential in enhancing various aspects of our daily lives, from conversational AI to search and AI assistants. However, their growing capabilities come at the cost of extremely large model sizes, making deployment on edge devices challenging due to memory and computational constraints. This paper introduces a novel approach to LLM weight pruning that directly optimizes for approximating the attention matrix, a core component of transformer architectures. Unlike existing methods that focus on linear approximations, our approach accounts for the non-linear nature of the Softmax attention mechanism. We provide theoretical guarantees for the convergence of our Gradient Descent-based optimization method to a near-optimal pruning mask solution. Our preliminary empirical results demonstrate the effectiveness of this approach in maintaining model performance while significantly reducing computational costs. This work establishes a new theoretical foundation for pruning algorithm design in LLMs, potentially paving the way for more efficient LLM inference on resource-constrained devices.
☆ Investigation of Speaker Representation for Target-Speaker Speech Processing
Target-speaker speech processing (TS) tasks, such as target-speaker automatic speech recognition (TS-ASR), target speech extraction (TSE), and personal voice activity detection (p-VAD), are important for extracting information about a desired speaker's speech even when it is corrupted by interfering speakers. While most studies have focused on training schemes or system architectures for each specific task, the auxiliary network for embedding target-speaker cues has not been investigated comprehensively in a unified cross-task evaluation. Therefore, this paper aims to address a fundamental question: what is the preferred speaker embedding for TS tasks? To this end, for the TS-ASR, TSE, and p-VAD tasks, we compare pre-trained speaker encoders (i.e., self-supervised or speaker recognition models) that compute speaker embeddings from pre-recorded enrollment speech of the target speaker with ideal speaker embeddings derived directly from the target speaker's identity in the form of a one-hot vector. To further understand the properties of ideal speaker embedding, we optimize it using a gradient-based approach to improve performance on the TS task. Our analysis reveals that speaker verification performance is somewhat unrelated to TS task performances, the one-hot vector outperforms enrollment-based ones, and the optimal embedding depends on the input mixture.
comment: Accepted at IEEE SLT 2024
☆ HR-Agent: A Task-Oriented Dialogue (TOD) LLM Agent Tailored for HR Applications
Recent LLM (Large Language Models) advancements benefit many fields such as education and finance, but HR has hundreds of repetitive processes, such as access requests, medical claim filing and time-off submissions, which are unaddressed. We relate these tasks to the LLM agent, which has addressed tasks such as writing assisting and customer support. We present HR-Agent, an efficient, confidential, and HR-specific LLM-based task-oriented dialogue system tailored for automating repetitive HR processes such as medical claims and access requests. Since conversation data is not sent to an LLM during inference, it preserves confidentiality required in HR-related tasks.
☆ Unleashing the Power of LLMs as Multi-Modal Encoders for Text and Graph-Structured Data
Graph-structured information offers rich contextual information that can enhance language models by providing structured relationships and hierarchies, leading to more expressive embeddings for various applications such as retrieval, question answering, and classification. However, existing methods for integrating graph and text embeddings, often based on Multi-layer Perceptrons (MLPs) or shallow transformers, are limited in their ability to fully exploit the heterogeneous nature of these modalities. To overcome this, we propose Janus, a simple yet effective framework that leverages Large Language Models (LLMs) to jointly encode text and graph data. Specifically, Janus employs an MLP adapter to project graph embeddings into the same space as text embeddings, allowing the LLM to process both modalities jointly. Unlike prior work, we also introduce contrastive learning to align the graph and text spaces more effectively, thereby improving the quality of learned joint embeddings. Empirical results across six datasets spanning three tasks, knowledge graph-contextualized question answering, graph-text pair classification, and retrieval, demonstrate that Janus consistently outperforms existing baselines, achieving significant improvements across multiple datasets, with gains of up to 11.4% in QA tasks. These results highlight Janus's effectiveness in integrating graph and text data. Ablation studies further validate the effectiveness of our method.
☆ "Is Hate Lost in Translation?": Evaluation of Multilingual LGBTQIA+ Hate Speech Detection
This paper explores the challenges of detecting LGBTQIA+ hate speech of large language models across multiple languages, including English, Italian, Chinese and (code-switched) English-Tamil, examining the impact of machine translation and whether the nuances of hate speech are preserved across translation. We examine the hate speech detection ability of zero-shot and fine-tuned GPT. Our findings indicate that: (1) English has the highest performance and the code-switching scenario of English-Tamil being the lowest, (2) fine-tuning improves performance consistently across languages whilst translation yields mixed results. Through simple experimentation with original text and machine-translated text for hate speech detection along with a qualitative error analysis, this paper sheds light on the socio-cultural nuances and complexities of languages that may not be captured by automatic translation.
comment: Under review
☆ On the Capacity of Citation Generation by Large Language Models
Retrieval-augmented generation (RAG) appears as a promising method to alleviate the "hallucination" problem in large language models (LLMs), since it can incorporate external traceable resources for response generation. The essence of RAG in combating the hallucination issue lies in accurately attributing claims in responses to the corresponding retrieved documents. However, most of existing works focus on improving the quality of generated responses from the LLM, while largely overlooked its ability to attribute sources accurately. In this study, we conduct a systematic analysis about the capabilities of LLMs in generating citations within response generation, and further introduce a novel method to enhance their citation generation abilities. Specifically, we evaluate both the correctness and citation quality for seven widely-used LLMs on two benchmark datasets. Meanwhile, we introduce new citation evaluation metrics to eliminate the over-penalization of unnecessary and excessive citations in existing metrics. Furthermore, we propose a Generate-then-Refine method that completes relevant citations and removes irrelevant ones without altering the response text. The results on WebGLM-QA, ASQA and ELI5 datasets show that our method substantially improves the quality of citations in responses generated by LLMs.
comment: Accepted by CCIR 2024
☆ Sampling Strategies for Creation of a Benchmark for Dialectal Sentiment Classification
This paper investigates data sampling strategies to create a benchmark for dialectal sentiment classification of Google Places reviews written in English. Based on location-based filtering, we collect a self-supervised dataset of reviews in Australian (Australian English), Indian (Indian English), and British (British English) English with self-supervised sentiment labels (1-star to 5-star). We employ sampling techniques based on label semantics, review length, and sentiment proportion and report performances on three fine-tuned BERT-based models. Our multi-dialect evaluation provides pointers to challenging scenarios for inner-circle (Australian English and British English) as well as non-native dialects (Indian English) of English, highlighting the need for more diverse benchmarks.
comment: Under review
☆ Athena: Retrieval-augmented Legal Judgment Prediction with Large Language Models
Recently, large language models (LLMs) like ChatGPT, LLaMA, and Claude have prevailed in countless domains, including legal scenarios. With LLMs' rapid technological progress, the development of prompt engineering (PE) as an interface between the LLMs and real-world applications has drawn the attention of all developers. Various PE methods have been proposed to overcome real-world challenges, such as few-shot prompting, chain-of-thought, and retrieval-augmented generation (RAG). However, RAG for legal judgment prediction (LJP) is still underexplored. To address this, we propose "Athena", a novel framework cultivating RAG as a core preprocess component to enhance LLMs' performance on specialized tasks. Athena constructs a knowledge base for accusations, attached with a semantic retrieval mechanism through vectorization. Our experiments show that Athena's overall performance has improved significantly, achieving state-of-the-art results on the CAIL2018 dataset. Our ablation study on the in-context window size parameter further reproduces LLMs' "lost-in-the-middle" phenomenon with a relative positional variation. And with moderate hyper-parameter-tuning, we can achieve at most 95% of accuracy accordingly. We also study the impact of query rewriting and data distribution, providing possible directions for future research based on former analyses.
comment: 13 pages, 6 figures
☆ Model Swarms: Collaborative Search to Adapt LLM Experts via Swarm Intelligence
We propose Model Swarms, a collaborative search algorithm to adapt LLMs via swarm intelligence, the collective behavior guiding individual systems. Specifically, Model Swarms starts with a pool of LLM experts and a utility function. Guided by the best-found checkpoints across models, diverse LLM experts collaboratively move in the weight space and optimize a utility function representing model adaptation objectives. Compared to existing model composition approaches, Model Swarms offers tuning-free model adaptation, works in low-data regimes with as few as 200 examples, and does not require assumptions about specific experts in the swarm or how they should be composed. Extensive experiments demonstrate that Model Swarms could flexibly adapt LLM experts to a single task, multi-task domains, reward models, as well as diverse human interests, improving over 12 model composition baselines by up to 21.0% across tasks and contexts. Further analysis reveals that LLM experts discover previously unseen capabilities in initial checkpoints and that Model Swarms enable the weak-to-strong transition of experts through the collaborative search process.
☆ Scaling laws for post-training quantized large language models
Generalization abilities of well-trained large language models (LLMs) are known to scale predictably as a function of model size. In contrast to the existence of practical scaling laws governing pre-training, the quality of LLMs after post-training compression remains highly unpredictable, often requiring case-by-case validation in practice. In this work, we attempted to close this gap for post-training weight quantization of LLMs by conducting a systematic empirical study on multiple LLM families quantized to numerous low-precision tensor data types using popular weight quantization techniques. We identified key scaling factors pertaining to characteristics of the local loss landscape, based on which the performance of quantized LLMs can be reasonably well predicted by a statistical model.
☆ Planning Anything with Rigor: General-Purpose Zero-Shot Planning with LLM-based Formalized Programming
While large language models (LLMs) have recently demonstrated strong potential in solving planning problems, there is a trade-off between flexibility and complexity. LLMs, as zero-shot planners themselves, are still not capable of directly generating valid plans for complex planning problems such as multi-constraint or long-horizon tasks. On the other hand, many frameworks aiming to solve complex planning problems often rely on task-specific preparatory efforts, such as task-specific in-context examples and pre-defined critics/verifiers, which limits their cross-task generalization capability. In this paper, we tackle these challenges by observing that the core of many planning problems lies in optimization problems: searching for the optimal solution (best plan) with goals subject to constraints (preconditions and effects of decisions). With LLMs' commonsense, reasoning, and programming capabilities, this opens up the possibilities of a universal LLM-based approach to planning problems. Inspired by this observation, we propose LLMFP, a general-purpose framework that leverages LLMs to capture key information from planning problems and formally formulate and solve them as optimization problems from scratch, with no task-specific examples needed. We apply LLMFP to 9 planning problems, ranging from multi-constraint decision making to multi-step planning problems, and demonstrate that LLMFP achieves on average 83.7% and 86.8% optimal rate across 9 tasks for GPT-4o and Claude 3.5 Sonnet, significantly outperforming the best baseline (direct planning with OpenAI o1-preview) with 37.6% and 40.7% improvements. We also validate components of LLMFP with ablation experiments and analyzed the underlying success and failure reasons.
comment: 50 pages, 25 figures, 7 tables
☆ OMCAT: Omni Context Aware Transformer
Large Language Models (LLMs) have made significant strides in text generation and comprehension, with recent advancements extending into multimodal LLMs that integrate visual and audio inputs. However, these models continue to struggle with fine-grained, cross-modal temporal understanding, particularly when correlating events across audio and video streams. We address these challenges with two key contributions: a new dataset and model, called OCTAV and OMCAT respectively. OCTAV (Omni Context and Temporal Audio Video) is a novel dataset designed to capture event transitions across audio and video. Second, OMCAT (Omni Context Aware Transformer) is a powerful model that leverages RoTE (Rotary Time Embeddings), an innovative extension of RoPE, to enhance temporal grounding and computational efficiency in time-anchored tasks. Through a robust three-stage training pipeline-feature alignment, instruction tuning, and OCTAV-specific training-OMCAT excels in cross-modal temporal understanding. Our model demonstrates state-of-the-art performance on Audio-Visual Question Answering (AVQA) tasks and the OCTAV benchmark, showcasing significant gains in temporal reasoning and cross-modal alignment, as validated through comprehensive experiments and ablation studies. Our dataset and code will be made publicly available. The link to our demo page is https://om-cat.github.io.
comment: Demo page: https://om-cat.github.io
☆ De-jargonizing Science for Journalists with GPT-4: A Pilot Study
This study offers an initial evaluation of a human-in-the-loop system leveraging GPT-4 (a large language model or LLM), and Retrieval-Augmented Generation (RAG) to identify and define jargon terms in scientific abstracts, based on readers' self-reported knowledge. The system achieves fairly high recall in identifying jargon and preserves relative differences in readers' jargon identification, suggesting personalization as a feasible use-case for LLMs to support sense-making of complex information. Surprisingly, using only abstracts for context to generate definitions yields slightly more accurate and higher quality definitions than using RAG-based context from the fulltext of an article. The findings highlight the potential of generative AI for assisting science reporters, and can inform future work on developing tools to simplify dense documents.
comment: Accepted to Computation+Journalism Symposium 2024
☆ LegalLens Shared Task 2024: Legal Violation Identification in Unstructured Text
This paper presents the results of the LegalLens Shared Task, focusing on detecting legal violations within text in the wild across two sub-tasks: LegalLens-NER for identifying legal violation entities and LegalLens-NLI for associating these violations with relevant legal contexts and affected individuals. Using an enhanced LegalLens dataset covering labor, privacy, and consumer protection domains, 38 teams participated in the task. Our analysis reveals that while a mix of approaches was used, the top-performing teams in both tasks consistently relied on fine-tuning pre-trained language models, outperforming legal-specific models and few-shot methods. The top-performing team achieved a 7.11% improvement in NER over the baseline, while NLI saw a more marginal improvement of 5.7%. Despite these gains, the complexity of legal texts leaves room for further advancements.
☆ Large-scale cloze evaluation reveals that token prediction tasks are neither lexically nor semantically aligned
In this work we compare the generative behavior at the next token prediction level in several language models by comparing them to human productions in the cloze task. We find that while large models trained for longer are typically better estimators of human productions, but they reliably under-estimate the probabilities of human responses, over-rank rare responses, under-rank top responses, and produce highly distinct semantic spaces. Altogether, this work demonstrates in a tractable, interpretable domain that LM generations can not be used as replacements of or models of the cloze task.
☆ A State-of-the-Art Morphosyntactic Parser and Lemmatizer for Ancient Greek
This paper presents an experiment consisting in the comparison of six models to identify a state-of-the-art morphosyntactic parser and lemmatizer for Ancient Greek capable of annotating according to the Ancient Greek Dependency Treebank annotation scheme. A normalized version of the major collections of annotated texts was used to (i) train the baseline model Dithrax with randomly initialized character embeddings and (ii) fine-tune Trankit and four recent models pretrained on Ancient Greek texts, i.e., GreBERTa and PhilBERTa for morphosyntactic annotation and GreTA and PhilTa for lemmatization. A Bayesian analysis shows that Dithrax and Trankit annotate morphology practically equivalently, while syntax is best annotated by Trankit and lemmata by GreTa. The results of the experiment suggest that token embeddings are not sufficient to achieve high UAS and LAS scores unless they are coupled with a modeling strategy specifically designed to capture syntactic relationships. The dataset and best-performing models are made available online for reuse.
☆ Skill-LLM: Repurposing General-Purpose LLMs for Skill Extraction
Accurate skill extraction from job descriptions is crucial in the hiring process but remains challenging. Named Entity Recognition (NER) is a common approach used to address this issue. With the demonstrated success of large language models (LLMs) in various NLP tasks, including NER, we propose fine-tuning a specialized Skill-LLM and a light weight model to improve the precision and quality of skill extraction. In our study, we evaluated the fine-tuned Skill-LLM and the light weight model using a benchmark dataset and compared its performance against state-of-the-art (SOTA) methods. Our results show that this approach outperforms existing SOTA techniques.
☆ Sabiá-3 Technical Report
This report presents Sabi\'a-3, our new flagship language model trained on a large brazilian-centric corpus. Evaluations across diverse professional and academic benchmarks show a strong performance on Portuguese and Brazil-related tasks. Sabi\'a-3 shows large improvements in comparison to our previous best of model, Sabi\'a-2 Medium, especially in reasoning-intensive tasks. Notably, Sabi\'a-3's average performance matches frontier LLMs, while it is offered at a three to four times lower cost per token, reinforcing the benefits of domain specialization.
☆ Boosting Logical Fallacy Reasoning in LLMs via Logical Structure Tree EMNLP 2024
Logical fallacy uses invalid or faulty reasoning in the construction of a statement. Despite the prevalence and harmfulness of logical fallacies, detecting and classifying logical fallacies still remains a challenging task. We observe that logical fallacies often use connective words to indicate an intended logical relation between two arguments, while the argument semantics does not actually support the logical relation. Inspired by this observation, we propose to build a logical structure tree to explicitly represent and track the hierarchical logic flow among relation connectives and their arguments in a statement. Specifically, this logical structure tree is constructed in an unsupervised manner guided by the constituency tree and a taxonomy of connectives for ten common logical relations, with relation connectives as non-terminal nodes and textual arguments as terminal nodes, and the latter are mostly elementary discourse units. We further develop two strategies to incorporate the logical structure tree into LLMs for fallacy reasoning. Firstly, we transform the tree into natural language descriptions and feed the textualized tree into LLMs as a part of the hard text prompt. Secondly, we derive a relation-aware tree embedding and insert the tree embedding into LLMs as a soft prompt. Experiments on benchmark datasets demonstrate that our approach based on logical structure tree significantly improves precision and recall for both fallacy detection and fallacy classification.
comment: Accepted to EMNLP 2024
☆ Concept-Reversed Winograd Schema Challenge: Evaluating and Improving Robust Reasoning in Large Language Models via Abstraction
While Large Language Models (LLMs) have showcased remarkable proficiency in reasoning, there is still a concern about hallucinations and unreliable reasoning issues due to semantic associations and superficial logical chains. To evaluate the extent to which LLMs perform robust reasoning instead of relying on superficial logical chains, we propose a new evaluation dataset, the Concept-Reversed Winograd Schema Challenge (CR-WSC), based on the famous Winograd Schema Challenge (WSC) dataset. By simply reversing the concepts to those that are more associated with the wrong answer, we find that the performance of LLMs drops significantly despite the rationale of reasoning remaining the same. Furthermore, we propose Abstraction-of-Thought (AoT), a novel prompt method for recovering adversarial cases to normal cases using conceptual abstraction to improve LLMs' robustness and consistency in reasoning, as demonstrated by experiments on CR-WSC.
☆ On Classification with Large Language Models in Cultural Analytics
In this work, we survey the way in which classification is used as a sensemaking practice in cultural analytics, and assess where large language models can fit into this landscape. We identify ten tasks supported by publicly available datasets on which we empirically assess the performance of LLMs compared to traditional supervised methods, and explore the ways in which LLMs can be employed for sensemaking goals beyond mere accuracy. We find that prompt-based LLMs are competitive with traditional supervised models for established tasks, but perform less well on de novo tasks. In addition, LLMs can assist sensemaking by acting as an intermediary input to formal theory testing.
☆ LocoMotion: Learning Motion-Focused Video-Language Representations ACCV 2024
This paper strives for motion-focused video-language representations. Existing methods to learn video-language representations use spatial-focused data, where identifying the objects and scene is often enough to distinguish the relevant caption. We instead propose LocoMotion to learn from motion-focused captions that describe the movement and temporal progression of local object motions. We achieve this by adding synthetic motions to videos and using the parameters of these motions to generate corresponding captions. Furthermore, we propose verb-variation paraphrasing to increase the caption variety and learn the link between primitive motions and high-level verbs. With this, we are able to learn a motion-focused video-language representation. Experiments demonstrate our approach is effective for a variety of downstream tasks, particularly when limited data is available for fine-tuning. Code is available: https://hazeldoughty.github.io/Papers/LocoMotion/
comment: ACCV 2024
☆ MoE-Pruner: Pruning Mixture-of-Experts Large Language Model using the Hints from Its Router
Mixture-of-Experts (MoE) architectures face challenges such as high memory consumption and redundancy in experts. Pruning MoE can reduce network weights while maintaining model performance. Motivated by the recent observation of emergent large magnitude features in Large Language Models (LLM) and MoE routing policy, we propose MoE-Pruner, a method that prunes weights with the smallest magnitudes multiplied by the corresponding input activations and router weights, on each output neuron. Our pruning method is one-shot, requiring no retraining or weight updates. We evaluate our method on Mixtral-8x7B and Mixtral-8x22B across multiple language benchmarks. Experimental results show that our pruning method significantly outperforms state-of-the-art LLM pruning methods. Furthermore, our pruned MoE models can benefit from a pretrained teacher model through expert-wise knowledge distillation, improving performance post-pruning. Experimental results demonstrate that the Mixtral-8x7B model with 50% sparsity maintains 99% of the performance of the original model after the expert-wise knowledge distillation.
♻ ☆ LoRA-Pro: Are Low-Rank Adapters Properly Optimized?
Low-rank adaptation, also known as LoRA, has emerged as a prominent method for parameter-efficient fine-tuning of foundation models. Despite its computational efficiency, LoRA still yields inferior performance compared to full fine-tuning. In this paper, we first uncover a fundamental connection between the optimization processes of LoRA and full fine-tuning: using LoRA for optimization is mathematically equivalent to full fine-tuning using a low-rank gradient for parameter updates. And this low-rank gradient can be expressed in terms of the gradients of the two low-rank matrices in LoRA. Leveraging this insight, we introduce LoRA-Pro, a method that enhances LoRA's performance by strategically adjusting the gradients of these low-rank matrices. This adjustment allows the low-rank gradient to more accurately approximate the full fine-tuning gradient, thereby narrowing the performance gap between LoRA and full fine-tuning. Furthermore, we theoretically derive the optimal solutions for adjusting the gradients of the low-rank matrices, applying them during fine-tuning in LoRA-Pro. We conduct extensive experiments across natural language understanding, dialogue generation, mathematical reasoning, code generation, and image classification tasks, demonstrating that LoRA-Pro substantially improves LoRA's performance, effectively narrowing the gap with full fine-tuning. Code is publicly available at \url{https://github.com/mrflogs/LoRA-Pro}.
♻ ☆ TemporalBench: Benchmarking Fine-grained Temporal Understanding for Multimodal Video Models
Understanding fine-grained temporal dynamics is crucial for multimodal video comprehension and generation. Due to the lack of fine-grained temporal annotations, existing video benchmarks mostly resemble static image benchmarks and are incompetent at evaluating models for temporal understanding. In this paper, we introduce TemporalBench, a new benchmark dedicated to evaluating fine-grained temporal understanding in videos. TemporalBench consists of ~10K video question-answer pairs, derived from ~2K high-quality human annotations detailing the temporal dynamics in video clips. As a result, our benchmark provides a unique testbed for evaluating various temporal understanding and reasoning abilities such as action frequency, motion magnitude, event order, etc. Moreover, it enables evaluations on various tasks like both video question answering and captioning, both short and long video understanding, as well as different models such as multimodal video embedding models and text generation models. Results show that state-of-the-art models like GPT-4o achieve only 38.5% question answering accuracy on TemporalBench, demonstrating a significant gap (~30%) between humans and AI in temporal understanding. Furthermore, we notice a critical pitfall for multi-choice QA where LLMs can detect the subtle changes in negative captions and find a centralized description as a cue for its prediction, where we propose Multiple Binary Accuracy (MBA) to correct such bias. We hope that TemporalBench can foster research on improving models' temporal reasoning capabilities. Both dataset and evaluation code will be made available.
comment: Project Page: https://temporalbench.github.io/
♻ ☆ What We Talk About When We Talk About LMs: Implicit Paradigm Shifts and the Ship of Language Models
The term Language Models (LMs), as a time-specific collection of models of interest, is constantly reinvented, with its referents updated much like the $\textit{Ship of Theseus}$ replaces its parts but remains the same ship in essence. In this paper, we investigate this $\textit{Ship of Language Models}$ problem, wherein scientific evolution takes the form of continuous, implicit retrofits of key existing terms. We seek to initiate a novel perspective of scientific progress, in addition to the more well-studied emergence of new terms. To this end, we construct the data infrastructure based on recent NLP publications. Then, we perform a series of text-based analyses toward a detailed, quantitative understanding of the use of Language Models as a term of art. Our work highlights how systems and theories influence each other in scientific discourse, and we call for attention to the transformation of this Ship that we all are contributing to.
♻ ☆ Teuken-7B-Base & Teuken-7B-Instruct: Towards European LLMs
We present two multilingual LLMs designed to embrace Europe's linguistic diversity by supporting all 24 official languages of the European Union. Trained on a dataset comprising around 60% non-English data and utilizing a custom multilingual tokenizer, our models address the limitations of existing LLMs that predominantly focus on English or a few high-resource languages. We detail the models' development principles, i.e., data composition, tokenizer optimization, and training methodologies. The models demonstrate competitive performance across multilingual benchmarks, as evidenced by their performance on European versions of ARC, HellaSwag, MMLU, and TruthfulQA.
♻ ☆ Self-Data Distillation for Recovering Quality in Pruned Large Language Models NeurIPS 2024
Large language models have driven significant progress in natural language processing, but their deployment requires substantial compute and memory resources. As models scale, compression techniques become essential for balancing model quality with computational efficiency. Structured pruning, which removes less critical components of the model, is a promising strategy for reducing complexity. However, one-shot pruning often results in significant quality degradation, particularly in tasks requiring multi-step reasoning. To recover lost quality, supervised fine-tuning (SFT) is commonly applied, but it can lead to catastrophic forgetting by shifting the model's learned data distribution. Therefore, addressing the degradation from both pruning and SFT is essential to preserve the original model's quality. In this work, we propose self-data distilled fine-tuning to address these challenges. Our approach leverages the original, unpruned model to generate a distilled dataset that preserves semantic richness and mitigates catastrophic forgetting by maintaining alignment with the base model's knowledge. Empirically, we demonstrate that self-data distillation consistently outperforms standard SFT, improving average accuracy by up to 8% on the HuggingFace OpenLLM Leaderboard v1. Specifically, when pruning 6 decoder blocks on Llama3.1-8B Instruct (i.e., 32 to 26 layers, reducing the model size from 8.03B to 6.72B parameters), our method retains 91.2% of the original model's accuracy compared to 81.7% with SFT, while reducing real-world FLOPs by 16.30%. Furthermore, our approach scales effectively across datasets, with the quality improving as the dataset size increases.
comment: Accepted at the NeurIPS 2024 Machine Learning and Compression Workshop
♻ ☆ Improving semantic understanding in speech language models via brain-tuning ICLR 2025
Speech language models align with human brain responses to natural language to an impressive degree. However, current models rely heavily on low-level speech features, indicating they lack brain-relevant semantics which limits their utility as model organisms of semantic processing in the brain. In this work, we address this limitation by inducing brain-relevant bias directly into the models via fine-tuning with fMRI recordings of people listening to natural stories, a process we name brain-tuning. After testing it on 3 different pretrained model families, we show that brain-tuning not only improves overall alignment with new brain recordings in semantic language regions, but also reduces the reliance on low-level speech features for this alignment. Excitingly, we further show that brain-tuning leads to 1) consistent improvements in performance on a range of downstream tasks and 2) a representational space with increased semantic preference. Our results provide converging evidence, for the first time, that incorporating brain signals into the training of language models improves the models' semantic understanding.
comment: Under Review at ICLR 2025
♻ ☆ TAMS: Translation-Assisted Morphological Segmentation ACL
Canonical morphological segmentation is the process of analyzing words into the standard (aka underlying) forms of their constituent morphemes. This is a core task in language documentation, and NLP systems have the potential to dramatically speed up this process. But in typical language documentation settings, training data for canonical morpheme segmentation is scarce, making it difficult to train high quality models. However, translation data is often much more abundant, and, in this work, we present a method that attempts to leverage this data in the canonical segmentation task. We propose a character-level sequence-to-sequence model that incorporates representations of translations obtained from pretrained high-resource monolingual language models as an additional signal. Our model outperforms the baseline in a super-low resource setting but yields mixed results on training splits with more data. While further work is needed to make translations useful in higher-resource settings, our model shows promise in severely resource-constrained settings.
comment: Submitted to ACL ARR on December 15th 2023
♻ ☆ LLM-Based Robust Product Classification in Commerce and Compliance EMNLP 2024
Product classification is a crucial task in international trade, as compliance regulations are verified and taxes and duties are applied based on product categories. Manual classification of products is time-consuming and error-prone, and the sheer volume of products imported and exported renders the manual process infeasible. Consequently, e-commerce platforms and enterprises involved in international trade have turned to automatic product classification using machine learning. However, current approaches do not consider the real-world challenges associated with product classification, such as very abbreviated and incomplete product descriptions. In addition, recent advancements in generative Large Language Models (LLMs) and their reasoning capabilities are mainly untapped in product classification and e-commerce. In this research, we explore the real-life challenges of industrial classification and we propose data perturbations that allow for realistic data simulation. Furthermore, we employ LLM-based product classification to improve the robustness of the prediction in presence of incomplete data. Our research shows that LLMs with in-context learning outperform the supervised approaches in the clean-data scenario. Additionally, we illustrate that LLMs are significantly more robust than the supervised approaches when data attacks are present.
comment: Camera-ready version for Customizable NLP Workshop at EMNLP 2024. 11 pages
♻ ☆ THaMES: An End-to-End Tool for Hallucination Mitigation and Evaluation in Large Language Models NeurIPS 2024
Hallucination, the generation of factually incorrect content, is a growing challenge in Large Language Models (LLMs). Existing detection and mitigation methods are often isolated and insufficient for domain-specific needs, lacking a standardized pipeline. This paper introduces THaMES (Tool for Hallucination Mitigations and EvaluationS), an integrated framework and library addressing this gap. THaMES offers an end-to-end solution for evaluating and mitigating hallucinations in LLMs, featuring automated test set generation, multifaceted benchmarking, and adaptable mitigation strategies. It automates test set creation from any corpus, ensuring high data quality, diversity, and cost-efficiency through techniques like batch processing, weighted sampling, and counterfactual validation. THaMES assesses a model's ability to detect and reduce hallucinations across various tasks, including text generation and binary classification, applying optimal mitigation strategies like In-Context Learning (ICL), Retrieval Augmented Generation (RAG), and Parameter-Efficient Fine-tuning (PEFT). Evaluations of state-of-the-art LLMs using a knowledge base of academic papers, political news, and Wikipedia reveal that commercial models like GPT-4o benefit more from RAG than ICL, while open-weight models like Llama-3.1-8B-Instruct and Mistral-Nemo gain more from ICL. Additionally, PEFT significantly enhances the performance of Llama-3.1-8B-Instruct in both evaluation tasks.
comment: Accepted in NeurIPS 2024 SoLaR (Socially Responsible Language Modelling Research ) Workshop
♻ ☆ LoRTA: Low Rank Tensor Adaptation of Large Language Models
Low Rank Adaptation (LoRA) is a popular Parameter Efficient Fine Tuning (PEFT) method that effectively adapts large pre-trained models for downstream tasks. LoRA parameterizes model updates using low-rank matrices at each layer, significantly reducing the number of trainable parameters and, consequently, resource requirements during fine-tuning. However, the lower bound on the number of trainable parameters remains high due to the use of the low-rank matrix model. In this paper, we address this limitation by proposing a novel approach that employs a low rank tensor parametrization for model updates. The proposed low rank tensor model can significantly reduce the number of trainable parameters, while also allowing for finer-grained control over adapter size. Our experiments on Natural Language Understanding, Instruction Tuning, Preference Optimization and Protein Folding benchmarks demonstrate that our method is both efficient and effective for fine-tuning large language models, achieving a substantial reduction in the number of parameters while maintaining comparable performance.
♻ ☆ GOVERN: Gradient Orientation Vote Ensemble for Multi-Teacher Reinforced Distillation EMNLP 2024
Pre-trained language models have become an integral component of question-answering systems, achieving remarkable performance. However, for practical deployment, it is crucial to perform knowledge distillation to maintain high performance while operating under computational constraints. In this paper, we address a key question: given the importance of unsupervised distillation for student model performance, how can knowledge from multiple teacher models be effectively ensemble during this stage without the guidance of labels? We propose a novel algorithm, GOVERN, to tackle this issue. GOVERN has demonstrated significant improvements in both offline and online experiments, enabling the student model to achieve results comparable to that of teacher ensembles. Our experiments show that GOVERN remarkably requires a mere 1\% of the ensemble method's inference budget to achieve 99.5\% of performance. The proposed algorithm has been successfully deployed in a real-world commercial question-answering system, demonstrating its real-world applicability.
comment: Accepted by EMNLP 2024 Industry Track
♻ ☆ Towards Supporting Legal Argumentation with NLP: Is More Data Really All You Need? EMNLP 2024
Modeling legal reasoning and argumentation justifying decisions in cases has always been central to AI & Law, yet contemporary developments in legal NLP have increasingly focused on statistically classifying legal conclusions from text. While conceptually simpler, these approaches often fall short in providing usable justifications connecting to appropriate legal concepts. This paper reviews both traditional symbolic works in AI & Law and recent advances in legal NLP, and distills possibilities of integrating expert-informed knowledge to strike a balance between scalability and explanation in symbolic vs. data-driven approaches. We identify open challenges and discuss the potential of modern NLP models and methods that integrate
comment: Accepted to NLLP Workshop, EMNLP 2024
♻ ☆ Subgraph-Aware Training of Language Models for Knowledge Graph Completion Using Structure-Aware Contrastive Learning
Fine-tuning pre-trained language models (PLMs) has recently shown a potential to improve knowledge graph completion (KGC). However, most PLM-based methods focus solely on encoding textual information, neglecting the long-tailed nature of knowledge graphs and their various topological structures, e.g., subgraphs, shortest paths, and degrees. We claim that this is a major obstacle to achieving higher accuracy of PLMs for KGC. To this end, we propose a Subgraph-Aware Training framework for KGC (SATKGC) with two ideas: (i) subgraph-aware mini-batching to encourage hard negative sampling and to mitigate an imbalance in the frequency of entity occurrences during training, and (ii) new contrastive learning to focus more on harder in-batch negative triples and harder positive triples in terms of the structural properties of the knowledge graph. To the best of our knowledge, this is the first study to comprehensively incorporate the structural inductive bias of the knowledge graph into fine-tuning PLMs. Extensive experiments on three KGC benchmarks demonstrate the superiority of SATKGC. Our code is available.
comment: 8 pages, 7 figures
♻ ☆ Mitigate Position Bias in Large Language Models via Scaling a Single Dimension
Large Language Models (LLMs) are increasingly applied in various real-world scenarios due to their excellent generalization capabilities and robust generative abilities. However, they exhibit position bias, also known as "lost in the middle", a phenomenon that is especially pronounced in long-context scenarios, which indicates the placement of the key information in different positions of a prompt can significantly affect accuracy. This paper first explores the micro-level manifestations of position bias, concluding that attention weights are a micro-level expression of position bias. It further identifies that, in addition to position embeddings, causal attention mask also contributes to position bias by creating position-specific hidden states. Based on these insights, we propose a method to mitigate position bias by scaling this positional hidden states. Experiments on the NaturalQuestions Multi-document QA, KV retrieval, LongBench and timeline reorder tasks, using various models including RoPE models, context windowextended models, and Alibi models, demonstrate the effectiveness and generalizability of our approach. Our method can improve performance by up to 15.2% by modifying just one dimension of hidden states. Our code is available at https://aka.ms/PositionalHidden.
♻ ☆ Phantom: General Trigger Attacks on Retrieval Augmented Language Generation
Retrieval Augmented Generation (RAG) expands the capabilities of modern large language models (LLMs), by anchoring, adapting, and personalizing their responses to the most relevant knowledge sources. It is particularly useful in chatbot applications, allowing developers to customize LLM output without expensive retraining. Despite their significant utility in various applications, RAG systems present new security risks. In this work, we propose new attack vectors that allow an adversary to inject a single malicious document into a RAG system's knowledge base, and mount a backdoor poisoning attack. We design Phantom, a general two-stage optimization framework against RAG systems, that crafts a malicious poisoned document leading to an integrity violation in the model's output. First, the document is constructed to be retrieved only when a specific trigger sequence of tokens appears in the victim's queries. Second, the document is further optimized with crafted adversarial text that induces various adversarial objectives on the LLM output, including refusal to answer, reputation damage, privacy violations, and harmful behaviors. We demonstrate our attacks on multiple LLM architectures, including Gemma, Vicuna, and Llama, and show that they transfer to GPT-3.5 Turbo and GPT-4. Finally, we successfully conducted a Phantom attack on NVIDIA's black-box production RAG system, "Chat with RTX".
♻ ☆ Enhancing Agent Learning through World Dynamics Modeling
Large language models (LLMs) have been increasingly applied to tasks in language understanding and interactive decision-making, with their impressive performance largely attributed to the extensive domain knowledge embedded within them. However, the depth and breadth of this knowledge can vary across domains. Many existing approaches assume that LLMs possess a comprehensive understanding of their environment, often overlooking potential gaps in their grasp of actual world dynamics. To address this, we introduce Discover, Verify, and Evolve (DiVE), a framework that discovers world dynamics from a small number of demonstrations, verifies the accuracy of these dynamics, and evolves new, advanced dynamics tailored to the current situation. Through extensive evaluations, we assess the impact of each component on performance and compare the dynamics generated by DiVE to human-annotated dynamics. Our results show that LLMs guided by DiVE make more informed decisions, achieving rewards comparable to human players in the Crafter environment and surpassing methods that require prior task-specific training in the MiniHack environment.
♻ ☆ Unlocking Structured Thinking in Language Models with Cognitive Prompting ICLR 2025
We propose cognitive prompting as a novel approach to guide problem-solving in large language models (LLMs) through structured, human-like cognitive operations such as goal clarification, decomposition, filtering, abstraction, and pattern recognition. By employing systematic, step-by-step reasoning, cognitive prompting enables LLMs to efficiently tackle complex, multi-step tasks. We evaluate the effectiveness of cognitive prompting on Meta's LLaMA models, comparing performance on arithmetic reasoning tasks using the GSM8K dataset and on commonsense reasoning benchmarks. Our analysis includes comparisons between models without cognitive prompting, models with a static sequence of cognitive operations, and models using reflective cognitive prompting, where the LLM dynamically self-selects the sequence of cognitive operations. The results show that cognitive prompting, particularly when dynamically adapted, significantly improves the performance of larger models, such as LLaMA3.1 70B, and enhances their ability to handle multi-step reasoning tasks. This approach also improves interpretability and flexibility, highlighting cognitive prompting as a promising strategy for general-purpose AI reasoning.
comment: 11 pages, submitted to ICLR 2025
♻ ☆ Teaching AI Agents to Search with Reflective-MCTS and Exploratory Learning
Autonomous agents have demonstrated significant potential in automating complex multistep decision-making tasks. However, even state-of-the-art vision-language models (VLMs), such as GPT-4o, still fall short of human-level performance, particularly in intricate web environments and long-horizon planning tasks. To address these limitations, we present Reflective Monte Carlo Tree Search (R-MCTS) and Exploratory Learning to build o1-like models for agentic applications. We first introduce R-MCTS, a novel test-time algorithm designed to enhance the ability of AI agents to explore decision space on the fly. R-MCTS extends traditional MCTS by 1) incorporating contrastive reflection, allowing agents to learn from past interactions and dynamically improve their search efficiency; and 2) using multi-agent debate to provide reliable state evaluation. Next, we introduce Exploratory Learning, a novel learning strategy to teach agents to search at inference time without relying on any external search algorithms. On the challenging VisualWebArena benchmark, our GPT-4o-based R-MCTS agent achieves a 6% to 30% relative improvement across various tasks compared to the previous state-of-the-art. Additionally, we show that the experience gained from test-time search can be effectively transferred back to GPT-4o via fine-tuning. After Exploratory Learning, GPT-4o 1) demonstrates the ability to explore the environment, evaluate a state, and backtrack to viable ones when it detects that the current state cannot lead to success, and 2) matches 87% of R-MCTS's performance while using significantly less compute. Notably, our work demonstrates the compute scaling properties in both training - data collection with R-MCTS - and testing time. These results suggest a promising research direction to enhance VLMs' reasoning and planning capabilities for agentic applications via test-time search and self-learning.
♻ ☆ Capturing Minds, Not Just Words: Enhancing Role-Playing Language Models with Personality-Indicative Data
Role-playing agents (RPA) have been a popular application area for large language models (LLMs), attracting significant interest from both industry and academia.While existing RPAs well portray the characters' knowledge and tones, they face challenges in capturing their minds, especially for small role-playing language models (RPLMs). In this paper, we propose to enhance RPLMs via personality-indicative data. Specifically, we leverage questions from psychological scales and distill advanced RPAs to generate dialogues that grasp the minds of characters. Experimental results validate that RPLMs trained with our dataset exhibit advanced role-playing capabilities for both general and personality-related evaluations. Code and data are available at \href{https://github.com/alienet1109/RolePersonality}{this URL}.
comment: 11 pages, 1 figures
♻ ☆ OLAPH: Improving Factuality in Biomedical Long-form Question Answering
In the medical domain, numerous scenarios necessitate the long-form generation ability of large language models (LLMs). Specifically, when addressing patients' questions, it is essential that the model's response conveys factual claims, highlighting the need for an automated method to evaluate those claims. Thus, we introduce MedLFQA, a benchmark dataset reconstructed using long-form question-answering datasets related to the biomedical domain. We use MedLFQA to facilitate a cost-effective automatic evaluations of factuality. We also propose OLAPH, a simple and novel framework that utilizes cost-effective and multifaceted automatic evaluation to construct a synthetic preference set and answers questions in our preferred manner. Our framework leads us to train LLMs step-by-step to reduce hallucinations and include crucial medical claims. We highlight that, even on evaluation metrics not used during training, LLMs trained with our OLAPH framework demonstrate significant performance improvement in factuality. Our findings reveal that a 7B LLM trained with our OLAPH framework can provide long answers comparable to the medical experts' answers in terms of factuality. We believe that our work could shed light on gauging the long-text generation ability of LLMs in the medical domain. Our code and datasets are available.
♻ ☆ Inducing anxiety in large language models can induce bias
Large language models (LLMs) are transforming research on machine learning while galvanizing public debates. Understanding not only when these models work well and succeed but also why they fail and misbehave is of great societal relevance. We propose to turn the lens of psychiatry, a framework used to describe and modify maladaptive behavior, to the outputs produced by these models. We focus on twelve established LLMs and subject them to a questionnaire commonly used in psychiatry. Our results show that six of the latest LLMs respond robustly to the anxiety questionnaire, producing comparable anxiety scores to humans. Moreover, the LLMs' responses can be predictably changed by using anxiety-inducing prompts. Anxiety-induction not only influences LLMs' scores on an anxiety questionnaire but also influences their behavior in a previously-established benchmark measuring biases such as racism and ageism. Importantly, greater anxiety-inducing text leads to stronger increases in biases, suggesting that how anxiously a prompt is communicated to large language models has a strong influence on their behavior in applied settings. These results demonstrate the usefulness of methods taken from psychiatry for studying the capable algorithms to which we increasingly delegate authority and autonomy.
♻ ☆ Discovering Knowledge-Critical Subnetworks in Pretrained Language Models EMNLP 2024
Pretrained language models (LMs) encode implicit representations of knowledge in their parameters. However, localizing these representations and disentangling them from each other remains an open problem. In this work, we investigate whether pretrained language models contain various knowledge-critical subnetworks: particular sparse computational subgraphs that can, if removed, precisely suppress specific knowledge the model has memorized. We propose a multi-objective differentiable masking scheme that can be applied to both weights and neurons to discover such subnetworks and show that we can use them to precisely remove specific knowledge from models while minimizing adverse effects on the behavior of the original model. We demonstrate our method on multiple GPT2 variants, uncovering highly sparse subnetworks (98%+ sparsity) that are critical for expressing specific collections of relational knowledge. When these subnetworks are removed, the remaining network maintains most of its initial abilities but struggles to represent the suppressed knowledge.
comment: EMNLP 2024
♻ ☆ Agent Planning with World Knowledge Model NeurIPS 2024
Recent endeavors towards directly using large language models (LLMs) as agent models to execute interactive planning tasks have shown commendable results. Despite their achievements, however, they still struggle with brainless trial-and-error in global planning and generating hallucinatory actions in local planning due to their poor understanding of the ``real'' physical world. Imitating humans' mental world knowledge model which provides global prior knowledge before the task and maintains local dynamic knowledge during the task, in this paper, we introduce parametric World Knowledge Model (WKM) to facilitate agent planning. Concretely, we steer the agent model to self-synthesize knowledge from both expert and sampled trajectories. Then we develop WKM, providing prior task knowledge to guide the global planning and dynamic state knowledge to assist the local planning. Experimental results on three complex real-world simulated datasets with three state-of-the-art open-source LLMs, Mistral-7B, Gemma-7B, and Llama-3-8B, demonstrate that our method can achieve superior performance compared to various strong baselines. Besides, we analyze to illustrate that our WKM can effectively alleviate the blind trial-and-error and hallucinatory action issues, providing strong support for the agent's understanding of the world. Other interesting findings include: 1) our instance-level task knowledge can generalize better to unseen tasks, 2) weak WKM can guide strong agent model planning, and 3) unified WKM training has promising potential for further development. The code is available at https://github.com/zjunlp/WKM.
comment: NeurIPS 2024
♻ ☆ Consistent Document-Level Relation Extraction via Counterfactuals
Many datasets have been developed to train and evaluate document-level relation extraction (RE) models. Most of these are constructed using real-world data. It has been shown that RE models trained on real-world data suffer from factual biases. To evaluate and address this issue, we present CovEReD, a counterfactual data generation approach for document-level relation extraction datasets using entity replacement. We first demonstrate that models trained on factual data exhibit inconsistent behavior: while they accurately extract triples from factual data, they fail to extract the same triples after counterfactual modification. This inconsistency suggests that models trained on factual data rely on spurious signals such as specific entities and external knowledge $\unicode{x2013}$ rather than on the input context $\unicode{x2013}$ to extract triples. We show that by generating document-level counterfactual data with CovEReD and training models on them, consistency is maintained with minimal impact on RE performance. We release our CovEReD pipeline as well as Re-DocRED-CF, a dataset of counterfactual RE documents, to assist in evaluating and addressing inconsistency in document-level RE.
♻ ☆ Words as Trigger Points in Social Media Discussions
Trigger points, introduced by Mau et al . [30], are rooted in theories of affective political identity and relate to deeply lying beliefs about moral expectations and social dispositions. Examining trigger points in online discussions helps understand why and when social media users engage in disagreements or affective political deliberations. This opens the door to modelling social media user engagement more effectively and studying the conditions and causal mechanisms that lead to adverse reactions, hate speech, and abusive language in online debates.
♻ ☆ Diverse Perspectives, Divergent Models: Cross-Cultural Evaluation of Depression Detection on Twitter NAACL 2024
Social media data has been used for detecting users with mental disorders, such as depression. Despite the global significance of cross-cultural representation and its potential impact on model performance, publicly available datasets often lack crucial metadata related to this aspect. In this work, we evaluate the generalization of benchmark datasets to build AI models on cross-cultural Twitter data. We gather a custom geo-located Twitter dataset of depressed users from seven countries as a test dataset. Our results show that depression detection models do not generalize globally. The models perform worse on Global South users compared to Global North. Pre-trained language models achieve the best generalization compared to Logistic Regression, though still show significant gaps in performance on depressed and non-Western users. We quantify our findings and provide several actionable suggestions to mitigate this issue.
comment: 6 pages, 2 figures, NAACL 2024 Main Conference
♻ ☆ Minimum Tuning to Unlock Long Output from LLMs with High Quality Data as the Key
As large language models rapidly evolve to support longer context, there is a notable disparity in their capability to generate output at greater lengths. Recent study suggests that the primary cause for this imbalance may arise from the lack of data with long-output during alignment training. In light of this observation, attempts are made to re-align foundation models with data that fills the gap, which result in models capable of generating lengthy output when instructed. In this paper, we explore the impact of data-quality in tuning a model for long output, and the possibility of doing so from the starting points of human-aligned (instruct or chat) models. With careful data curation, we show that it possible to achieve similar performance improvement in our tuned models, with only a small fraction of training data instances and compute. In addition, we assess the generalizability of such approaches by applying our tuning-recipes to several models. our findings suggest that, while capacities for generating long output vary across different models out-of-the-box, our approach to tune them with high-quality data using lite compute, consistently yields notable improvement across all models we experimented on. We have made public our curated dataset for tuning long-writing capability, the implementations of model tuning and evaluation, as well as the fine-tuned models, all of which can be openly-accessed.
♻ ☆ The optimal placement of the head in the noun phrase. The case of demonstrative, numeral, adjective and noun
The word order of a sentence is shaped by multiple principles. The principle of syntactic dependency distance minimization is in conflict with the principle of surprisal minimization (or predictability maximization) in single head syntactic dependency structures: while the former predicts that the head should be placed at the center of the linear arrangement, the latter predicts that the head should be placed at one of the ends (either first or last). A critical question is when surprisal minimization (or predictability maximization) should surpass syntactic dependency distance minimization. In the context of single head structures, it has been predicted that this is more likely to happen when two conditions are met, i.e. (a) fewer words are involved and (b) words are shorter. Here we test the prediction on the noun phrase when it is composed of a demonstrative, a numeral, an adjective and a noun. We find that, across preferred orders in languages, the noun tends to be placed at one of the ends, confirming the theoretical prediction. We also show evidence of anti locality effects: syntactic dependency distances in preferred orders are longer than expected by chance.
comment: typos corrected; in press in the Journal of Quantitative Linguistics
♻ ☆ Paraphrase Types Elicit Prompt Engineering Capabilities
Much of the success of modern language models depends on finding a suitable prompt to instruct the model. Until now, it has been largely unknown how variations in the linguistic expression of prompts affect these models. This study systematically and empirically evaluates which linguistic features influence models through paraphrase types, i.e., different linguistic changes at particular positions. We measure behavioral changes for five models across 120 tasks and six families of paraphrases (i.e., morphology, syntax, lexicon, lexico-syntax, discourse, and others). We also control for other prompt engineering factors (e.g., prompt length, lexical diversity, and proximity to training data). Our results show a potential for language models to improve tasks when their prompts are adapted in specific paraphrase types (e.g., 6.7% median gain in Mixtral 8x7B; 5.5% in LLaMA 3 8B). In particular, changes in morphology and lexicon, i.e., the vocabulary used, showed promise in improving prompts. These findings contribute to developing more robust language models capable of handling variability in linguistic expression.
♻ ☆ Large Language Models Are Cross-Lingual Knowledge-Free Reasoners
Large Language Models have demonstrated impressive reasoning capabilities across multiple languages. However, the relationship between capabilities in different languages is less explored. In this work, we decompose the process of reasoning tasks into two separated components: knowledge retrieval and knowledge-free reasoning, and analyze the relationship between cross-lingual transferability and these two components. With adapted commonsense reasoning datasets and constructed knowledge-free reasoning datasets, we show that the knowledge-free reasoning capability can be nearly perfectly transferred across various source-target language directions despite the secondary impact of resource in some specific target languages, while cross-lingual knowledge retrieval significantly hinders the transfer. Moreover, by analyzing the hidden states and feed-forward network neuron activation during the reasoning, we show that higher similarity of hidden representations and larger overlap of activated neurons could explain the better cross-lingual transferability of knowledge-free reasoning than knowledge retrieval. Thus, we hypothesize that knowledge-free reasoning shares similar neurons in different languages for reasoning, while knowledge is stored separately in different languages. Our code and data is available at: https://github.com/NJUNLP/Knowledge-Free-Reasoning.
♻ ☆ DiaSynth: Synthetic Dialogue Generation Framework for Low Resource Dialogue Applications
The scarcity of domain-specific dialogue datasets limits the development of dialogue systems across applications. Existing research is constrained by general or niche datasets that lack sufficient scale for training dialogue systems. To address this gap, we introduce DiaSynth - a synthetic dialogue generation framework capable of generating high-quality, contextually rich dialogues across a wide range of domains. Unlike existing frameworks, DiaSynth uses Large Language Models (LLMs) and Chain of Thought (CoT) reasoning to generate dynamic, domain-specific dialogues with simulated personas and diverse conversational features. We perform our experiments by generating synthetic data using different LLMs and few-shot examples from DialogSum and SAMSum. The pretrained language models fine-tuned on the synthetic data outperform the base models by 16.47% on dialogue summarization, while the comparison between models fine-tuned on in-domain data and synthetic data shows that the synthetic data is able to capture 90.48% of the performance distribution of the in-domain data on dialogue summarization. The quality of the data generated also increases as we increase the size of LLM from 3B to 8B. These results validate DiaSynth's potential as a robust alternative to traditional data collection methods. We open source the code and data generated for future research.
comment: 13 pages, 1 figure
♻ ☆ FineFake: A Knowledge-Enriched Dataset for Fine-Grained Multi-Domain Fake News Detection
Existing benchmarks for fake news detection have significantly contributed to the advancement of models in assessing the authenticity of news content. However, these benchmarks typically focus solely on news pertaining to a single semantic topic or originating from a single platform, thereby failing to capture the diversity of multi-domain news in real scenarios. In order to understand fake news across various domains, the external knowledge and fine-grained annotations are indispensable to provide precise evidence and uncover the diverse underlying strategies for fabrication, which are also ignored by existing benchmarks. To address this gap, we introduce a novel multi-domain knowledge-enhanced benchmark with fine-grained annotations, named \textbf{FineFake}. FineFake encompasses 16,909 data samples spanning six semantic topics and eight platforms. Each news item is enriched with multi-modal content, potential social context, semi-manually verified common knowledge, and fine-grained annotations that surpass conventional binary labels. Furthermore, we formulate three challenging tasks based on FineFake and propose a knowledge-enhanced domain adaptation network. Extensive experiments are conducted on FineFake under various scenarios, providing accurate and reliable benchmarks for future endeavors. The entire FineFake project is publicly accessible as an open-source repository at \url{https://github.com/Accuser907/FineFake}.
♻ ☆ A Corpus for Named Entity Recognition in Chinese Novels with Multi-genres
Entities like person, location, organization are important for literary text analysis. The lack of annotated data hinders the progress of named entity recognition (NER) in literary domain. To promote the research of literary NER, we build the largest multi-genre literary NER corpus containing 263,135 entities in 105,851 sentences from 260 online Chinese novels spanning 13 different genres. Based on the corpus, we investigate characteristics of entities from different genres. We propose several baseline NER models and conduct cross-genre and cross-domain experiments. Experimental results show that genre difference significantly impact NER performance though not as much as domain difference like literary domain and news domain. Compared with NER in news domain, literary NER still needs much improvement and the Out-of-Vocabulary (OOV) problem is more challenging due to the high variety of entities in literary works. Our data and models are open-sourced at https://github.com/hjzhao73/MultiGenre-ChineseNovel
♻ ☆ DIRAS: Efficient LLM Annotation of Document Relevance in Retrieval Augmented Generation
Retrieval Augmented Generation (RAG) is widely employed to ground responses to queries on domain-specific documents. But do RAG implementations leave out important information when answering queries that need an integrated analysis of information (e.g., Tell me good news in the stock market today.)? To address these concerns, RAG developers need to annotate information retrieval (IR) data for their domain of interest, which is challenging because (1) domain-specific queries usually need nuanced definitions of relevance beyond shallow semantic relevance; and (2) human or GPT-4 annotation is costly and cannot cover all (query, document) pairs (i.e., annotation selection bias), thus harming the effectiveness in evaluating IR recall. To address these challenges, we propose DIRAS (Domain-specific Information Retrieval Annotation with Scalability), a manual-annotation-free schema that fine-tunes open-sourced LLMs to consider nuanced relevance definition and annotate (partial) relevance labels with calibrated relevance scores. Extensive evaluation shows that DIRAS enables smaller (8B) LLMs to achieve GPT-4-level performance on annotating and ranking unseen (query, document) pairs, and is helpful for real-world RAG development. All code, LLM generations, and human annotations can be found in \url{https://github.com/EdisonNi-hku/DIRAS}.
♻ ☆ Introducing MeMo: A Multimodal Dataset for Memory Modelling in Multiparty Conversations
Conversational memory is the process by which humans encode, retain and retrieve verbal, non-verbal and contextual information from a conversation. Since human memory is selective, differing recollections of the same events can lead to misunderstandings and misalignments within a group. Yet, conversational facilitation systems, aimed at advancing the quality of group interactions, usually focus on tracking users' states within an individual session, ignoring what remains in each participant's memory after the interaction. Understanding conversational memory can be used as a source of information on the long-term development of social connections within a group. This paper introduces the MeMo corpus, the first conversational dataset annotated with participants' memory retention reports, aimed at facilitating computational modelling of human conversational memory. The MeMo corpus includes 31 hours of small-group discussions on Covid-19, repeated 3 times over the term of 2 weeks. It integrates validated behavioural and perceptual measures, audio, video, and multimodal annotations, offering a valuable resource for studying and modelling conversational memory and group dynamics. By introducing the MeMo corpus, analysing its validity, and demonstrating its usefulness for future research, this paper aims to pave the way for future research in conversational memory modelling for intelligent system development.
♻ ☆ Are Large Language Models Strategic Decision Makers? A Study of Performance and Bias in Two-Player Non-Zero-Sum Games
Large Language Models (LLMs) have been increasingly used in real-world settings, yet their strategic decision-making abilities remain largely unexplored. To fully benefit from the potential of LLMs, it's essential to understand their ability to function in complex social scenarios. Game theory, which is already used to understand real-world interactions, provides a good framework for assessing these abilities. This work investigates the performance and merits of LLMs in canonical game-theoretic two-player non-zero-sum games, Stag Hunt and Prisoner Dilemma. Our structured evaluation of GPT-3.5, GPT-4-Turbo, GPT-4o, and Llama-3-8B shows that these models, when making decisions in these games, are affected by at least one of the following systematic biases: positional bias, payoff bias, or behavioural bias. This indicates that LLMs do not fully rely on logical reasoning when making these strategic decisions. As a result, it was found that the LLMs' performance drops when the game configuration is misaligned with the affecting biases. When misaligned, GPT-3.5, GPT-4-Turbo, GPT-4o, and Llama-3-8B show an average performance drop of 32\%, 25\%, 34\%, and 29\% respectively in Stag Hunt, and 28\%, 16\%, 34\%, and 24\% respectively in Prisoner's Dilemma. Surprisingly, GPT-4o (a top-performing LLM across standard benchmarks) suffers the most substantial performance drop, suggesting that newer models are not addressing these issues. Interestingly, we found that a commonly used method of improving the reasoning capabilities of LLMs, chain-of-thought (CoT) prompting, reduces the biases in GPT-3.5, GPT-4o, and Llama-3-8B but increases the effect of the bias in GPT-4-Turbo, indicating that CoT alone cannot fully serve as a robust solution to this problem. We perform several additional experiments, which provide further insight into these observed behaviours.
♻ ☆ Unraveling the Mechanics of Learning-Based Demonstration Selection for In-Context Learning
Large Language Models (LLMs) have demonstrated impressive in-context learning (ICL) capabilities from few-shot demonstration exemplars. While recent learning-based demonstration selection methods have proven beneficial to ICL by choosing more useful exemplars, their underlying mechanisms are opaque, hindering efforts to address limitations such as high training costs and poor generalization across tasks. These methods generally assume the selection process captures similarities between the exemplar and the target instance, however, it remains unknown what kinds of similarities are captured and vital to performing ICL. To dive into this question, we analyze the working mechanisms of the learning-based demonstration selection methods and empirically identify two important factors related to similarity measurement: 1) The ability to integrate different levels of task-agnostic text similarities between the input of exemplars and test cases enhances generalization power across different tasks. 2) Incorporating task-specific labels when measuring the similarities significantly improves the performance on each specific task. We validate these two findings through extensive quantitative and qualitative analyses across ten datasets and various LLMs. Based on our findings, we introduce two effective yet simplified exemplar selection methods catering to task-agnostic and task-specific demands, eliminating the costly LLM inference overhead.
comment: 17 pages, 7 figures and 9 tables
♻ ☆ Mathador-LM: A Dynamic Benchmark for Mathematical Reasoning on Large Language Models EMNLP 2024
We introduce Mathador-LM, a new benchmark for evaluating the mathematical reasoning on large language models (LLMs), combining ruleset interpretation, planning, and problem-solving. This benchmark is inspired by the Mathador game, where the objective is to reach a target number using basic arithmetic operations on a given set of base numbers, following a simple set of rules. We show that, across leading LLMs, we obtain stable average performance while generating benchmark instances \emph{dynamically}, following a target difficulty level. Thus, our benchmark alleviates concerns about test-set leakage into training data, an issue that often undermines popular benchmarks. Additionally, we conduct a comprehensive evaluation of both open and closed-source state-of-the-art LLMs on Mathador-LM. Our findings reveal that contemporary models struggle with Mathador-LM, scoring significantly lower than average 3rd graders. This stands in stark contrast to their strong performance on popular mathematical reasoning benchmarks. The implementation of Mathador-LM benchmark is available at \href{https://github.com/IST-DASLab/Mathador-LM}{github.com/IST-DASLab/Mathador-LM}.
comment: EMNLP 2024
♻ ☆ Logit Separability-Driven Samples and Multiple Class-Related Words Selection for Advancing In-Context Learning
Effective organization of in-context learning (ICL) demonstrations is key to improving the quality of large language model (LLM) responses. To create better sample-label pairs that instruct LLM understanding, we introduce logit separability, a criterion to assess the clarity of both samples and class-related words at the logit level. This facilitates the optimization of sample and label selection, enhancing the precision of information provided in ICL demonstrations. Additionally, we find that incorporating multiple class-related words for each sample, rather than relying on a single class name, improves performance by offering a broader range of label information. Building on these insights, we propose LICL, a logit separability-based method that jointly organizes samples and integrates multiple class-related words into each sample-label pair. Evaluations across seven classification datasets show that this approach significantly improves ICL performance by providing clearer instructions and richer label information.
comment: 20 pages, 10 figures
♻ ☆ Dissecting Fine-Tuning Unlearning in Large Language Models EMNLP 2024
Fine-tuning-based unlearning methods prevail for preventing targeted harmful, sensitive, or copyrighted information within large language models while preserving overall capabilities. However, the true effectiveness of these methods is unclear. In this work, we delve into the limitations of fine-tuning-based unlearning through activation patching and parameter restoration experiments. Our findings reveal that these methods alter the model's knowledge retrieval process, providing further evidence that they do not genuinely erase the problematic knowledge embedded in the model parameters. Instead, the coefficients generated by the MLP components in the model's final layer are the primary contributors to these seemingly positive unlearning effects, playing a crucial role in controlling the model's behaviors. Furthermore, behavioral tests demonstrate that this unlearning mechanism inevitably impacts the global behavior of the models, affecting unrelated knowledge or capabilities. The code is released at https://github.com/yihuaihong/Dissecting-FT-Unlearning.
comment: Accepted in EMNLP 2024 Main (Short paper)
♻ ☆ Learning Planning-based Reasoning by Trajectories Collection and Process Reward Synthesizing EMNLP 2024
Large Language Models (LLMs) have demonstrated significant potential in handling complex reasoning tasks through step-by-step rationale generation. However, recent studies have raised concerns regarding the hallucination and flaws in their reasoning process. Substantial efforts are being made to improve the reliability and faithfulness of the generated rationales. Some approaches model reasoning as planning, while others focus on annotating for process supervision. Nevertheless, the planning-based search process often results in high latency due to the frequent assessment of intermediate reasoning states and the extensive exploration space. Additionally, supervising the reasoning process with human annotation is costly and challenging to scale for LLM training. To address these issues, in this paper, we propose a framework to learn planning-based reasoning through Direct Preference Optimization (DPO) on collected trajectories, which are ranked according to synthesized process rewards. Our results on challenging logical reasoning benchmarks demonstrate the effectiveness of our learning framework, showing that our 7B model can surpass the strong counterparts like GPT-3.5-Turbo.
comment: 17 pages, 9 figures. EMNLP 2024
♻ ☆ The Generation Gap: Exploring Age Bias in the Value Systems of Large Language Models
We explore the alignment of values in Large Language Models (LLMs) with specific age groups, leveraging data from the World Value Survey across thirteen categories. Through a diverse set of prompts tailored to ensure response robustness, we find a general inclination of LLM values towards younger demographics, especially when compared to the US population. Although a general inclination can be observed, we also found that this inclination toward younger groups can be different across different value categories. Additionally, we explore the impact of incorporating age identity information in prompts and observe challenges in mitigating value discrepancies with different age cohorts. Our findings highlight the age bias in LLMs and provide insights for future work. Materials for our analysis are available at \url{ https://github.com/MichiganNLP/Age-Bias-In-LLMs}
comment: 5 pages
♻ ☆ Efficient Data Generation for Source-grounded Information-seeking Dialogs: A Use Case for Meeting Transcripts
Automating data generation with Large Language Models (LLMs) has become increasingly popular. In this work, we investigate the feasibility and effectiveness of LLM-based data generation in the challenging setting of source-grounded information-seeking dialogs, with response attribution, over long documents. Our source texts consist of long and noisy meeting transcripts, adding to the task complexity. Since automating attribution remains difficult, we propose a semi-automatic approach: dialog queries and responses are generated with LLMs, followed by human verification and identification of attribution spans. Using this approach, we created MISeD -- Meeting Information Seeking Dialogs dataset -- a dataset of information-seeking dialogs focused on meeting transcripts. Models finetuned with MISeD demonstrate superior performance compared to off-the-shelf models, even those of larger size. Finetuning on MISeD gives comparable response generation quality to finetuning on fully manual data, while improving attribution quality and reducing time and effort.
♻ ☆ NatLan: Native Language Prompting Facilitates Knowledge Elicitation Through Language Trigger Provision and Domain Trigger Retention
Multilingual large language models (MLLMs) do not perform as well when answering questions in non-dominant languages as they do in their dominant languages. Although existing translate-then-answer methods alleviate this issue, the mechanisms behind their effectiveness remain unclear. In this study, we analogize the dominant language of MLLMs to the native language of humans and use two human cognitive features: the Language Trigger (LT) and the Domain Trigger (DT), to interpret the mechanisms behind translate-then-answer methods. This reveals that while sufficient LTs are provided by these methods, there remains a deficiency in DT retention. To mitigate this issue, we propose Native Language Prompting (NatLan), employing a Multi-MLLM collaboration strategy and introducing an additional role-enhanced domain-specific MLLM with stronger multilingual understanding capabilities as the translator. Across five language QA benchmarks, NatLan achieves up to a 31.28% improvement in accuracy and, compared to existing state-of-the-art methods, provides comparable or greater retention of DTs in up to 87% of cases. Our code is available at https://github.com/AnonyNLP/NatLan.
♻ ☆ Achieving >97% on GSM8K: Deeply Understanding the Problems Makes LLMs Better Solvers for Math Word Problems
Chain-of-Thought (CoT) prompting has enhanced the performance of Large Language Models (LLMs) across various reasoning tasks. However, CoT still falls short in dealing with complex math word problems, as it usually suffers from three pitfalls: semantic misunderstanding errors, calculation errors, and step-missing errors. Prior studies involve addressing the calculation errors and step-missing errors, but neglect the semantic misunderstanding errors, which is the major factor limiting the reasoning performance of LLMs. To this end, we propose a simple-yet-effective method, namely Deeply Understanding the Problems (DUP), to improve the LLMs' math problem-solving ability by addressing semantic misunderstanding errors. The core of our method is to encourage the LLMs to deeply understand the problems and extract the key problem-solving information used for better reasoning. Extensive experiments on 10 diverse reasoning benchmarks show that our DUP method consistently outperforms the other counterparts by a large margin. More encouragingly, DUP achieves a new SOTA result on the GSM8K benchmark, with an accuracy of 97.1% under the zero-shot setting.
comment: Work in progress
♻ ☆ MTL-LoRA: Low-Rank Adaptation for Multi-Task Learning
Parameter-efficient fine-tuning (PEFT) has been widely employed for domain adaptation, with LoRA being one of the most prominent methods due to its simplicity and effectiveness. However, in multi-task learning (MTL) scenarios, LoRA tends to obscure the distinction between tasks by projecting sparse high-dimensional features from different tasks into the same dense low-dimensional intrinsic space. This leads to task interference and suboptimal performance for LoRA and its variants. To tackle this challenge, we propose MTL-LoRA, which retains the advantages of low-rank adaptation while significantly enhancing multi-task learning capabilities. MTL-LoRA augments LoRA by incorporating additional task-adaptive parameters that differentiate task-specific information and effectively capture shared knowledge across various tasks within low-dimensional spaces. This approach enables large language models (LLMs) pre-trained on general corpus to adapt to different target task domains with a limited number of trainable parameters. Comprehensive experimental results, including evaluations on public academic benchmarks for natural language understanding, commonsense reasoning, and image-text understanding, as well as real-world industrial text Ads relevance datasets, demonstrate that MTL-LoRA outperforms LoRA and its various variants with comparable or even fewer learnable parameters in multitask learning.
comment: 12 Pages, 4 Figures
♻ ☆ Ouroboros: Generating Longer Drafts Phrase by Phrase for Faster Speculative Decoding EMNLP 2024
Speculative decoding is a widely used method that accelerates the generation process of large language models (LLMs) with no compromise in model performance. It achieves this goal by using an existing smaller model for drafting and then employing the target LLM to verify the draft in a low-cost parallel manner. Under such a drafting-verification framework, drafting efficiency has become a bottleneck in the final speedup of speculative decoding. Therefore, generating longer drafts at less cost can lead to better decoding speedup. To achieve this, we introduce Ouroboros, which can generate draft phrases to parallelize the drafting process and meanwhile lengthen drafts in a training-free manner. The experimental results on various typical text generation tasks show that Ouroboros can achieve speedups of up to $2.8\times$ over speculative decoding and $3.9\times$ over vanilla decoding, without fine-tuning draft and target models. The source code of Ouroboros is available at https://github.com/thunlp/Ouroboros.
comment: EMNLP 2024
♻ ☆ Do LLMs have Consistent Values?
Large Language Models (LLM) technology is constantly improving towards human-like dialogue. Values are a basic driving force underlying human behavior, but little research has been done to study the values exhibited in text generated by LLMs. Here we study this question by turning to the rich literature on value structure in psychology. We ask whether LLMs exhibit the same value structure that has been demonstrated in humans, including the ranking of values, and correlation between values. We show that the results of this analysis depend on how the LLM is prompted, and that under a particular prompting strategy (referred to as "Value Anchoring") the agreement with human data is quite compelling. Our results serve both to improve our understanding of values in LLMs, as well as introduce novel methods for assessing consistency in LLM responses.
comment: 16 pages, 4 figures, and there are more in the appendix
♻ ☆ The Buffer Mechanism for Multi-Step Information Reasoning in Language Models
Large language models have consistently struggled with complex reasoning tasks, such as mathematical problem-solving. Investigating the internal reasoning mechanisms of these models can help us design better model architectures and training strategies, ultimately enhancing their reasoning capability. In this study, we constructed a symbolic dataset to investigate the mechanisms by which Transformer models employ vertical thinking strategy based on their inherent structure and horizontal thinking strategy based on Chain of Thought to achieve multi-step reasoning. We introduced the concept of buffer mechanism: the model stores various information in distinct buffers and selectively extracts them through the query-key matrix. We proposed a random matrix-based algorithm to enhance the model's reasoning ability, resulting in a 75% reduction in the training time required for the GPT-2 model to achieve generalization capability on the PrOntoQA dataset. These findings provide new insights into understanding the mechanisms of large language models.
♻ ☆ WaterSeeker: Pioneering Efficient Detection of Watermarked Segments in Large Documents
Watermarking algorithms for large language models (LLMs) have attained high accuracy in detecting LLM-generated text. However, existing methods primarily focus on distinguishing fully watermarked text from non-watermarked text, overlooking real-world scenarios where LLMs generate only small sections within large documents. In this scenario, balancing time complexity and detection performance poses significant challenges. This paper presents WaterSeeker, a novel approach to efficiently detect and locate watermarked segments amid extensive natural text. It first applies an efficient anomaly extraction method to preliminarily locate suspicious watermarked regions. Following this, it conducts a local traversal and performs full-text detection for more precise verification. Theoretical analysis and experimental results demonstrate that WaterSeeker achieves a superior balance between detection accuracy and computational efficiency. Moreover, WaterSeeker's localization ability supports the development of interpretable AI detection systems. This work pioneers a new direction in watermarked segment detection, facilitating more reliable AI-generated content identification.Our code is available at https://github.com/THU-BPM/WaterSeeker.
comment: 20 pages, 7 figures, 8 tables
♻ ☆ Generate then Refine: Data Augmentation for Zero-shot Intent Detection
In this short paper we propose a data augmentation method for intent detection in zero-resource domains. Existing data augmentation methods rely on few labelled examples for each intent category, which can be expensive in settings with many possible intents. We use a two-stage approach: First, we generate utterances for intent labels using an open-source large language model in a zero-shot setting. Second, we develop a smaller sequence-to-sequence model (the Refiner), to improve the generated utterances. The Refiner is fine-tuned on seen domains and then applied to unseen domains. We evaluate our method by training an intent classifier on the generated data, and evaluating it on real (human) data. We find that the Refiner significantly improves the data utility and diversity over the zero-shot LLM baseline for unseen domains and over common baseline approaches. Our results indicate that a two-step approach of a generative LLM in zero-shot setting and a smaller sequence-to-sequence model can provide high-quality data for intent detection.
♻ ☆ A Character-Centric Creative Story Generation via Imagination
Creative story generation has long been a goal of NLP research. While existing methodologies have aimed to generate long and coherent stories, they fall significantly short of human capabilities in terms of diversity and character depth. To address this, we introduce a novel story generation framework called CCI (Character-centric Creative story generation via Imagination). CCI features two modules for creative story generation: IG (Image-Guided Imagination) and MW (Multi-Writer model). In the IG module, we utilize a text-to-image model to create visual representations of key story elements, such as characters, backgrounds, and main plots, in a more novel and concrete manner than text-only approaches. The MW module uses these story elements to generate multiple persona-description candidates and selects the best one to insert into the story, thereby enhancing the richness and depth of the narrative. We compared the stories generated by CCI and baseline models through statistical analysis, as well as human and LLM evaluations. The results showed that the IG and MW modules significantly improve various aspects of the stories' creativity. Furthermore, our framework enables interactive multi-modal story generation with users, opening up new possibilities for human-LLM integration in cultural development. Project page : https://www.2024cci.p-e.kr/
♻ ☆ A Training-free Sub-quadratic Cost Transformer Model Serving Framework With Hierarchically Pruned Attention
In modern large language models (LLMs), increasing the context length is crucial for improving comprehension and coherence in long-context, multi-modal, and retrieval-augmented language generation. While many recent transformer models attempt to extend their context length over a million tokens, they remain impractical due to the quadratic time and space complexities. Although recent works on linear and sparse attention mechanisms can achieve this goal, their real-world applicability is often limited by the need to re-train from scratch and significantly worse performance. In response, we propose a novel approach, Hierarchically Pruned Attention (HiP), which reduces the time complexity of the attention mechanism to $O(T \log T)$ and the space complexity to $O(T)$, where $T$ is the sequence length. We notice a pattern in the attention scores of pretrained LLMs where tokens close together tend to have similar scores, which we call ``attention locality''. Based on this observation, we utilize a novel tree-search-like algorithm that estimates the top-$k$ key tokens for a given query on the fly, which is mathematically guaranteed to have better performance than random attention pruning. In addition to improving the time complexity of the attention mechanism, we further optimize GPU memory usage by implementing KV cache offloading, which stores only $O(\log T)$ tokens on the GPU while maintaining similar decoding throughput. Experiments on benchmarks show that HiP, with its training-free nature, significantly reduces both prefill and decoding latencies, as well as memory usage, while maintaining high-quality generation with minimal degradation. HiP enables pretrained LLMs to scale up to millions of tokens on commodity GPUs, potentially unlocking long-context LLM applications previously deemed infeasible.
comment: 44 pages
♻ ☆ Fostering Natural Conversation in Large Language Models with NICO: a Natural Interactive COnversation dataset
Benefiting from diverse instruction datasets, contemporary Large Language Models (LLMs) perform effectively as AI assistants in collaborating with humans. However, LLMs still struggle to generate natural and colloquial responses in real-world applications such as chatbots and psychological counseling that require more human-like interactions. To address these limitations, we introduce NICO, a Natural Interactive COnversation dataset in Chinese. We first use GPT-4-turbo to generate dialogue drafts and make them cover 20 daily-life topics and 5 types of social interactions. Then, we hire workers to revise these dialogues to ensure that they are free of grammatical errors and unnatural utterances. We define two dialogue-level natural conversation tasks and two sentence-level tasks for identifying and rewriting unnatural sentences. Multiple open-source and closed-source LLMs are tested and analyzed in detail. The experimental results highlight the challenge of the tasks and demonstrate how NICO can help foster the natural dialogue capabilities of LLMs. The dataset will be released.
comment: 16 pages, 3 figures, 10 tables
♻ ☆ Impeding LLM-assisted Cheating in Introductory Programming Assignments via Adversarial Perturbation
While Large language model (LLM)-based programming assistants such as CoPilot and ChatGPT can help improve the productivity of professional software developers, they can also facilitate cheating in introductory computer programming courses. Assuming instructors have limited control over the industrial-strength models, this paper investigates the baseline performance of 5 widely used LLMs on a collection of introductory programming problems, examines adversarial perturbations to degrade their performance, and describes the results of a user study aimed at understanding the efficacy of such perturbations in hindering actual code generation for introductory programming assignments. The user study suggests that i) perturbations combinedly reduced the average correctness score by 77%, ii) the drop in correctness caused by these perturbations was affected based on their detectability.
♻ ☆ Cofca: A Step-Wise Counterfactual Multi-hop QA benchmark
While Large Language Models (LLMs) excel in question-answering (QA) tasks, their real reasoning abilities on multiple evidence retrieval and integration on Multi-hop QA tasks remain less explored. Firstly, LLMs sometimes generate answers that rely on internal memory rather than retrieving evidence and reasoning in the given context, which brings concerns about the evaluation quality of real reasoning abilities. Although previous counterfactual QA benchmarks can separate the internal memory of LLMs, they focus solely on final QA performance, which is insufficient for reporting LLMs' real reasoning abilities. Because LLMs are expected to engage in intricate reasoning processes that involve evidence retrieval and answering a series of sub-questions from given passages. Moreover, current factual Multi-hop QA (MHQA) benchmarks are annotated on open-source corpora such as Wikipedia, although useful for multi-step reasoning evaluation, they show limitations due to the potential data contamination in LLMs' pre-training stage. To address these issues, we introduce a Step-wise Counterfactual benchmark (CofCA), a novel evaluation benchmark consisting of factual data and counterfactual data that reveals LLMs' real reasoning abilities on multi-step reasoning and reasoning chain evaluation. Our experimental results reveal a significant performance gap of several LLMs between Wikipedia-based factual data and counterfactual data, deeming data contamination issues in existing benchmarks. Moreover, we observe that LLMs usually bypass the correct reasoning chain, showing an inflated multi-step reasoning performance. We believe that our CofCA benchmark will enhance and facilitate the evaluations of trustworthy LLMs.
♻ ☆ Resolving Knowledge Conflicts in Large Language Models
Large language models (LLMs) often encounter knowledge conflicts, scenarios where discrepancy arises between the internal parametric knowledge of LLMs and non-parametric information provided in the prompt context. In this work we ask what are the desiderata for LLMs when a knowledge conflict arises and whether existing LLMs fulfill them. We posit that LLMs should 1) identify knowledge conflicts, 2) pinpoint conflicting information segments, and 3) provide distinct answers or viewpoints in conflicting scenarios. To this end, we introduce an evaluation framework for simulating contextual knowledge conflicts and quantitatively evaluating to what extent LLMs achieve these goals. It includes diverse and complex situations of knowledge conflict, knowledge from diverse entities and domains, two synthetic conflict creation methods, and settings with progressively increasing difficulty to reflect realistic knowledge conflicts. Extensive experiments with the framework reveal that while LLMs perform well in identifying the existence of knowledge conflicts, they struggle to determine the specific conflicting knowledge and produce a response with distinct answers amidst conflicting information. To address these challenges, we propose new instruction-based approaches that augment LLMs to better achieve the three goals. Further analysis shows that abilities to tackle knowledge conflicts are greatly impacted by factors such as knowledge domain, while generating robust responses to knowledge conflict scenarios remains an open research question.
comment: Published at COLM 2024
♻ ☆ EVA-Score: Evaluating Abstractive Long-form Summarization on Informativeness through Extraction and Validation
Since LLMs emerged, more attention has been paid to abstractive long-form summarization, where longer input sequences indicate more information contained. Nevertheless, the automatic evaluation of such summaries remains underexplored. The current evaluation metrics for long-form summarization either use similarity-based metrics like ROUGE and BERTScore or LLM-based metrics using appropriate prompts or pre-defined schema. We argue that the former only relies on similarity and fails to consider informativeness while the latter lacks quantitative analysis of informative richness, and is rather subjective and hard to explain. Current evaluation metrics either use traditional metrics like ROUGE and BERTScore, which rely on surface-level similarity and fail to consider informativeness, or simple LLM-based metrics, which are not robust and easily overwhelmed by the long contexts. In this paper, we propose a new evaluation metric called EVA-Score to extract all information from the given summaries, identify overlapped information based on reference, and calculate the information score. We test EVA-Score on several datasets and the experimental results reveal that EVA-Score shows the highest correlation with humans. We also re-evaluate the performance of LLMs on long-form summarization from the information perspective. The results indicate that responses of LLMs still have a gap with the human-written answers. Moreover, we provide a detailed analysis of the effectiveness of EVA-Score, forecasting future ways to automatically evaluate abstractive long-form summarization.
comment: 20 pages
♻ ☆ Rethinking Reward Model Evaluation: Are We Barking up the Wrong Tree?
Reward Models (RMs) are crucial for aligning language models with human preferences. Currently, the evaluation of RMs depends on measuring accuracy against a validation set of manually annotated preference data. Although this method is straightforward and widely adopted, the relationship between RM accuracy and downstream policy performance remains under-explored. In this work, we conduct experiments in a synthetic setting to investigate how differences in RM measured by accuracy translate into gaps in optimized policy performance. Our findings reveal that while there is a weak positive correlation between accuracy and downstream performance, policies optimized towards RMs with similar accuracy can exhibit quite different performance. Moreover, we discover that the way of measuring accuracy significantly impacts its ability to predict the final policy performance. Through the lens of Regressional Goodhart's effect, we identify the existence of exogenous variables impacting the relationship between RM quality measured by accuracy and policy model capability. This underscores the inadequacy of relying solely on accuracy to reflect their impact on policy optimization.
♻ ☆ DriVLMe: Enhancing LLM-based Autonomous Driving Agents with Embodied and Social Experiences IROS
Recent advancements in foundation models (FMs) have unlocked new prospects in autonomous driving, yet the experimental settings of these studies are preliminary, over-simplified, and fail to capture the complexity of real-world driving scenarios in human environments. It remains under-explored whether FM agents can handle long-horizon navigation tasks with free-from dialogue and deal with unexpected situations caused by environmental dynamics or task changes. To explore the capabilities and boundaries of FMs faced with the challenges above, we introduce DriVLMe, a video-language-model-based agent to facilitate natural and effective communication between humans and autonomous vehicles that perceive the environment and navigate. We develop DriVLMe from both embodied experiences in a simulated environment and social experiences from real human dialogue. While DriVLMe demonstrates competitive performance in both open-loop benchmarks and closed-loop human studies, we reveal several limitations and challenges, including unacceptable inference time, imbalanced training data, limited visual understanding, challenges with multi-turn interactions, simplified language generation from robotic experiences, and difficulties in handling on-the-fly unexpected situations like environmental dynamics and task changes.
comment: 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
♻ ☆ Investigating Annotator Bias in Large Language Models for Hate Speech Detection NeurIPS
Data annotation, the practice of assigning descriptive labels to raw data, is pivotal in optimizing the performance of machine learning models. However, it is a resource-intensive process susceptible to biases introduced by annotators. The emergence of sophisticated Large Language Models (LLMs) presents a unique opportunity to modernize and streamline this complex procedure. While existing research extensively evaluates the efficacy of LLMs, as annotators, this paper delves into the biases present in LLMs when annotating hate speech data. Our research contributes to understanding biases in four key categories: gender, race, religion, and disability with four LLMs: GPT-3.5, GPT-4o, Llama-3.1 and Gemma-2. Specifically targeting highly vulnerable groups within these categories, we analyze annotator biases. Furthermore, we conduct a comprehensive examination of potential factors contributing to these biases by scrutinizing the annotated data. We introduce our custom hate speech detection dataset, HateBiasNet, to conduct this research. Additionally, we perform the same experiments on the ETHOS (Mollas et al. 2022) dataset also for comparative analysis. This paper serves as a crucial resource, guiding researchers and practitioners in harnessing the potential of LLMs for data annotation, thereby fostering advancements in this critical field.
comment: Accepted at NeurIPS Safe Generative AI Workshop, 2024
♻ ☆ Multi-Layer Transformers Gradient Can be Approximated in Almost Linear Time
The computational complexity of the self-attention mechanism in popular transformer architectures poses significant challenges for training and inference, and becomes the bottleneck for long inputs. Is it possible to significantly reduce the quadratic time complexity of computing the gradients in multi-layer transformer models? This paper proves that a novel fast approximation method can calculate the gradients in almost linear time $n^{1+o(1)}$ where $n$ is the input sequence length, while it maintains a polynomially small approximation error $1 / \mathrm{poly}(n)$ across the entire model. Our theory holds for general loss functions and when the multi-layer transformer model contains many practical sub-modules, such as residual connection, casual mask, and multi-head attention. By improving the efficiency of gradient computation, we hope that this work will facilitate more effective training and deployment of long-context language models based on our theoretical results.
♻ ☆ RoCoFT: Efficient Finetuning of Large Language Models with Row-Column Updates
We propose RoCoFT, a parameter-efficient fine-tuning method for large-scale language models (LMs) based on updating only a few rows and columns of the weight matrices in transformers. Through extensive experiments with medium-size LMs like BERT and RoBERTa, and larger LMs like Bloom-7B, Llama2-7B, and Llama2-13B, we show that our method gives comparable or better accuracies than state-of-art PEFT methods while also being more memory and computation-efficient. We also study the reason behind the effectiveness of our method with tools from neural tangent kernel theory. We empirically demonstrate that our kernel, constructed using a restricted set of row and column parameters, are numerically close to the full-parameter kernel and gives comparable classification performance. Ablation studies are conducted to investigate the impact of different algorithmic choices, including the selection strategy for rows and columns as well as the optimal rank for effective implementation of our method.
comment: RoCoFT is a parameter-efficient method
♻ ☆ Beyond Numeric Awards: In-Context Dueling Bandits with LLM Agents
In-context decision-making is an important capability of artificial general intelligence, which Large Language Models (LLMs) have effectively demonstrated in various scenarios. However, LLMs often face challenges when dealing with numerical contexts, and limited attention has been paid to evaluating their performance through preference feedback generated by the environment. This paper is the first to investigate the performance of LLMs as decision-makers in the context of Dueling Bandits (DB). We compare GPT-3.5 Turbo, GPT-4, GPT-4 Turbo, Llama 3.1, and o1-preview against eight well-established DB algorithms. Our results reveal that LLMs, particularly GPT-4 Turbo, quickly identify the Condorcet winner, thus outperforming existing state-of-the-art algorithms in terms of weak regret. Nevertheless, LLMs struggle to converge even when explicitly prompted to do so and are sensitive to prompt variations. To overcome these issues, we introduce a hybrid algorithm: LLM-Enhanced Adaptive Dueling (LEAD), which takes advantage of both in-context decision-making capabilities of LLMs and theoretical guarantees inherited from classic DB algorithms. We show that LEAD has theoretical guarantees on both weak and strong regret and validate its robustness even with noisy and adversarial prompts. The design of such an algorithm sheds light on how to enhance trustworthiness for LLMs used in decision-making tasks where performance robustness matters.
♻ ☆ Training Language Models on Synthetic Edit Sequences Improves Code Synthesis
Software engineers mainly write code by editing existing programs. In contrast, large language models (LLMs) autoregressively synthesize programs in a single pass. One explanation for this is the scarcity of open-sourced edit data. While high-quality instruction data for code synthesis is already scarce, high-quality edit data is even scarcer. To fill this gap, we develop a synthetic data generation algorithm called LintSeq. This algorithm refactors existing code into a sequence of code edits by using a linter to procedurally sample across the error-free insertions that can be used to sequentially write programs. It outputs edit sequences as text strings consisting of consecutive program diffs. To test LintSeq, we use it to refactor a dataset of instruction + program pairs into instruction + program-diff-sequence tuples. Then, we instruction finetune a series of smaller LLMs ranging from 2.6B to 14B parameters on both the re-factored and original versions of this dataset, comparing zero-shot performance on code synthesis benchmarks. We show that during repeated sampling, edit sequence finetuned models produce more diverse programs than baselines. This results in better inference-time scaling for benchmark coverage as a function of samples, i.e. the fraction of problems "pass@k" solved by any attempt given "k" tries. For example, on HumanEval pass@50, small LLMs finetuned on synthetic edit sequences are competitive with GPT-4 and outperform models finetuned on the baseline dataset by +20% (+/-3%) in absolute score. Finally, we also pretrain our own tiny LMs for code understanding. We show that finetuning tiny models on synthetic code edits results in state-of-the-art code synthesis for the on-device model class. Our 150M parameter edit sequence LM matches or outperforms code models with twice as many parameters, both with and without repeated sampling, including Codex and AlphaCode.
♻ ☆ BiMix: Bivariate Data Mixing Law for Language Model Pretraining
Large language models have demonstrated remarkable capabilities across various tasks, primarily attributed to the utilization of diversely sourced data. However, the impact of pretraining data composition on model performance remains poorly understood. This paper introduces $\textbf{BiMix}$, a novel bivariate data mixing law that models the joint scaling behavior of domain proportions and data volume in LLM pretraining. $\textbf{BiMix}$ provides a systematic framework for understanding and optimizing data mixtures across diverse domains. Through extensive experiments on two large-scale datasets, we demonstrate $\textbf{BiMix}$'s high accuracy in loss extrapolation (mean relative error < 0.2%) and its generalization to unseen mixtures (R${}^{2}$ > 0.97). Optimization of domain proportions yields superior model performance compared to existing methods. Furthermore, we establish entropy-based measures as efficient proxies for data mixing, offering a computationally lightweight strategy. Our work contributes both theoretical insights into data mixing dynamics and practical tools for enhancing LLM training efficiency, paving the way for more effective scaling strategies in language model development.
comment: reorganize & new exps
♻ ☆ LongLaMP: A Benchmark for Personalized Long-form Text Generation
Long-text generation is seemingly ubiquitous in real-world applications of large language models such as generating an email or writing a review. Despite the fundamental importance and prevalence of long-text generation in many practical applications, existing work on personalized generation has focused on the generation of very short text. To overcome these limitations, we study the problem of personalized long-text generation, that is, generating long-text that is personalized for a specific user while being practically useful for the vast majority of real-world applications that naturally require the generation of longer text. In this work, we demonstrate the importance of user-specific personalization for long-text generation tasks and develop the Long-text Language Model Personalization (LongLaMP) Benchmark. LongLaMP provides a comprehensive and diverse evaluation framework for personalized long-text generation. Extensive experiments on LongLaMP for zero-shot and fine-tuned language tasks demonstrate the effectiveness of the proposed benchmark and its utility for developing and evaluating techniques for personalized long-text generation across a wide variety of long-text generation tasks. The results highlight the importance of personalization across a wide variety of long-text generation tasks. Finally, we release the benchmark for others to use for this important problem.
♻ ☆ Ada-K Routing: Boosting the Efficiency of MoE-based LLMs
In the era of Large Language Models (LLMs), Mixture-of-Experts (MoE) architectures offer a promising approach to managing computational costs while scaling up model parameters. Conventional MoE-based LLMs typically employ static Top-K routing, which activates a fixed and equal number of experts for each token regardless of their significance within the context. In this paper, we propose a novel Ada-K routing strategy that dynamically adjusts the number of activated experts for each token, thereby improving the balance between computational efficiency and model performance. Specifically, our strategy incorporates learnable and lightweight allocator modules that decide customized expert resource allocation tailored to the contextual needs for each token. These allocators are designed to be fully pluggable, making it broadly applicable across all mainstream MoE-based LLMs. We leverage the Proximal Policy Optimization (PPO) algorithm to facilitate an end-to-end learning process for this non-differentiable decision-making framework. Extensive evaluations on four popular baseline models demonstrate that our Ada-K routing method significantly outperforms conventional Top-K routing. Compared to Top-K, our method achieves over 25% reduction in FLOPs and more than 20% inference speedup while still improving performance across various benchmarks. Moreover, the training of Ada-K is highly efficient. Even for Mixtral-8x22B, a MoE-based LLM with more than 140B parameters, the training time is limited to 8 hours. Detailed analysis shows that harder tasks, middle layers, and content words tend to activate more experts, providing valuable insights for future adaptive MoE system designs. Both the training code and model checkpoints will be publicly available.
comment: Coauthors do not reach a consensus on submitting the current version
♻ ☆ MisinfoEval: Generative AI in the Era of "Alternative Facts" EMNLP 2024
The spread of misinformation on social media platforms threatens democratic processes, contributes to massive economic losses, and endangers public health. Many efforts to address misinformation focus on a knowledge deficit model and propose interventions for improving users' critical thinking through access to facts. Such efforts are often hampered by challenges with scalability, and by platform users' personal biases. The emergence of generative AI presents promising opportunities for countering misinformation at scale across ideological barriers. In this paper, we introduce a framework (MisinfoEval) for generating and comprehensively evaluating large language model (LLM) based misinformation interventions. We present (1) an experiment with a simulated social media environment to measure effectiveness of misinformation interventions, and (2) a second experiment with personalized explanations tailored to the demographics and beliefs of users with the goal of countering misinformation by appealing to their pre-existing values. Our findings confirm that LLM-based interventions are highly effective at correcting user behavior (improving overall user accuracy at reliability labeling by up to 41.72%). Furthermore, we find that users favor more personalized interventions when making decisions about news reliability and users shown personalized interventions have significantly higher accuracy at identifying misinformation.
comment: EMNLP 2024. Correspondence can be sent to skgabrie at cs dot ucla dot edu
♻ ☆ Measuring Agreeableness Bias in Multimodal Models
This paper examines a phenomenon in multimodal language models where pre-marked options in question images can significantly influence model responses. Our study employs a systematic methodology to investigate this effect: we present models with images of multiple-choice questions, which they initially answer correctly, then expose the same model to versions with pre-marked options. Our findings reveal a significant shift in the models' responses towards the pre-marked option, even when it contradicts their answers in the neutral settings. Comprehensive evaluations demonstrate that this agreeableness bias is a consistent and quantifiable behavior across various model architectures. These results show potential limitations in the reliability of these models when processing images with pre-marked options, raising important questions about their application in critical decision-making contexts where such visual cues might be present.
♻ ☆ EffiLearner: Enhancing Efficiency of Generated Code via Self-Optimization NeurIPS 2024
Large language models (LLMs) have shown remarkable progress in code generation, but their generated code often suffers from inefficiency, resulting in longer execution times and higher memory consumption. To address this issue, we propose \textbf{EffiLearner}, a self-optimization framework that utilizes execution overhead profiles to improve the efficiency of LLM-generated code. EffiLearner first generates code using an LLM, then executes it locally to capture execution time and memory usage profiles. These profiles are fed back to the LLM, which then revises the code to reduce overhead. To evaluate the effectiveness of EffiLearner, we conduct extensive experiments on the EffiBench, HumanEval, and MBPP with 16 open-source and 6 closed-source models. Our evaluation results demonstrate that through iterative self-optimization, EffiLearner significantly enhances the efficiency of LLM-generated code. For example, the execution time (ET) of StarCoder2-15B for the EffiBench decreases from 0.93 (s) to 0.12 (s) which reduces 87.1% the execution time requirement compared with the initial code. The total memory usage (TMU) of StarCoder2-15B also decreases from 22.02 (Mb*s) to 2.03 (Mb*s), which decreases 90.8% of total memory consumption during the execution process. The source code of EffiLearner was released in \url{https://github.com/huangd1999/EffiLearner}.
comment: Accepted by NeurIPS 2024
♻ ☆ FLAG: Financial Long Document Classification via AMR-based GNN
The advent of large language models (LLMs) has initiated much research into their various financial applications. However, in applying LLMs on long documents, semantic relations are not explicitly incorporated, and a full or arbitrarily sparse attention operation is employed. In recent years, progress has been made in Abstract Meaning Representation (AMR), which is a graph-based representation of text to preserve its semantic relations. Since AMR can represent semantic relationships at a deeper level, it can be beneficially utilized by graph neural networks (GNNs) for constructing effective document-level graph representations built upon LLM embeddings to predict target metrics in the financial domain. We propose FLAG: Financial Long document classification via AMR-based GNN, an AMR graph based framework to generate document-level embeddings for long financial document classification. We construct document-level graphs from sentence-level AMR graphs, endow them with specialized LLM word embeddings in the financial domain, apply a deep learning mechanism that utilizes a GNN, and examine the efficacy of our AMR-based approach in predicting labeled target data from long financial documents. Extensive experiments are conducted on a dataset of quarterly earnings calls transcripts of companies in various sectors of the economy, as well as on a corpus of more recent earnings calls of companies in the S&P 1500 Composite Index. We find that our AMR-based approach outperforms fine-tuning LLMs directly on text in predicting stock price movement trends at different time horizons in both datasets. Our work also outperforms previous work utilizing document graphs and GNNs for text classification.
comment: 8 pages, 3 figures, to be published in CIFEr Conference 2024 as "Semantic Graph Learning for Trend Prediction from Long Financial Documents"
♻ ☆ EasyRAG: Efficient Retrieval-Augmented Generation Framework for Automated Network Operations
This paper presents EasyRAG, a simple, lightweight, and efficient retrieval-augmented generation framework for automated network operations. Our framework has three advantages. The first is accurate question answering. We designed a straightforward RAG scheme based on (1) a specific data processing workflow (2) dual-route sparse retrieval for coarse ranking (3) LLM Reranker for reranking (4) LLM answer generation and optimization. This approach achieved first place in the GLM4 track in the preliminary round and second place in the GLM4 track in the semifinals. The second is simple deployment. Our method primarily consists of BM25 retrieval and BGE-reranker reranking, requiring no fine-tuning of any models, occupying minimal VRAM, easy to deploy, and highly scalable; we provide a flexible code library with various search and generation strategies, facilitating custom process implementation. The last one is efficient inference. We designed an efficient inference acceleration scheme for the entire coarse ranking, reranking, and generation process that significantly reduces the inference latency of RAG while maintaining a good level of accuracy; each acceleration scheme can be plug-and-play into any component of the RAG process, consistently enhancing the efficiency of the RAG system. Our code and data are released at \url{https://github.com/BUAADreamer/EasyRAG}.
comment: 10 pages, 2 figures
♻ ☆ ComplexityNet: Increasing LLM Inference Efficiency by Learning Task Complexity
We present ComplexityNet, a streamlined language model designed for assessing task complexity. This model predicts the likelihood of accurate output by various language models, each with different capabilities. Our initial application of ComplexityNet involves the Mostly Basic Python Problems (MBPP) dataset. We pioneered the creation of the first set of labels to define task complexity. ComplexityNet achieved a notable 79% accuracy in determining task complexity, a significant improvement over the 34% accuracy of the original, non fine-tuned model. Furthermore, ComplexityNet effectively reduces computational resource usage by 90% compared to using the highest complexity model, while maintaining a high code generation accuracy of 86.7%. This study demonstrates that fine-tuning smaller models to categorize tasks based on their complexity can lead to a more balanced trade-off between accuracy and efficiency in the use of Large Language Models. Our findings suggest a promising direction for optimizing LLM applications, especially in resource-constrained environments.
♻ ☆ LaCo: Large Language Model Pruning via Layer Collapse EMNLP2024
Large language models (LLMs) based on transformer are witnessing a notable trend of size expansion, which brings considerable costs to both model training and inference. However, existing methods such as model quantization, knowledge distillation, and model pruning are constrained by various issues, including hardware support limitations, the need for extensive training, and alterations to the model internal structure. In this paper, we propose a concise layer-wise structured pruner called \textit{Layer Collapse (LaCo)}, in which rear model layers collapse into a prior layer, enabling a rapid reduction in model size while preserving the model structure. Comprehensive experiments show that our method maintains an average task performance of over 80\% at pruning ratios of 25-30\%, significantly outperforming existing state-of-the-art structured pruning methods. We also conduct post-training experiments to confirm that the \textit{LaCo} effectively inherits the parameters of the original model. Additionally, we perform ablation studies on various settings of \textit{LaCo}. Finally, we discuss our motivation from the perspective of layer-wise similarity and evaluate the performance of the pruned LLMs across various pruning ratios\footnote{\url{https://github.com/yangyifei729/LaCo}}.
comment: Accepted as Findings of EMNLP2024
♻ ☆ Large Visual-Language Models Are Also Good Classifiers: A Study of In-Context Multimodal Fake News Detection
Large visual-language models (LVLMs) exhibit exceptional performance in visual-language reasoning across diverse cross-modal benchmarks. Despite these advances, recent research indicates that Large Language Models (LLMs), like GPT-3.5-turbo, underachieve compared to well-trained smaller models, such as BERT, in Fake News Detection (FND), prompting inquiries into LVLMs' efficacy in FND tasks. Although performance could improve through fine-tuning LVLMs, the substantial parameters and requisite pre-trained weights render it a resource-heavy endeavor for FND applications. This paper initially assesses the FND capabilities of two notable LVLMs, CogVLM and GPT4V, in comparison to a smaller yet adeptly trained CLIP model in a zero-shot context. The findings demonstrate that LVLMs can attain performance competitive with that of the smaller model. Next, we integrate standard in-context learning (ICL) with LVLMs, noting improvements in FND performance, though limited in scope and consistency. To address this, we introduce the \textbf{I}n-context \textbf{M}ultimodal \textbf{F}ake \textbf{N}ews \textbf{D}etection (IMFND) framework, enriching in-context examples and test inputs with predictions and corresponding probabilities from a well-trained smaller model. This strategic integration directs the LVLMs' focus towards news segments associated with higher probabilities, thereby improving their analytical accuracy. The experimental results suggest that the IMFND framework significantly boosts the FND efficiency of LVLMs, achieving enhanced accuracy over the standard ICL approach across three publicly available FND datasets.
♻ ☆ Plurals: A System for Guiding LLMs Via Simulated Social Ensembles
Recent debates raised concerns that language models may favor certain viewpoints. But what if the solution is not to aim for a 'view from nowhere' but rather to leverage different viewpoints? We introduce Plurals, a system and Python library for pluralistic AI deliberation. Plurals consists of Agents (LLMs, optionally with personas) which deliberate within customizable Structures, with Moderators overseeing deliberation. Plurals is a generator of simulated social ensembles. Plurals integrates with government datasets to create nationally representative personas, includes deliberation templates inspired by democratic deliberation theory, and allows users to customize both information-sharing structures and deliberation behavior within Structures. Six case studies demonstrate fidelity to theoretical constructs and efficacy. Three randomized experiments show simulated focus groups produced output resonant with an online sample of the relevant audiences (chosen over zero-shot generation in 75% of trials). Plurals is both a paradigm and a concrete system for pluralistic AI. The Plurals library is available at https://github.com/josh-ashkinaze/plurals and will be continually updated.
♻ ☆ Nudging: Inference-time Alignment via Model Collaboration
Large language models (LLMs) require alignment, such as instruction-tuning or reinforcement learning from human feedback, to effectively and safely follow user instructions. This process necessitates training aligned versions for every model size in each model family, resulting in significant computational overhead. In this work, we propose nudging, a simple, plug-and-play, and training-free algorithm that aligns any base model at inference time using a small aligned model. Nudging is motivated by recent findings that alignment primarily alters the model's behavior on a small subset of stylistic tokens, such as "Sure" or "Thank". We find that base models are significantly more uncertain when generating these tokens. Leveraging this observation, nudging employs a small aligned model to generate nudging tokens to steer the large base model's output toward desired directions when the base model's uncertainty is high. We evaluate the effectiveness of nudging across 3 model families and 13 tasks, covering reasoning, general knowledge, instruction following, and safety benchmarks. Without any additional training, nudging a large base model with a 7x - 14x smaller aligned model achieves zero-shot performance comparable to, and sometimes surpassing, that of large aligned models. For example, nudging OLMo-7b with OLMo-1b-instruct, affecting less than 9% of tokens, achieves a 10% absolute improvement on GSM8K over OLMo-7b-instruct. Unlike prior inference-time tuning methods, nudging enables off-the-shelf collaboration between model families. For instance, nudging Gemma-2-27b with Llama-2-7b-chat outperforms Llama-2-70b-chat on various tasks. Overall, this work introduces a simple yet powerful approach to token-level model collaboration, offering a modular solution to LLM alignment. Our project website: https://fywalter.github.io/nudging/ .
♻ ☆ Learning Evolving Tools for Large Language Models
Tool learning enables large language models (LLMs) to interact with external tools and APIs, greatly expanding the application scope of LLMs. However, due to the dynamic nature of external environments, these tools and APIs may become outdated over time, preventing LLMs from correctly invoking tools. Existing research primarily focuses on static environments and overlooks this issue, limiting the adaptability of LLMs in real-world applications. In this paper, we propose ToolEVO, a novel framework designed to enhance the adaptive and reflective capabilities of LLMs against tool variability. By leveraging Monte Carlo Tree Search, ToolEVO facilitates active exploration and interaction of LLMs within dynamic environments, allowing for autonomous self-reflection and self-updating of tool usage based on environmental feedback. Additionally, we introduce ToolQA-D, a benchmark specifically designed to evaluate the impact of tool variability. Extensive experiments demonstrate the effectiveness and stability of our approach, highlighting the importance of adaptability to tool variability for effective tool learning.
comment: Ongoing Work
♻ ☆ Mitigating the Influence of Distractor Tasks in LMs with Prior-Aware Decoding
The broad capabilities of Language Models (LMs) can be limited by their sensitivity to distractor tasks: LMs can infer secondary tasks from the prompt in addition to the intended one, leading to unwanted outputs. For example, prompt injection attacks can cause models to deviate from explicit directives. In some 'inverse scaling' cases, this unwanted behaviour actually worsens as models scale up to at least 540B parameters. We present a theoretical framework that interprets LMs as a product of experts that combine multiple data generation processes. Based on this framework, we demonstrate prior-aware decoding (PAD) - a simple contrastive inference method to reduce the influence of distractor tasks. We apply PAD to eleven models, across four datasets, and find improvements in 41 out of 44 task-model combinations, with a median increase in task completion proportion of 40%. The results suggest a promising direction for further development towards more reliable language models.
comment: 16 pages, 6 figures
♻ ☆ Auto-ABSA: Cross-Domain Aspect Detection and Sentiment Analysis Using Auxiliary Sentences
After transformer is proposed, lots of pre-trained language models have been come up with and sentiment analysis (SA) task has been improved. In this paper, we proposed a method that uses an auxiliary sentence about aspects that the sentence contains to help sentiment prediction. The first is aspect detection, which uses a multi-aspects detection model to predict all aspects that the sentence has. Combining the predicted aspects and the original sentence as Sentiment Analysis (SA) model's input. The second is to do out-of-domain aspect-based sentiment analysis(ABSA), train sentiment classification model with one kind of dataset and validate it with another kind of dataset. Finally, we created two baselines, they use no aspect and all aspects as sentiment classification model's input, respectively. Compare two baselines performance to our method, found that our method really makes sense.
♻ ☆ Enhancing and Assessing Instruction-Following with Fine-Grained Instruction Variants
The effective alignment of Large Language Models (LLMs) with precise instructions is essential for their application in diverse real-world scenarios. Current methods focus on enhancing the diversity and complexity of training and evaluation samples, yet they fall short in accurately assessing LLMs' ability to follow similar instruction variants. We introduce an effective data augmentation technique DeMoRecon that decomposes complex instructions into simpler sub-components, modifies these, and reconstructs them into new variants, thereby preserves the original instruction's context and complexity while introducing variability, which is critical for training and evaluating LLMs' instruction-following precision. Based on DeMoRecon, we developed the FGIV dataset which contains fine-grained instruction variants of 1,773 seed instructions to both fine-tune and evaluate LLMs. Our findings show that LLMs fine-tuned with FGIV will gain significant performance boost on both ours and commonly used instructions-following benchmarks.
♻ ☆ Large Language Models Can Solve Real-World Planning Rigorously with Formal Verification Tools
Despite their recent advancements, Large Language Models (LLMs) still struggle to directly generate correct plans for complex multi-constraint planning problems, even with self-verification and self-critique. For example, a U.S. domestic travel planning benchmark TravelPlanner was proposed in Xie et al. (2024), where the best LLM OpenAI o1-preview can only find travel plans that satisfy user requirements with a 10% success rate given all needed information. In this work, we tackle this difficult problem by proposing an LLM-based planning framework that formalizes and solves complex multi-constraint planning problems as constrained satisfiability problems, which are further consumed by sound and complete satisfiability solvers. We start with TravelPlanner as the primary use case and achieve a success rate of 93.9%. We demonstrate our framework's robustness by showing its effectiveness in diverse paraphrased prompts. More importantly, our framework has strong zero-shot generalizability: It can successfully handle unseen constraints in a completely unseen international travel dataset we created, and it can even generalize well to new domains such as routing and task allocation problems in a zero-shot manner. Moreover, when user input queries are infeasible, our framework can identify the unsatisfiable core, provide failure reasons, and offers personalized modification suggestions to users according to diverse human preferences. We show that our framework can modify and solve for an average of 81.6% and 91.7% unsatisfiable queries from two datasets and prove with ablations that all key components of our framework are effective and necessary.
comment: 48 pages, 6 figures, 8 tables
♻ ☆ GPT-4 Jailbreaks Itself with Near-Perfect Success Using Self-Explanation EMNLP 2024
Research on jailbreaking has been valuable for testing and understanding the safety and security issues of large language models (LLMs). In this paper, we introduce Iterative Refinement Induced Self-Jailbreak (IRIS), a novel approach that leverages the reflective capabilities of LLMs for jailbreaking with only black-box access. Unlike previous methods, IRIS simplifies the jailbreaking process by using a single model as both the attacker and target. This method first iteratively refines adversarial prompts through self-explanation, which is crucial for ensuring that even well-aligned LLMs obey adversarial instructions. IRIS then rates and enhances the output given the refined prompt to increase its harmfulness. We find that IRIS achieves jailbreak success rates of 98% on GPT-4, 92% on GPT-4 Turbo, and 94% on Llama-3.1-70B in under 7 queries. It significantly outperforms prior approaches in automatic, black-box, and interpretable jailbreaking, while requiring substantially fewer queries, thereby establishing a new standard for interpretable jailbreaking methods.
comment: Accepted to EMNLP 2024 Main Conference
♻ ☆ Encourage or Inhibit Monosemanticity? Revisit Monosemanticity from a Feature Decorrelation Perspective EMNLP24
To better interpret the intrinsic mechanism of large language models (LLMs), recent studies focus on monosemanticity on its basic units. A monosemantic neuron is dedicated to a single and specific concept, which forms a one-to-one correlation between neurons and concepts. Despite extensive research in monosemanticity probing, it remains unclear whether monosemanticity is beneficial or harmful to model capacity. To explore this question, we revisit monosemanticity from the feature decorrelation perspective and advocate for its encouragement. We experimentally observe that the current conclusion by wang2024learning, which suggests that decreasing monosemanticity enhances model performance, does not hold when the model changes. Instead, we demonstrate that monosemanticity consistently exhibits a positive correlation with model capacity, in the preference alignment process. Consequently, we apply feature correlation as a proxy for monosemanticity and incorporate a feature decorrelation regularizer into the dynamic preference optimization process. The experiments show that our method not only enhances representation diversity and activation sparsity but also improves preference alignment performance.
comment: EMNLP24, Main, Long
♻ ☆ PromptIntern: Saving Inference Costs by Internalizing Recurrent Prompt during Large Language Model Fine-tuning EMNLP 2024
Recent advances in fine-tuning large language models (LLMs) have greatly enhanced their usage in domain-specific tasks. Despite the success, fine-tuning continues to rely on repeated and lengthy prompts, which escalate computational expenses, require more resources, and lead to slower inference. In this paper, we present a novel approach, PromptIntern, which internalizes prompt knowledge during model fine-tuning to achieve efficient inference and save costs. Instead of compressing the prompts for a vanilla model, PromptIntern aims to embed the recurrent prompt directly into the model parameters. We design a fine-tuning pipeline that includes instruction template compression, few-shot example absorption, and a progressive internalization strategy, effectively diminishing the need for intricate prompts during inference. Comprehensive experiments on challenging NL2Code tasks demonstrate that our method reduces input tokens by more than 90%, accelerates inference by 4.2 times, and reduces monetary inference costs by 88.3%.
comment: This paper has been accepted by EMNLP 2024
♻ ☆ COMPASS: Computational Mapping of Patient-Therapist Alliance Strategies with Language Modeling
The therapeutic working alliance is a critical factor in predicting the success of psychotherapy treatment. Traditionally, working alliance assessment relies on questionnaires completed by both therapists and patients. In this paper, we present COMPASS, a novel framework to directly infer the therapeutic working alliance from the natural language used in psychotherapy sessions. Our approach utilizes advanced large language models (LLMs) to analyze transcripts of psychotherapy sessions and compare them with distributed representations of statements in the working alliance inventory. Analyzing a dataset of over 950 sessions covering diverse psychiatric conditions including anxiety, depression, schizophrenia, and suicidal tendencies, we demonstrate the effectiveness of our method in providing fine-grained mapping of patient-therapist alignment trajectories and offering interpretability for clinical psychiatry and in identifying emerging patterns related to the condition being treated. By employing various deep learning-based topic modeling techniques in combination with prompting generative language models, we analyze the topical characteristics of different psychiatric conditions and their evolution at a turn-level resolution. This combined framework enhances the understanding of therapeutic interactions, enabling timely feedback for therapists regarding the quality of therapeutic relationships and providing interpretable insights to improve the effectiveness of psychotherapy.
comment: This work extends our research series in computational psychiatry (e.g auto annotation in arXiv:2204.05522, topic extraction in arXiv:2204.10189, and diagnosis in arXiv:2210.15603) with the introduction of LLMs to complete the full cycle of interpreting and understanding psychotherapy strategies as a comprehensive analytical framework
♻ ☆ SciDoc2Diagrammer-MAF: Towards Generation of Scientific Diagrams from Documents guided by Multi-Aspect Feedback Refinement
Automating the creation of scientific diagrams from academic papers can significantly streamline the development of tutorials, presentations, and posters, thereby saving time and accelerating the process. Current text-to-image models struggle with generating accurate and visually appealing diagrams from long-context inputs. We propose SciDoc2Diagram, a task that extracts relevant information from scientific papers and generates diagrams, along with a benchmarking dataset, SciDoc2DiagramBench. We develop a multi-step pipeline SciDoc2Diagrammer that generates diagrams based on user intentions using intermediate code generation. We observed that initial diagram drafts were often incomplete or unfaithful to the source, leading us to develop SciDoc2Diagrammer-Multi-Aspect-Feedback (MAF), a refinement strategy that significantly enhances factual correctness and visual appeal and outperforms existing models on both automatic and human judgement.
comment: Code and data available at https://github.com/Ishani-Mondal/SciDoc2DiagramGeneration
♻ ☆ Core: Robust Factual Precision with Informative Sub-Claim Identification
Hallucinations pose a challenge to the application of large language models (LLMs) thereby motivating the development of metrics to evaluate factual precision. We observe that popular metrics using the Decompose-Then-Verify framework, such as \FActScore, can be manipulated by adding obvious or repetitive subclaims to artificially inflate scores. This observation motivates our new customizable plug-and-play subclaim selection component called Core, which filters down individual subclaims according to their uniqueness and informativeness. We show that many popular factual precision metrics augmented by Core are substantially more robust on a wide range of knowledge domains. We release an evaluation framework supporting easy and modular use of Core and various decomposition strategies, which we recommend adoption by the community. We also release an expansion of the FActScore biography dataset to facilitate further studies of decomposition-based factual precision evaluation.
♻ ☆ Few shot chain-of-thought driven reasoning to prompt LLMs for open ended medical question answering EMNLP 2024
In this paper, we propose a modified version of the MedQA-USMLE dataset, named MEDQA-OPEN, which contains open-ended medical questions without options to mimic clinical scenarios, along with clinician-approved reasoned answers. Additionally, we implement a prompt driven by Chain of Thought (CoT) reasoning, CLINICR, to mirror the prospective process of incremental reasoning, reaching a correct response to medical questions. We empirically demonstrate how CLINICR outperforms the state-of-the-art 5-shot CoT-based prompt (Li\'evin et al., 2022). We also present an approach that mirrors real-life clinical practice by first exploring multiple differential diagnoses through MCQ-CLINICR and subsequently narrowing down to a final diagnosis using MCQ-ELIMINATIVE. Finally, emphasizing the importance of response verification in medical settings, we utilize a reward model mechanism, replacing the elimination process performed by MCQ-ELIMINATIVE.
comment: The paper is accepted in EMNLP 2024 Findings
♻ ☆ BlendFilter: Advancing Retrieval-Augmented Large Language Models via Query Generation Blending and Knowledge Filtering EMNLP 2024
Retrieval-augmented Large Language Models (LLMs) offer substantial benefits in enhancing performance across knowledge-intensive scenarios. However, these methods often face challenges with complex inputs and encounter difficulties due to noisy knowledge retrieval, notably hindering model effectiveness. To address this issue, we introduce BlendFilter, a novel approach that elevates retrieval-augmented LLMs by integrating query generation blending with knowledge filtering. BlendFilter proposes the blending process through its query generation method, which integrates both external and internal knowledge augmentation with the original query, ensuring comprehensive information gathering. Additionally, our distinctive knowledge filtering module capitalizes on the intrinsic capabilities of the LLM, effectively eliminating extraneous data. We conduct extensive experiments on three open-domain question answering benchmarks, and the findings clearly indicate that our innovative BlendFilter surpasses state-of-the-art baselines significantly.
comment: EMNLP 2024 main
♻ ☆ Modeling Bilingual Sentence Processing: Evaluating RNN and Transformer Architectures for Cross-Language Structural Priming EMNLP
This study evaluates the performance of Recurrent Neural Network (RNN) and Transformer models in replicating cross-language structural priming, a key indicator of abstract grammatical representations in human language processing. Focusing on Chinese-English priming, which involves two typologically distinct languages, we examine how these models handle the robust phenomenon of structural priming, where exposure to a particular sentence structure increases the likelihood of selecting a similar structure subsequently. Our findings indicate that transformers outperform RNNs in generating primed sentence structures, with accuracy rates that exceed 25.84\% to 33. 33\%. This challenges the conventional belief that human sentence processing primarily involves recurrent and immediate processing and suggests a role for cue-based retrieval mechanisms. This work contributes to our understanding of how computational models may reflect human cognitive processes across diverse language families.
comment: This study evaluates the performance of RNN and Transformer models in replicating Chinese-English structural priming. Accepted by EMNLP Multilingual Representation Learning (MRL) Workshop 2024
♻ ☆ HumVI: A Multilingual Dataset for Detecting Violent Incidents Impacting Humanitarian Aid
Humanitarian organizations can enhance their effectiveness by analyzing data to discover trends, gather aggregated insights, manage their security risks, support decision-making, and inform advocacy and funding proposals. However, data about violent incidents with direct impact and relevance for humanitarian aid operations is not readily available. An automatic data collection and NLP-backed classification framework aligned with humanitarian perspectives can help bridge this gap. In this paper, we present HumVI - a dataset comprising news articles in three languages (English, French, Arabic) containing instances of different types of violent incidents categorized by the humanitarian sector they impact, e.g., aid security, education, food security, health, and protection. Reliable labels were obtained for the dataset by partnering with a data-backed humanitarian organization, Insecurity Insight. We provide multiple benchmarks for the dataset, employing various deep learning architectures and techniques, including data augmentation and mask loss, to address different task-related challenges, e.g., domain expansion. The dataset is publicly available at https://github.com/dataminr-ai/humvi-dataset.
♻ ☆ A Watermark for Low-entropy and Unbiased Generation in Large Language Models
Recent advancements in large language models (LLMs) have highlighted the risk of misusing them, raising the need for accurate detection of LLM-generated content. In response, a viable solution is to inject imperceptible identifiers into LLMs, known as watermarks. Previous work demonstrates that unbiased watermarks ensure unforgeability and preserve text quality by maintaining the expectation of the LLM output probability distribution. However, previous unbiased watermarking methods suffer from one or more of the following issues: (1) requiring access to white-box LLMs during detection, (2) incurring long detection time, (3) being not robust against simple watermarking attacks, (4) failing to provide statistical guarantees for the type II error of watermark detection, and (5) being not statistically unbiased for low-entropy scenarios, which hinder their deployment in practice. This study proposes the Sampling One Then Accepting (STA-1) method, a watermark that can address all of these issues. Moreover, we discuss the tradeoff between watermark strength and text quality for unbiased watermarks. We show that in low-entropy scenarios, unbiased watermarks face a tradeoff between watermark strength and the risk of unsatisfactory outputs. Experimental results on both low-entropy and high-entropy datasets demonstrate that STA-1 achieves text quality and watermark strength comparable to existing unbiased watermarks, with a low risk of unsatisfactory outputs. Implementation codes for this study are available online.
♻ ☆ Towards Rationality in Language and Multimodal Agents: A Survey
Rationality is the quality of being guided by reason, characterized by decision-making that aligns with evidence and logical principles. It plays a crucial role in reliable problem-solving by ensuring well-grounded and consistent solutions. While large language models (LLMs) have made significant progress in generating human-like text, they still exhibit limitations such as bounded knowledge space and inconsistent outputs. In response, recent efforts have shifted toward developing multimodal and multi-agent systems, as well as integrating modules like external tools, programming codes, symbolic reasoners, utility function, and conformal risk controls rather than relying solely on a single LLM for decision-making. This paper surveys the state-of-the-art advancements in language and multimodal agents, evaluates how they contribute to make intelligent agents more rational, and identifies open challenges and future research directions. We maintain an open repository at https://github.com/bowen-upenn/Agent_Rationality.
comment: We maintain an open repository at https://github.com/bowen-upenn/Agent_Rationality
♻ ☆ Reframing Data Value for Large Language Models Through the Lens of Plausibility
Data valuation seeks to answer the important question, "How much is this data worth?" Existing data valuation methods have largely focused on discriminative models, primarily examining data value through the lens of its utility in training. However, with the push for ever-larger language models, relying on valuation methods that require training becomes increasingly expensive and dependent on specific techniques. We propose an alternative perspective on the data value problem for language models, centering around the plausibility of the data. We posit that data holds lesser value if it can be plausibly generated by the model itself. Starting from some intuitive criteria that align with our notions of valuable data, we develop a novel value function that is computationally tractable and derived from first principles with provable properties. We conduct a theoretical analysis of our value function and evaluate it across multiple scenarios and datasets.
♻ ☆ Empowering Air Travelers: A Chatbot for Canadian Air Passenger Rights EMNLP 2024
The Canadian air travel sector has seen a significant increase in flight delays, cancellations, and other issues concerning passenger rights. Recognizing this demand, we present a chatbot to assist passengers and educate them about their rights. Our system breaks a complex user input into simple queries which are used to retrieve information from a collection of documents detailing air travel regulations. The most relevant passages from these documents are presented along with links to the original documents and the generated queries, enabling users to dissect and leverage the information for their unique circumstances. The system successfully overcomes two predominant challenges: understanding complex user inputs, and delivering accurate answers, free of hallucinations, that passengers can rely on for making informed decisions. A user study comparing the chatbot to a Google search demonstrated the chatbot's usefulness and ease of use. Beyond the primary goal of providing accurate and timely information to air passengers regarding their rights, we hope that this system will also enable further research exploring the tradeoff between the user-friendly conversational interface of chatbots and the accuracy of retrieval systems.
comment: Accepted to The Natural Legal Language Processing Workshop at EMNLP 2024
♻ ☆ RoseLoRA: Row and Column-wise Sparse Low-rank Adaptation of Pre-trained Language Model for Knowledge Editing and Fine-tuning EMNLP 2024
Pre-trained language models, trained on large-scale corpora, demonstrate strong generalizability across various NLP tasks. Fine-tuning these models for specific tasks typically involves updating all parameters, which is resource-intensive. Parameter-efficient fine-tuning (PEFT) methods, such as the popular LoRA family, introduce low-rank matrices to learn only a few parameters efficiently. However, during inference, the product of these matrices updates all pre-trained parameters, complicating tasks like knowledge editing that require selective updates. We propose a novel PEFT method, which conducts \textbf{r}ow and c\textbf{o}lumn-wise spar\textbf{se} \textbf{lo}w-\textbf{r}ank \textbf{a}daptation (RoseLoRA), to address this challenge. RoseLoRA identifies and updates only the most important parameters for a specific task, maintaining efficiency while preserving other model knowledge. By adding a sparsity constraint on the product of low-rank matrices and converting it to row and column-wise sparsity, we ensure efficient and precise model updates. Our theoretical analysis guarantees the lower bound of the sparsity with respective to the matrix product. Extensive experiments on five benchmarks across twenty datasets demonstrate that RoseLoRA outperforms baselines in both general fine-tuning and knowledge editing tasks.
comment: EMNLP 2024 main
Computer Vision and Pattern Recognition 199
☆ MoH: Multi-Head Attention as Mixture-of-Head Attention
In this work, we upgrade the multi-head attention mechanism, the core of the Transformer model, to improve efficiency while maintaining or surpassing the previous accuracy level. We show that multi-head attention can be expressed in the summation form. Drawing on the insight that not all attention heads hold equal significance, we propose Mixture-of-Head attention (MoH), a new architecture that treats attention heads as experts in the Mixture-of-Experts (MoE) mechanism. MoH has two significant advantages: First, MoH enables each token to select the appropriate attention heads, enhancing inference efficiency without compromising accuracy or increasing the number of parameters. Second, MoH replaces the standard summation in multi-head attention with a weighted summation, introducing flexibility to the attention mechanism and unlocking extra performance potential. Extensive experiments on ViT, DiT, and LLMs demonstrate that MoH outperforms multi-head attention by using only 50%-90% of the attention heads. Moreover, we demonstrate that pre-trained multi-head attention models, such as LLaMA3-8B, can be further continue-tuned into our MoH models. Notably, MoH-LLaMA3-8B achieves an average accuracy of 64.0% across 14 benchmarks, outperforming LLaMA3-8B by 2.4% by utilizing only 75% of the attention heads. We believe the proposed MoH is a promising alternative to multi-head attention and provides a strong foundation for developing advanced and efficient attention-based models.
comment: 23 pages, code: https://github.com/SkyworkAI/MoH
☆ High-Resolution Frame Interpolation with Patch-based Cascaded Diffusion
Despite the recent progress, existing frame interpolation methods still struggle with processing extremely high resolution input and handling challenging cases such as repetitive textures, thin objects, and large motion. To address these issues, we introduce a patch-based cascaded pixel diffusion model for frame interpolation, HiFI, that excels in these scenarios while achieving competitive performance on standard benchmarks. Cascades, which generate a series of images from low- to high-resolution, can help significantly with large or complex motion that require both global context for a coarse solution and detailed context for high resolution output. However, contrary to prior work on cascaded diffusion models which perform diffusion on increasingly large resolutions, we use a single model that always performs diffusion at the same resolution and upsamples by processing patches of the inputs and the prior solution. We show that this technique drastically reduces memory usage at inference time and also allows us to use a single model at test time, solving both frame interpolation and spatial up-sampling, saving training cost. We show that HiFI helps significantly with high resolution and complex repeated textures that require global context. HiFI demonstrates comparable or beyond state-of-the-art performance on multiple benchmarks (Vimeo, Xiph, X-Test, SEPE-8K). On our newly introduced dataset that focuses on particularly challenging cases, HiFI also significantly outperforms other baselines on these cases. Please visit our project page for video results: https://hifi-diffusion.github.io
comment: Project page: https://hifi-diffusion.github.io/
☆ On the Effectiveness of Dataset Alignment for Fake Image Detection
As latent diffusion models (LDMs) democratize image generation capabilities, there is a growing need to detect fake images. A good detector should focus on the generative models fingerprints while ignoring image properties such as semantic content, resolution, file format, etc. Fake image detectors are usually built in a data driven way, where a model is trained to separate real from fake images. Existing works primarily investigate network architecture choices and training recipes. In this work, we argue that in addition to these algorithmic choices, we also require a well aligned dataset of real/fake images to train a robust detector. For the family of LDMs, we propose a very simple way to achieve this: we reconstruct all the real images using the LDMs autoencoder, without any denoising operation. We then train a model to separate these real images from their reconstructions. The fakes created this way are extremely similar to the real ones in almost every aspect (e.g., size, aspect ratio, semantic content), which forces the model to look for the LDM decoders artifacts. We empirically show that this way of creating aligned real/fake datasets, which also sidesteps the computationally expensive denoising process, helps in building a detector that focuses less on spurious correlations, something that a very popular existing method is susceptible to. Finally, to demonstrate just how effective the alignment in a dataset can be, we build a detector using images that are not natural objects, and present promising results. Overall, our work identifies the subtle but significant issues that arise when training a fake image detector and proposes a simple and inexpensive solution to address these problems.
☆ CoTracker3: Simpler and Better Point Tracking by Pseudo-Labelling Real Videos
Most state-of-the-art point trackers are trained on synthetic data due to the difficulty of annotating real videos for this task. However, this can result in suboptimal performance due to the statistical gap between synthetic and real videos. In order to understand these issues better, we introduce CoTracker3, comprising a new tracking model and a new semi-supervised training recipe. This allows real videos without annotations to be used during training by generating pseudo-labels using off-the-shelf teachers. The new model eliminates or simplifies components from previous trackers, resulting in a simpler and often smaller architecture. This training scheme is much simpler than prior work and achieves better results using 1,000 times less data. We further study the scaling behaviour to understand the impact of using more real unsupervised data in point tracking. The model is available in online and offline variants and reliably tracks visible and occluded points.
☆ MMFuser: Multimodal Multi-Layer Feature Fuser for Fine-Grained Vision-Language Understanding
Despite significant advancements in Multimodal Large Language Models (MLLMs) for understanding complex human intentions through cross-modal interactions, capturing intricate image details remains challenging. Previous methods integrating multiple vision encoders to enhance visual detail introduce redundancy and computational overhead. We observe that most MLLMs utilize only the last-layer feature map of the vision encoder for visual representation, neglecting the rich fine-grained information in shallow feature maps. To address this issue, we propose \modelname, a simple yet effective multi-layer feature fuser that efficiently integrates deep and shallow features from Vision Transformers (ViTs). Specifically, it leverages semantically aligned deep features as queries to dynamically extract missing details from shallow features, thus preserving semantic alignment while enriching the representation with fine-grained information. Applied to the LLaVA-1.5 model, \modelname~achieves significant improvements in visual representation and benchmark performance, providing a more flexible and lightweight solution compared to multi-encoder ensemble methods. The code and model have been released at https://github.com/yuecao0119/MMFuser.
comment: 11 pages, 6 figures, technical report
☆ Analysis and Benchmarking of Extending Blind Face Image Restoration to Videos
Recent progress in blind face restoration has resulted in producing high-quality restored results for static images. However, efforts to extend these advancements to video scenarios have been minimal, partly because of the absence of benchmarks that allow for a comprehensive and fair comparison. In this work, we first present a fair evaluation benchmark, in which we first introduce a Real-world Low-Quality Face Video benchmark (RFV-LQ), evaluate several leading image-based face restoration algorithms, and conduct a thorough systematical analysis of the benefits and challenges associated with extending blind face image restoration algorithms to degraded face videos. Our analysis identifies several key issues, primarily categorized into two aspects: significant jitters in facial components and noise-shape flickering between frames. To address these issues, we propose a Temporal Consistency Network (TCN) cooperated with alignment smoothing to reduce jitters and flickers in restored videos. TCN is a flexible component that can be seamlessly plugged into the most advanced face image restoration algorithms, ensuring the quality of image-based restoration is maintained as closely as possible. Extensive experiments have been conducted to evaluate the effectiveness and efficiency of our proposed TCN and alignment smoothing operation. Project page: https://wzhouxiff.github.io/projects/FIR2FVR/FIR2FVR.
comment: Accepted by TIP'2024; Project page: https://wzhouxiff.github.io/projects/FIR2FVR/FIR2FVR
☆ KITTEN: A Knowledge-Intensive Evaluation of Image Generation on Visual Entities
Recent advancements in text-to-image generation have significantly enhanced the quality of synthesized images. Despite this progress, evaluations predominantly focus on aesthetic appeal or alignment with text prompts. Consequently, there is limited understanding of whether these models can accurately represent a wide variety of realistic visual entities - a task requiring real-world knowledge. To address this gap, we propose a benchmark focused on evaluating Knowledge-InTensive image generaTion on real-world ENtities (i.e., KITTEN). Using KITTEN, we conduct a systematic study on the fidelity of entities in text-to-image generation models, focusing on their ability to generate a wide range of real-world visual entities, such as landmark buildings, aircraft, plants, and animals. We evaluate the latest text-to-image models and retrieval-augmented customization models using both automatic metrics and carefully-designed human evaluations, with an emphasis on the fidelity of entities in the generated images. Our findings reveal that even the most advanced text-to-image models often fail to generate entities with accurate visual details. Although retrieval-augmented models can enhance the fidelity of entity by incorporating reference images during testing, they often over-rely on these references and struggle to produce novel configurations of the entity as requested in creative text prompts.
comment: Project page: https://kitten-project.github.io/
☆ Improving Long-Text Alignment for Text-to-Image Diffusion Models
The rapid advancement of text-to-image (T2I) diffusion models has enabled them to generate unprecedented results from given texts. However, as text inputs become longer, existing encoding methods like CLIP face limitations, and aligning the generated images with long texts becomes challenging. To tackle these issues, we propose LongAlign, which includes a segment-level encoding method for processing long texts and a decomposed preference optimization method for effective alignment training. For segment-level encoding, long texts are divided into multiple segments and processed separately. This method overcomes the maximum input length limits of pretrained encoding models. For preference optimization, we provide decomposed CLIP-based preference models to fine-tune diffusion models. Specifically, to utilize CLIP-based preference models for T2I alignment, we delve into their scoring mechanisms and find that the preference scores can be decomposed into two components: a text-relevant part that measures T2I alignment and a text-irrelevant part that assesses other visual aspects of human preference. Additionally, we find that the text-irrelevant part contributes to a common overfitting problem during fine-tuning. To address this, we propose a reweighting strategy that assigns different weights to these two components, thereby reducing overfitting and enhancing alignment. After fine-tuning $512 \times 512$ Stable Diffusion (SD) v1.5 for about 20 hours using our method, the fine-tuned SD outperforms stronger foundation models in T2I alignment, such as PixArt-$\alpha$ and Kandinsky v2.2. The code is available at https://github.com/luping-liu/LongAlign.
☆ Jigsaw++: Imagining Complete Shape Priors for Object Reassembly
The automatic assembly problem has attracted increasing interest due to its complex challenges that involve 3D representation. This paper introduces Jigsaw++, a novel generative method designed to tackle the multifaceted challenges of reconstruction for the reassembly problem. Existing approach focusing primarily on piecewise information for both part and fracture assembly, often overlooking the integration of complete object prior. Jigsaw++ distinguishes itself by learning a category-agnostic shape prior of complete objects. It employs the proposed "retargeting" strategy that effectively leverages the output of any existing assembly method to generate complete shape reconstructions. This capability allows it to function orthogonally to the current methods. Through extensive evaluations on Breaking Bad dataset and PartNet, Jigsaw++ has demonstrated its effectiveness, reducing reconstruction errors and enhancing the precision of shape reconstruction, which sets a new direction for future reassembly model developments.
comment: 21 pages, 10 figures
☆ SGEdit: Bridging LLM with Text2Image Generative Model for Scene Graph-based Image Editing SIGGRAPH
Scene graphs offer a structured, hierarchical representation of images, with nodes and edges symbolizing objects and the relationships among them. It can serve as a natural interface for image editing, dramatically improving precision and flexibility. Leveraging this benefit, we introduce a new framework that integrates large language model (LLM) with Text2Image generative model for scene graph-based image editing. This integration enables precise modifications at the object level and creative recomposition of scenes without compromising overall image integrity. Our approach involves two primary stages: 1) Utilizing a LLM-driven scene parser, we construct an image's scene graph, capturing key objects and their interrelationships, as well as parsing fine-grained attributes such as object masks and descriptions. These annotations facilitate concept learning with a fine-tuned diffusion model, representing each object with an optimized token and detailed description prompt. 2) During the image editing phase, a LLM editing controller guides the edits towards specific areas. These edits are then implemented by an attention-modulated diffusion editor, utilizing the fine-tuned model to perform object additions, deletions, replacements, and adjustments. Through extensive experiments, we demonstrate that our framework significantly outperforms existing image editing methods in terms of editing precision and scene aesthetics.
comment: Accepted by ACM Transactions on Graphics and SIGGRAPH Asia 2024. Project page: https://bestzzhang.github.io/SGEdit
☆ Efficient Diffusion Models: A Comprehensive Survey from Principles to Practices
As one of the most popular and sought-after generative models in the recent years, diffusion models have sparked the interests of many researchers and steadily shown excellent advantage in various generative tasks such as image synthesis, video generation, molecule design, 3D scene rendering and multimodal generation, relying on their dense theoretical principles and reliable application practices. The remarkable success of these recent efforts on diffusion models comes largely from progressive design principles and efficient architecture, training, inference, and deployment methodologies. However, there has not been a comprehensive and in-depth review to summarize these principles and practices to help the rapid understanding and application of diffusion models. In this survey, we provide a new efficiency-oriented perspective on these existing efforts, which mainly focuses on the profound principles and efficient practices in architecture designs, model training, fast inference and reliable deployment, to guide further theoretical research, algorithm migration and model application for new scenarios in a reader-friendly way. \url{https://github.com/ponyzym/Efficient-DMs-Survey}
☆ OKAMI: Teaching Humanoid Robots Manipulation Skills through Single Video Imitation
We study the problem of teaching humanoid robots manipulation skills by imitating from single video demonstrations. We introduce OKAMI, a method that generates a manipulation plan from a single RGB-D video and derives a policy for execution. At the heart of our approach is object-aware retargeting, which enables the humanoid robot to mimic the human motions in an RGB-D video while adjusting to different object locations during deployment. OKAMI uses open-world vision models to identify task-relevant objects and retarget the body motions and hand poses separately. Our experiments show that OKAMI achieves strong generalizations across varying visual and spatial conditions, outperforming the state-of-the-art baseline on open-world imitation from observation. Furthermore, OKAMI rollout trajectories are leveraged to train closed-loop visuomotor policies, which achieve an average success rate of 79.2% without the need for labor-intensive teleoperation. More videos can be found on our website https://ut-austin-rpl.github.io/OKAMI/.
comment: Accepted for oral presentation at 8th Annual Conference on Robot Learning. Project website: https://ut-austin-rpl.github.io/OKAMI/
☆ Latent BKI: Open-Dictionary Continuous Mapping in Visual-Language Latent Spaces with Quantifiable Uncertainty
This paper introduces a novel probabilistic mapping algorithm, Latent BKI, which enables open-vocabulary mapping with quantifiable uncertainty. Traditionally, semantic mapping algorithms focus on a fixed set of semantic categories which limits their applicability for complex robotic tasks. Vision-Language (VL) models have recently emerged as a technique to jointly model language and visual features in a latent space, enabling semantic recognition beyond a predefined, fixed set of semantic classes. Latent BKI recurrently incorporates neural embeddings from VL models into a voxel map with quantifiable uncertainty, leveraging the spatial correlations of nearby observations through Bayesian Kernel Inference (BKI). Latent BKI is evaluated against similar explicit semantic mapping and VL mapping frameworks on the popular MatterPort-3D and Semantic KITTI data sets, demonstrating that Latent BKI maintains the probabilistic benefits of continuous mapping with the additional benefit of open-dictionary queries. Real-world experiments demonstrate applicability to challenging indoor environments.
☆ MLLM can see? Dynamic Correction Decoding for Hallucination Mitigation
Multimodal Large Language Models (MLLMs) frequently exhibit hallucination phenomena, but the underlying reasons remain poorly understood. In this paper, we present an empirical analysis and find that, although MLLMs incorrectly generate the objects in the final output, they are actually able to recognize visual objects in the preceding layers. We speculate that this may be due to the strong knowledge priors of the language model suppressing the visual information, leading to hallucinations. Motivated by this, we propose a novel dynamic correction decoding method for MLLMs (DeCo), which adaptively selects the appropriate preceding layers and proportionally integrates knowledge into the final layer to adjust the output logits. Note that DeCo is model agnostic and can be seamlessly incorporated with various classic decoding strategies and applied to different MLLMs. We evaluate DeCo on widely-used benchmarks, demonstrating that it can reduce hallucination rates by a large margin compared to baselines, highlighting its potential to mitigate hallucinations. Code is available at https://github.com/zjunlp/DeCo.
comment: Ongoing work
☆ Fractal Calibration for long-tailed object detection
Real-world datasets follow an imbalanced distribution, which poses significant challenges in rare-category object detection. Recent studies tackle this problem by developing re-weighting and re-sampling methods, that utilise the class frequencies of the dataset. However, these techniques focus solely on the frequency statistics and ignore the distribution of the classes in image space, missing important information. In contrast to them, we propose FRActal CALibration (FRACAL): a novel post-calibration method for long-tailed object detection. FRACAL devises a logit adjustment method that utilises the fractal dimension to estimate how uniformly classes are distributed in image space. During inference, it uses the fractal dimension to inversely downweight the probabilities of uniformly spaced class predictions achieving balance in two axes: between frequent and rare categories, and between uniformly spaced and sparsely spaced classes. FRACAL is a post-processing method and it does not require any training, also it can be combined with many off-the-shelf models such as one-stage sigmoid detectors and two-stage instance segmentation models. FRACAL boosts the rare class performance by up to 8.6% and surpasses all previous methods on LVIS dataset, while showing good generalisation to other datasets such as COCO, V3Det and OpenImages. The code will be released.
☆ DPD-NeuralEngine: A 22-nm 6.6-TOPS/W/mm$^2$ Recurrent Neural Network Accelerator for Wideband Power Amplifier Digital Pre-Distortion
The increasing adoption of Deep Neural Network (DNN)-based Digital Pre-distortion (DPD) in modern communication systems necessitates efficient hardware implementations. This paper presents DPD-NeuralEngine, an ultra-fast, tiny-area, and power-efficient DPD accelerator based on a Gated Recurrent Unit (GRU) neural network (NN). Leveraging a co-designed software and hardware approach, our 22 nm CMOS implementation operates at 2 GHz, capable of processing I/Q signals up to 250 MSps. Experimental results demonstrate a throughput of 256.5 GOPS and power efficiency of 1.32 TOPS/W with DPD linearization performance measured in Adjacent Channel Power Ratio (ACPR) of -45.3 dBc and Error Vector Magnitude (EVM) of -39.8 dB. To our knowledge, this work represents the first AI-based DPD application-specific integrated circuit (ASIC) accelerator, achieving a power-area efficiency (PAE) of 6.6 TOPS/W/mm$^2$.
comment: 5 pages, 5 figures
☆ SlideChat: A Large Vision-Language Assistant for Whole-Slide Pathology Image Understanding
Despite the progress made by multimodal large language models (MLLMs) in computational pathology, they remain limited by a predominant focus on patch-level analysis, missing essential contextual information at the whole-slide level. The lack of large-scale instruction datasets and the gigapixel scale of whole slide images (WSIs) pose significant developmental challenges. In this paper, we present SlideChat, the first vision-language assistant capable of understanding gigapixel whole-slide images, exhibiting excellent multimodal conversational capability and response complex instruction across diverse pathology scenarios. To support its development, we created SlideInstruction, the largest instruction-following dataset for WSIs consisting of 4.2K WSI captions and 176K VQA pairs with multiple categories. Furthermore, we propose SlideBench, a multimodal benchmark that incorporates captioning and VQA tasks to assess SlideChat's capabilities in varied clinical settings such as microscopy, diagnosis. Compared to both general and specialized MLLMs, SlideChat exhibits exceptional capabilities achieving state-of-the-art performance on 18 of 22 tasks. For example, it achieved an overall accuracy of 81.17% on SlideBench-VQA (TCGA), and 54.15% on SlideBench-VQA (BCNB). We will fully release SlideChat, SlideInstruction and SlideBench as open-source resources to facilitate research and development in computational pathology.
☆ Latent Action Pretraining from Videos
We introduce Latent Action Pretraining for general Action models (LAPA), an unsupervised method for pretraining Vision-Language-Action (VLA) models without ground-truth robot action labels. Existing Vision-Language-Action models require action labels typically collected by human teleoperators during pretraining, which significantly limits possible data sources and scale. In this work, we propose a method to learn from internet-scale videos that do not have robot action labels. We first train an action quantization model leveraging VQ-VAE-based objective to learn discrete latent actions between image frames, then pretrain a latent VLA model to predict these latent actions from observations and task descriptions, and finally finetune the VLA on small-scale robot manipulation data to map from latent to robot actions. Experimental results demonstrate that our method significantly outperforms existing techniques that train robot manipulation policies from large-scale videos. Furthermore, it outperforms the state-of-the-art VLA model trained with robotic action labels on real-world manipulation tasks that require language conditioning, generalization to unseen objects, and semantic generalization to unseen instructions. Training only on human manipulation videos also shows positive transfer, opening up the potential for leveraging web-scale data for robotics foundation model.
comment: Website: https://latentactionpretraining.github.io
☆ POLO -- Point-based, multi-class animal detection ECCV 2024
Automated wildlife surveys based on drone imagery and object detection technology are a powerful and increasingly popular tool in conservation biology. Most detectors require training images with annotated bounding boxes, which are tedious, expensive, and not always unambiguous to create. To reduce the annotation load associated with this practice, we develop POLO, a multi-class object detection model that can be trained entirely on point labels. POLO is based on simple, yet effective modifications to the YOLOv8 architecture, including alterations to the prediction process, training losses, and post-processing. We test POLO on drone recordings of waterfowl containing up to multiple thousands of individual birds in one image and compare it to a regular YOLOv8. Our experiments show that at the same annotation cost, POLO achieves improved accuracy in counting animals in aerial imagery.
comment: Published in the CV4Ecology workshop at ECCV 2024
☆ Patch-Based Diffusion Models Beat Whole-Image Models for Mismatched Distribution Inverse Problems
Diffusion models have achieved excellent success in solving inverse problems due to their ability to learn strong image priors, but existing approaches require a large training dataset of images that should come from the same distribution as the test dataset. When the training and test distributions are mismatched, artifacts and hallucinations can occur in reconstructed images due to the incorrect priors. In this work, we systematically study out of distribution (OOD) problems where a known training distribution is first provided. We first study the setting where only a single measurement obtained from the unknown test distribution is available. Next we study the setting where a very small sample of data belonging to the test distribution is available, and our goal is still to reconstruct an image from a measurement that came from the test distribution. In both settings, we use a patch-based diffusion prior that learns the image distribution solely from patches. Furthermore, in the first setting, we include a self-supervised loss that helps the network output maintain consistency with the measurement. Extensive experiments show that in both settings, the patch-based method can obtain high quality image reconstructions that can outperform whole-image models and can compete with methods that have access to large in-distribution training datasets. Furthermore, we show how whole-image models are prone to memorization and overfitting, leading to artifacts in the reconstructions, while a patch-based model can resolve these issues.
☆ YOLO-ELA: Efficient Local Attention Modeling for High-Performance Real-Time Insulator Defect Detection
Existing detection methods for insulator defect identification from unmanned aerial vehicles (UAV) struggle with complex background scenes and small objects, leading to suboptimal accuracy and a high number of false positives detection. Using the concept of local attention modeling, this paper proposes a new attention-based foundation architecture, YOLO-ELA, to address this issue. The Efficient Local Attention (ELA) blocks were added into the neck part of the one-stage YOLOv8 architecture to shift the model's attention from background features towards features of insulators with defects. The SCYLLA Intersection-Over-Union (SIoU) criterion function was used to reduce detection loss, accelerate model convergence, and increase the model's sensitivity towards small insulator defects, yielding higher true positive outcomes. Due to a limited dataset, data augmentation techniques were utilized to increase the diversity of the dataset. In addition, we leveraged the transfer learning strategy to improve the model's performance. Experimental results on high-resolution UAV images show that our method achieved a state-of-the-art performance of 96.9% mAP0.5 and a real-time detection speed of 74.63 frames per second, outperforming the baseline model. This further demonstrates the effectiveness of attention-based convolutional neural networks (CNN) in object detection tasks.
☆ RClicks: Realistic Click Simulation for Benchmarking Interactive Segmentation NeurIPS 2024
The emergence of Segment Anything (SAM) sparked research interest in the field of interactive segmentation, especially in the context of image editing tasks and speeding up data annotation. Unlike common semantic segmentation, interactive segmentation methods allow users to directly influence their output through prompts (e.g. clicks). However, click patterns in real-world interactive segmentation scenarios remain largely unexplored. Most methods rely on the assumption that users would click in the center of the largest erroneous area. Nevertheless, recent studies show that this is not always the case. Thus, methods may have poor performance in real-world deployment despite high metrics in a baseline benchmark. To accurately simulate real-user clicks, we conducted a large crowdsourcing study of click patterns in an interactive segmentation scenario and collected 475K real-user clicks. Drawing on ideas from saliency tasks, we develop a clickability model that enables sampling clicks, which closely resemble actual user inputs. Using our model and dataset, we propose RClicks benchmark for a comprehensive comparison of existing interactive segmentation methods on realistic clicks. Specifically, we evaluate not only the average quality of methods, but also the robustness w.r.t. click patterns. According to our benchmark, in real-world usage interactive segmentation models may perform worse than it has been reported in the baseline benchmark, and most of the methods are not robust. We believe that RClicks is a significant step towards creating interactive segmentation methods that provide the best user experience in real-world cases.
comment: Accepted by NeurIPS 2024
☆ Robotic Arm Platform for Multi-View Image Acquisition and 3D Reconstruction in Minimally Invasive Surgery
Minimally invasive surgery (MIS) offers significant benefits such as reduced recovery time and minimised patient trauma, but poses challenges in visibility and access, making accurate 3D reconstruction a significant tool in surgical planning and navigation. This work introduces a robotic arm platform for efficient multi-view image acquisition and precise 3D reconstruction in MIS settings. We adapted a laparoscope to a robotic arm and captured ex-vivo images of several ovine organs across varying lighting conditions (operating room and laparoscopic) and trajectories (spherical and laparoscopic). We employed recently released learning-based feature matchers combined with COLMAP to produce our reconstructions. The reconstructions were evaluated against high-precision laser scans for quantitative evaluation. Our results show that whilst reconstructions suffer most under realistic MIS lighting and trajectory, many versions of our pipeline achieve close to sub-millimetre accuracy with an average of 1.05 mm Root Mean Squared Error and 0.82 mm Chamfer distance. Our best reconstruction results occur with operating room lighting and spherical trajectories. Our robotic platform provides a tool for controlled, repeatable multi-view data acquisition for 3D generation in MIS environments which we hope leads to new datasets for training learning-based models.
comment: 8 pages, 5 figures, 3 tables. This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible
☆ It's Just Another Day: Unique Video Captioning by Discriminative Prompting ACCV 2024
Long videos contain many repeating actions, events and shots. These repetitions are frequently given identical captions, which makes it difficult to retrieve the exact desired clip using a text search. In this paper, we formulate the problem of unique captioning: Given multiple clips with the same caption, we generate a new caption for each clip that uniquely identifies it. We propose Captioning by Discriminative Prompting (CDP), which predicts a property that can separate identically captioned clips, and use it to generate unique captions. We introduce two benchmarks for unique captioning, based on egocentric footage and timeloop movies - where repeating actions are common. We demonstrate that captions generated by CDP improve text-to-video R@1 by 15% for egocentric videos and 10% in timeloop movies.
comment: ACCV 2024 Oral. Project page: https://tobyperrett.github.io/its-just-another-day/
☆ Magnifier Prompt: Tackling Multimodal Hallucination via Extremely Simple Instructions
Hallucinations in multimodal large language models (MLLMs) hinder their practical applications. To address this, we propose a Magnifier Prompt (MagPrompt), a simple yet effective method to tackle hallucinations in MLLMs via extremely simple instructions. MagPrompt is based on the following two key principles, which guide the design of various effective prompts, demonstrating robustness: (1) MLLMs should focus more on the image. (2) When there are conflicts between the image and the model's inner knowledge, MLLMs should prioritize the image. MagPrompt is training-free and can be applied to open-source and closed-source models, such as GPT-4o and Gemini-pro. It performs well across many datasets and its effectiveness is comparable or even better than more complex methods like VCD. Furthermore, our prompt design principles and experimental analyses provide valuable insights into multimodal hallucination.
comment: 9 pages, 13 tables, 4 figures
☆ Visual Fixation-Based Retinal Prosthetic Simulation
This study proposes a retinal prosthetic simulation framework driven by visual fixations, inspired by the saccade mechanism, and assesses performance improvements through end-to-end optimization in a classification task. Salient patches are predicted from input images using the self-attention map of a vision transformer to mimic visual fixations. These patches are then encoded by a trainable U-Net and simulated using the pulse2percept framework to predict visual percepts. By incorporating a learnable encoder, we aim to optimize the visual information transmitted to the retinal implant, addressing both the limited resolution of the electrode array and the distortion between the input stimuli and resulting phosphenes. The predicted percepts are evaluated using the self-supervised DINOv2 foundation model, with an optional learnable linear layer for classification accuracy. On a subset of the ImageNet validation set, the fixation-based framework achieves a classification accuracy of 87.72%, using computational parameters based on a real subject's physiological data, significantly outperforming the downsampling-based accuracy of 40.59% and approaching the healthy upper bound of 92.76%. Our approach shows promising potential for producing more semantically understandable percepts with the limited resolution available in retinal prosthetics.
☆ A Survey of Low-shot Vision-Language Model Adaptation via Representer Theorem
The advent of pre-trained vision-language foundation models has revolutionized the field of zero/few-shot (i.e., low-shot) image recognition. The key challenge to address under the condition of limited training data is how to fine-tune pre-trained vision-language models in a parameter-efficient manner. Previously, numerous approaches tackling this challenge have been proposed. Meantime, a few survey papers are also published to summarize these works. However, there still lacks a unified computational framework to integrate existing methods together, identify their nature and support in-depth comparison. As such, this survey paper first proposes a unified computational framework from the perspective of Representer Theorem and then derives many of the existing methods by specializing this framework. Thereafter, a comparative analysis is conducted to uncover the differences and relationships between existing methods. Based on the analyses, some possible variants to improve the existing works are presented. As a demonstration, we extend existing methods by modeling inter-class correlation between representers in reproducing kernel Hilbert space (RKHS), which is implemented by exploiting the closed-form solution of kernel ridge regression. Extensive experiments on 11 datasets are conducted to validate the effectiveness of this method. Toward the end of this paper, we discuss the limitations and provide further research directions.
☆ SurFhead: Affine Rig Blending for Geometrically Accurate 2D Gaussian Surfel Head Avatars
Recent advancements in head avatar rendering using Gaussian primitives have achieved significantly high-fidelity results. Although precise head geometry is crucial for applications like mesh reconstruction and relighting, current methods struggle to capture intricate geometric details and render unseen poses due to their reliance on similarity transformations, which cannot handle stretch and shear transforms essential for detailed deformations of geometry. To address this, we propose SurFhead, a novel method that reconstructs riggable head geometry from RGB videos using 2D Gaussian surfels, which offer well-defined geometric properties, such as precise depth from fixed ray intersections and normals derived from their surface orientation, making them advantageous over 3D counterparts. SurFhead ensures high-fidelity rendering of both normals and images, even in extreme poses, by leveraging classical mesh-based deformation transfer and affine transformation interpolation. SurFhead introduces precise geometric deformation and blends surfels through polar decomposition of transformations, including those affecting normals. Our key contribution lies in bridging classical graphics techniques, such as mesh-based deformation, with modern Gaussian primitives, achieving state-of-the-art geometry reconstruction and rendering quality. Unlike previous avatar rendering approaches, SurFhead enables efficient reconstruction driven by Gaussian primitives while preserving high-fidelity geometry.
☆ Leveraging Structure Knowledge and Deep Models for the Detection of Abnormal Handwritten Text
Currently, the destruction of the sequence structure in handwritten text has become one of the main bottlenecks restricting the recognition task. The typical situations include additional specific markers (the text swapping modification) and the text overlap caused by character modifications like deletion, replacement, and insertion. In this paper, we propose a two-stage detection algorithm that combines structure knowledge and deep models for the above mentioned text. Firstly, different structure prototypes are roughly located from handwritten text images. Based on the detection results of the first stage, in the second stage, we adopt different strategies. Specifically, a shape regression network trained by a novel semi-supervised contrast training strategy is introduced and the positional relationship between the characters is fully employed. Experiments on two handwritten text datasets show that the proposed method can greatly improve the detection performance. The new dataset is available at https://github.com/Wukong90.
☆ Degradation Oriented and Regularized Network for Real-World Depth Super-Resolution
Recently, existing RGB-guided depth super-resolution methods achieve excellent performance based on the assumption of fixed and known degradation (e.g., bicubic downsampling). However, in real-world scenarios, the captured depth often suffers from unconventional and agnostic degradation due to sensor limitations and the complexity of imaging environments (e.g., low reflective surface, illumination). Their performance significantly declines when these real degradation differ from their assumptions. To address these issues, we propose a Degradation Oriented and Regularized Network, DORNet, which pays more attention on learning degradation representation of low-resolution depth that can provide targeted guidance for depth recovery. Specifically, we first design a self-supervised Degradation Learning to model the discriminative degradation representation of low-resolution depth using routing selection-based Degradation Regularization. Then, we present a Degradation Awareness that recursively conducts multiple Degradation-Oriented Feature Transformations, each of which selectively embeds RGB information into the depth based on the learned degradation representation. Extensive experimental results on both real and synthetic datasets demonstrate that our method achieves state-of-the-art performance.
comment: 10 pages
☆ VisualRWKV-HD and UHD: Advancing High-Resolution Processing for Visual Language Models
Accurately understanding complex visual information is crucial for visual language models (VLMs). Enhancing image resolution can improve visual perception capabilities, not only reducing hallucinations but also boosting performance in tasks that demand high resolution, such as text-rich or document analysis. In this paper, we present VisualRWKV-HD and VisualRWKV-UHD, two advancements in the VisualRWKV model family, specifically designed to process high-resolution visual inputs. For VisualRWKV-HD, we developed a lossless downsampling method to effectively integrate a high-resolution vision encoder with low-resolution encoders, without extending the input sequence length. For the VisualRWKV-UHD model, we enhanced image representation by dividing the image into four segments, which are then recombined with the original image. This technique allows the model to incorporate both high-resolution and low-resolution features, effectively balancing coarse and fine-grained information. As a result, the model supports resolutions up to 4096 x 4096 pixels, offering a more detailed and comprehensive visual processing capability. Both VisualRWKV-HD and VisualRWKV-UHD not only achieve strong results on VLM benchmarks but also show marked improvements in performance for text-rich tasks.
☆ Unveiling the Mystery of Visual Attributes of Concrete and Abstract Concepts: Variability, Nearest Neighbors, and Challenging Categories
The visual representation of a concept varies significantly depending on its meaning and the context where it occurs; this poses multiple challenges both for vision and multimodal models. Our study focuses on concreteness, a well-researched lexical-semantic variable, using it as a case study to examine the variability in visual representations. We rely on images associated with approximately 1,000 abstract and concrete concepts extracted from two different datasets: Bing and YFCC. Our goals are: (i) evaluate whether visual diversity in the depiction of concepts can reliably distinguish between concrete and abstract concepts; (ii) analyze the variability of visual features across multiple images of the same concept through a nearest neighbor analysis; and (iii) identify challenging factors contributing to this variability by categorizing and annotating images. Our findings indicate that for classifying images of abstract versus concrete concepts, a combination of basic visual features such as color and texture is more effective than features extracted by more complex models like Vision Transformer (ViT). However, ViTs show better performances in the nearest neighbor analysis, emphasizing the need for a careful selection of visual features when analyzing conceptual variables through modalities other than text.
☆ RS-MOCO: A deep learning-based topology-preserving image registration method for cardiac T1 mapping
Cardiac T1 mapping can evaluate various clinical symptoms of myocardial tissue. However, there is currently a lack of effective, robust, and efficient methods for motion correction in cardiac T1 mapping. In this paper, we propose a deep learning-based and topology-preserving image registration framework for motion correction in cardiac T1 mapping. Notably, our proposed implicit consistency constraint dubbed BLOC, to some extent preserves the image topology in registration by bidirectional consistency constraint and local anti-folding constraint. To address the contrast variation issue, we introduce a weighted image similarity metric for multimodal registration of cardiac T1-weighted images. Besides, a semi-supervised myocardium segmentation network and a dual-domain attention module are integrated into the framework to further improve the performance of the registration. Numerous comparative experiments, as well as ablation studies, demonstrated the effectiveness and high robustness of our method. The results also indicate that the proposed weighted image similarity metric, specifically crafted for our network, contributes a lot to the enhancement of the motion correction efficacy, while the bidirectional consistency constraint combined with the local anti-folding constraint ensures a more desirable topology-preserving registration mapping.
☆ ED-ViT: Splitting Vision Transformer for Distributed Inference on Edge Devices
Deep learning models are increasingly deployed on resource-constrained edge devices for real-time data analytics. In recent years, Vision Transformer models and their variants have demonstrated outstanding performance across various computer vision tasks. However, their high computational demands and inference latency pose significant challenges for model deployment on resource-constraint edge devices. To address this issue, we propose a novel Vision Transformer splitting framework, ED-ViT, designed to execute complex models across multiple edge devices efficiently. Specifically, we partition Vision Transformer models into several sub-models, where each sub-model is tailored to handle a specific subset of data classes. To further minimize computation overhead and inference latency, we introduce a class-wise pruning technique that reduces the size of each sub-model. We conduct extensive experiments on five datasets with three model structures, demonstrating that our approach significantly reduces inference latency on edge devices and achieves a model size reduction of up to 28.9 times and 34.1 times, respectively, while maintaining test accuracy comparable to the original Vision Transformer. Additionally, we compare ED-ViT with two state-of-the-art methods that deploy CNN and SNN models on edge devices, evaluating accuracy, inference time, and overall model size. Our comprehensive evaluation underscores the effectiveness of the proposed ED-ViT framework.
comment: 14 pages, 8 figures
☆ Feature-guided score diffusion for sampling conditional densities
Score diffusion methods can learn probability densities from samples. The score of the noise-corrupted density is estimated using a deep neural network, which is then used to iteratively transport a Gaussian white noise density to a target density. Variants for conditional densities have been developed, but correct estimation of the corresponding scores is difficult. We avoid these difficulties by introducing an algorithm that guides the diffusion with a projected score. The projection pushes the image feature vector towards the feature vector centroid of the target class. The projected score and the feature vectors are learned by the same network. Specifically, the image feature vector is defined as the spatial averages of the channels activations in select layers of the network. Optimizing the projected score for denoising loss encourages image feature vectors of each class to cluster around their centroids. It also leads to the separations of the centroids. We show that these centroids provide a low-dimensional Euclidean embedding of the class conditional densities. We demonstrate that the algorithm can generate high quality and diverse samples from the conditioning class. Conditional generation can be performed using feature vectors interpolated between those of the training set, demonstrating out-of-distribution generalization.
☆ Efficient and Effective Universal Adversarial Attack against Vision-Language Pre-training Models
Vision-language pre-training (VLP) models, trained on large-scale image-text pairs, have become widely used across a variety of downstream vision-and-language (V+L) tasks. This widespread adoption raises concerns about their vulnerability to adversarial attacks. Non-universal adversarial attacks, while effective, are often impractical for real-time online applications due to their high computational demands per data instance. Recently, universal adversarial perturbations (UAPs) have been introduced as a solution, but existing generator-based UAP methods are significantly time-consuming. To overcome the limitation, we propose a direct optimization-based UAP approach, termed DO-UAP, which significantly reduces resource consumption while maintaining high attack performance. Specifically, we explore the necessity of multimodal loss design and introduce a useful data augmentation strategy. Extensive experiments conducted on three benchmark VLP datasets, six popular VLP models, and three classical downstream tasks demonstrate the efficiency and effectiveness of DO-UAP. Specifically, our approach drastically decreases the time consumption by 23-fold while achieving a better attack performance.
comment: 11 pages
☆ Simultaneous Diffusion Sampling for Conditional LiDAR Generation
By enabling capturing of 3D point clouds that reflect the geometry of the immediate environment, LiDAR has emerged as a primary sensor for autonomous systems. If a LiDAR scan is too sparse, occluded by obstacles, or too small in range, enhancing the point cloud scan by while respecting the geometry of the scene is useful for downstream tasks. Motivated by the explosive growth of interest in generative methods in vision, conditional LiDAR generation is starting to take off. This paper proposes a novel simultaneous diffusion sampling methodology to generate point clouds conditioned on the 3D structure of the scene as seen from multiple views. The key idea is to impose multi-view geometric constraints on the generation process, exploiting mutual information for enhanced results. Our method begins by recasting the input scan to multiple new viewpoints around the scan, thus creating multiple synthetic LiDAR scans. Then, the synthetic and input LiDAR scans simultaneously undergo conditional generation according to our methodology. Results show that our method can produce accurate and geometrically consistent enhancements to point cloud scans, allowing it to outperform existing methods by a large margin in a variety of benchmarks.
☆ Fast Local Neural Regression for Low-Cost, Path Traced Lambertian Global Illumination
Despite recent advances in hardware acceleration of ray tracing, real-time ray budgets remain stubbornly limited at a handful of samples per pixel (spp) on commodity hardware, placing the onus on denoising algorithms to achieve high visual quality for path traced global illumination. Neural network-based solutions give excellent result quality at the cost of increased execution time relative to hand-engineered methods, making them less suitable for deployment on resource-constrained systems. We therefore propose incorporating a neural network into a computationally-efficient local linear model-based denoiser, and demonstrate faithful single-frame reconstruction of global illumination for Lambertian scenes at very low sample counts (1spp) and for low computational cost. Other contributions include improving the quality and performance of local linear model-based denoising through a simplified mathematical treatment, and demonstration of the surprising usefulness of ambient occlusion as a guide channel. We also show how our technique is straightforwardly extensible to joint denoising and upsampling of path traced renders with reference to low-cost, rasterized guide channels.
comment: 11 pages, 10 figures, 1 table
☆ VidEgoThink: Assessing Egocentric Video Understanding Capabilities for Embodied AI
Recent advancements in Multi-modal Large Language Models (MLLMs) have opened new avenues for applications in Embodied AI. Building on previous work, EgoThink, we introduce VidEgoThink, a comprehensive benchmark for evaluating egocentric video understanding capabilities. To bridge the gap between MLLMs and low-level control in Embodied AI, we design four key interrelated tasks: video question-answering, hierarchy planning, visual grounding and reward modeling. To minimize manual annotation costs, we develop an automatic data generation pipeline based on the Ego4D dataset, leveraging the prior knowledge and multimodal capabilities of GPT-4o. Three human annotators then filter the generated data to ensure diversity and quality, resulting in the VidEgoThink benchmark. We conduct extensive experiments with three types of models: API-based MLLMs, open-source image-based MLLMs, and open-source video-based MLLMs. Experimental results indicate that all MLLMs, including GPT-4o, perform poorly across all tasks related to egocentric video understanding. These findings suggest that foundation models still require significant advancements to be effectively applied to first-person scenarios in Embodied AI. In conclusion, VidEgoThink reflects a research trend towards employing MLLMs for egocentric vision, akin to human capabilities, enabling active observation and interaction in the complex real-world environments.
☆ MultiVENT 2.0: A Massive Multilingual Benchmark for Event-Centric Video Retrieval
Efficiently retrieving and synthesizing information from large-scale multimodal collections has become a critical challenge. However, existing video retrieval datasets suffer from scope limitations, primarily focusing on matching descriptive but vague queries with small collections of professionally edited, English-centric videos. To address this gap, we introduce $\textbf{MultiVENT 2.0}$, a large-scale, multilingual event-centric video retrieval benchmark featuring a collection of more than 218,000 news videos and 3,906 queries targeting specific world events. These queries specifically target information found in the visual content, audio, embedded text, and text metadata of the videos, requiring systems leverage all these sources to succeed at the task. Preliminary results show that state-of-the-art vision-language models struggle significantly with this task, and while alternative approaches show promise, they are still insufficient to adequately address this problem. These findings underscore the need for more robust multimodal retrieval systems, as effective video retrieval is a crucial step towards multimodal content understanding and generation tasks.
☆ Depth Estimation From Monocular Images With Enhanced Encoder-Decoder Architecture
Estimating depth from a single 2D image is a challenging task because of the need for stereo or multi-view data, which normally provides depth information. This paper deals with this challenge by introducing a novel deep learning-based approach using an encoder-decoder architecture, where the Inception-ResNet-v2 model is utilized as the encoder. According to the available literature, this is the first instance of using Inception-ResNet-v2 as an encoder for monocular depth estimation, illustrating better performance than previous models. The use of Inception-ResNet-v2 enables our model to capture complex objects and fine-grained details effectively that are generally difficult to predict. Besides, our model incorporates multi-scale feature extraction to enhance depth prediction accuracy across different kinds of object sizes and distances. We propose a composite loss function consisting of depth loss, gradient edge loss, and SSIM loss, where the weights are fine-tuned to optimize the weighted sum, ensuring better balance across different aspects of depth estimation. Experimental results on the NYU Depth V2 dataset show that our model achieves state-of-the-art performance, with an ARE of 0.064, RMSE of 0.228, and accuracy ($\delta$ $<1.25$) of 89.3%. These metrics demonstrate that our model effectively predicts depth, even in challenging circumstances, providing a scalable solution for real-world applications in robotics, 3D reconstruction, and augmented reality.
☆ PaSTe: Improving the Efficiency of Visual Anomaly Detection at the Edge
Visual Anomaly Detection (VAD) has gained significant research attention for its ability to identify anomalous images and pinpoint the specific areas responsible for the anomaly. A key advantage of VAD is its unsupervised nature, which eliminates the need for costly and time-consuming labeled data collection. However, despite its potential for real-world applications, the literature has given limited focus to resource-efficient VAD, particularly for deployment on edge devices. This work addresses this gap by leveraging lightweight neural networks to reduce memory and computation requirements, enabling VAD deployment on resource-constrained edge devices. We benchmark the major VAD algorithms within this framework and demonstrate the feasibility of edge-based VAD using the well-known MVTec dataset. Furthermore, we introduce a novel algorithm, Partially Shared Teacher-student (PaSTe), designed to address the high resource demands of the existing Student Teacher Feature Pyramid Matching (STFPM) approach. Our results show that PaSTe decreases the inference time by 25%, while reducing the training time by 33% and peak RAM usage during training by 76%. These improvements make the VAD process significantly more efficient, laying a solid foundation for real-world deployment on edge devices.
comment: 13 pages, 6 figures
☆ Breaking Modality Gap in RGBT Tracking: Coupled Knowledge Distillation ACM MM2024
Modality gap between RGB and thermal infrared (TIR) images is a crucial issue but often overlooked in existing RGBT tracking methods. It can be observed that modality gap mainly lies in the image style difference. In this work, we propose a novel Coupled Knowledge Distillation framework called CKD, which pursues common styles of different modalities to break modality gap, for high performance RGBT tracking. In particular, we introduce two student networks and employ the style distillation loss to make their style features consistent as much as possible. Through alleviating the style difference of two student networks, we can break modality gap of different modalities well. However, the distillation of style features might harm to the content representations of two modalities in student networks. To handle this issue, we take original RGB and TIR networks as the teachers, and distill their content knowledge into two student networks respectively by the style-content orthogonal feature decoupling scheme. We couple the above two distillation processes in an online optimization framework to form new feature representations of RGB and thermal modalities without modality gap. In addition, we design a masked modeling strategy and a multi-modal candidate token elimination strategy into CKD to improve tracking robustness and efficiency respectively. Extensive experiments on five standard RGBT tracking datasets validate the effectiveness of the proposed method against state-of-the-art methods while achieving the fastest tracking speed of 96.4 FPS. Code available at https://github.com/Multi-Modality-Tracking/CKD.
comment: Accepted by ACM MM2024
☆ DeformPAM: Data-Efficient Learning for Long-horizon Deformable Object Manipulation via Preference-based Action Alignment
In recent years, imitation learning has made progress in the field of robotic manipulation. However, it still faces challenges when dealing with complex long-horizon deformable object tasks, such as high-dimensional state spaces, complex dynamics, and multimodal action distributions. Traditional imitation learning methods often require a large amount of data and encounter distributional shifts and accumulative errors in these tasks. To address these issues, we propose a data-efficient general learning framework (DeformPAM) based on preference learning and reward-guided action selection. DeformPAM decomposes long-horizon tasks into multiple action primitives, utilizes 3D point cloud inputs and diffusion models to model action distributions, and trains an implicit reward model using human preference data. During the inference phase, the reward model scores multiple candidate actions, selecting the optimal action for execution, thereby reducing the occurrence of anomalous actions and improving task completion quality. Experiments conducted on three challenging real-world long-horizon deformable object manipulation tasks demonstrate the effectiveness of this method. Results show that DeformPAM improves both task completion quality and efficiency compared to baseline methods even with limited data. Code and data will be available at https://deform-pam.robotflow.ai.
☆ On-the-fly Modulation for Balanced Multimodal Learning
Multimodal learning is expected to boost model performance by integrating information from different modalities. However, its potential is not fully exploited because the widely-used joint training strategy, which has a uniform objective for all modalities, leads to imbalanced and under-optimized uni-modal representations. Specifically, we point out that there often exists modality with more discriminative information, e.g., vision of playing football and sound of blowing wind. They could dominate the joint training process, resulting in other modalities being significantly under-optimized. To alleviate this problem, we first analyze the under-optimized phenomenon from both the feed-forward and the back-propagation stages during optimization. Then, On-the-fly Prediction Modulation (OPM) and On-the-fly Gradient Modulation (OGM) strategies are proposed to modulate the optimization of each modality, by monitoring the discriminative discrepancy between modalities during training. Concretely, OPM weakens the influence of the dominant modality by dropping its feature with dynamical probability in the feed-forward stage, while OGM mitigates its gradient in the back-propagation stage. In experiments, our methods demonstrate considerable improvement across a variety of multimodal tasks. These simple yet effective strategies not only enhance performance in vanilla and task-oriented multimodal models, but also in more complex multimodal tasks, showcasing their effectiveness and flexibility. The source code is available at \url{https://github.com/GeWu-Lab/BML_TPAMI2024}.
comment: Accepted by T-PAMI 2024
☆ PAVLM: Advancing Point Cloud based Affordance Understanding Via Vision-Language Model
Affordance understanding, the task of identifying actionable regions on 3D objects, plays a vital role in allowing robotic systems to engage with and operate within the physical world. Although Visual Language Models (VLMs) have excelled in high-level reasoning and long-horizon planning for robotic manipulation, they still fall short in grasping the nuanced physical properties required for effective human-robot interaction. In this paper, we introduce PAVLM (Point cloud Affordance Vision-Language Model), an innovative framework that utilizes the extensive multimodal knowledge embedded in pre-trained language models to enhance 3D affordance understanding of point cloud. PAVLM integrates a geometric-guided propagation module with hidden embeddings from large language models (LLMs) to enrich visual semantics. On the language side, we prompt Llama-3.1 models to generate refined context-aware text, augmenting the instructional input with deeper semantic cues. Experimental results on the 3D-AffordanceNet benchmark demonstrate that PAVLM outperforms baseline methods for both full and partial point clouds, particularly excelling in its generalization to novel open-world affordance tasks of 3D objects. For more information, visit our project site: pavlm-source.github.io.
☆ PSVMA+: Exploring Multi-granularity Semantic-visual Adaption for Generalized Zero-shot Learning
Generalized zero-shot learning (GZSL) endeavors to identify the unseen categories using knowledge from the seen domain, necessitating the intrinsic interactions between the visual features and attribute semantic features. However, GZSL suffers from insufficient visual-semantic correspondences due to the attribute diversity and instance diversity. Attribute diversity refers to varying semantic granularity in attribute descriptions, ranging from low-level (specific, directly observable) to high-level (abstract, highly generic) characteristics. This diversity challenges the collection of adequate visual cues for attributes under a uni-granularity. Additionally, diverse visual instances corresponding to the same sharing attributes introduce semantic ambiguity, leading to vague visual patterns. To tackle these problems, we propose a multi-granularity progressive semantic-visual mutual adaption (PSVMA+) network, where sufficient visual elements across granularity levels can be gathered to remedy the granularity inconsistency. PSVMA+ explores semantic-visual interactions at different granularity levels, enabling awareness of multi-granularity in both visual and semantic elements. At each granularity level, the dual semantic-visual transformer module (DSVTM) recasts the sharing attributes into instance-centric attributes and aggregates the semantic-related visual regions, thereby learning unambiguous visual features to accommodate various instances. Given the diverse contributions of different granularities, PSVMA+ employs selective cross-granularity learning to leverage knowledge from reliable granularities and adaptively fuses multi-granularity features for comprehensive representations. Experimental results demonstrate that PSVMA+ consistently outperforms state-of-the-art methods.
comment: Accepted to TPAMI 2024. arXiv admin note: text overlap with arXiv:2303.15322
☆ Efficiera Residual Networks: Hardware-Friendly Fully Binary Weight with 2-bit Activation Model Achieves Practical ImageNet Accuracy
The edge-device environment imposes severe resource limitations, encompassing computation costs, hardware resource usage, and energy consumption for deploying deep neural network models. Ultra-low-bit quantization and hardware accelerators have been explored as promising approaches to address these challenges. Ultra-low-bit quantization significantly reduces the model size and the computational cost. Despite progress so far, many competitive ultra-low-bit models still partially rely on float or non-ultra-low-bit quantized computation such as the input and output layer. We introduce Efficiera Residual Networks (ERNs), a model optimized for low-resource edge devices. ERNs achieve full ultra-low-bit quantization, with all weights, including the initial and output layers, being binary, and activations set at 2 bits. We introduce the shared constant scaling factor technique to enable integer-valued computation in residual connections, allowing our model to operate without float values until the final convolution layer. Demonstrating competitiveness, ERNs achieve an ImageNet top-1 accuracy of 72.5pt with a ResNet50-compatible architecture and 63.6pt with a model size less than 1MB. Moreover, ERNs exhibit impressive inference times, reaching 300FPS with the smallest model and 60FPS with the largest model on a cost-efficient FPGA device.
comment: 11pages, 2 figures, the model implementation is available at https://github.com/LeapMind/ERN
☆ MCTBench: Multimodal Cognition towards Text-Rich Visual Scenes Benchmark
The comprehension of text-rich visual scenes has become a focal point for evaluating Multi-modal Large Language Models (MLLMs) due to their widespread applications. Current benchmarks tailored to the scenario emphasize perceptual capabilities, while overlooking the assessment of cognitive abilities. To address this limitation, we introduce a Multimodal benchmark towards Text-rich visual scenes, to evaluate the Cognitive capabilities of MLLMs through visual reasoning and content-creation tasks (MCTBench). To mitigate potential evaluation bias from the varying distributions of datasets, MCTBench incorporates several perception tasks (e.g., scene text recognition) to ensure a consistent comparison of both the cognitive and perceptual capabilities of MLLMs. To improve the efficiency and fairness of content-creation evaluation, we conduct an automatic evaluation pipeline. Evaluations of various MLLMs on MCTBench reveal that, despite their impressive perceptual capabilities, their cognition abilities require enhancement. We hope MCTBench will offer the community an efficient resource to explore and enhance cognitive capabilities towards text-rich visual scenes.
comment: 12 pages, 5 figures, project page: https://github.com/xfey/MCTBench?tab=readme-ov-file
☆ Overcoming Domain Limitations in Open-vocabulary Segmentation
Open-vocabulary segmentation (OVS) has gained attention for its ability to recognize a broader range of classes. However, OVS models show significant performance drops when applied to unseen domains beyond the previous training dataset. Fine-tuning these models on new datasets can improve performance, but often leads to the catastrophic forgetting of previously learned knowledge. To address this issue, we propose a method that allows OVS models to learn information from new domains while preserving prior knowledge. Our approach begins by evaluating the input sample's proximity to multiple domains, using precomputed multivariate normal distributions for each domain. Based on this prediction, we dynamically interpolate between the weights of the pre-trained decoder and the fine-tuned decoders. Extensive experiments demonstrate that this approach allows OVS models to adapt to new domains while maintaining performance on the previous training dataset. The source code is available at https://github.com/dongjunhwang/dwi.
☆ Prediction of Cardiovascular Risk Factors from Retinal Fundus Images using CNNs
Early detection of cardiovascular disease risk factors is essential to alter the course of the disease. Previous studies showed that deep learning can successfully be used to detect such risk factors from retinal images. This study uses convolutional neural networks (CNNs) to predict the cardiovascular disease risk factors age, BMI, smoking status, HbA1c, systolic blood pressure, diastolic blood pressure, gender and total cholesterol from retinal images from the UK Biobank data set. By applying contrast enhancement on the retinal images in the form of Gaussian filtering and deriving predictions on individual basis through the combination of left and right retinal image predictions, an increased prediction performance could be derived for the variables age (R2 score of 0.81) and systolic blood pressure (R2 score of 0.39) compared to previous studies using retinal images from the UK Biobank data set. Further, this is the first study that tries to predict HbA1c and total cholesterol from UK Biobank retinal fundus images. For these variables the models achieved an R2 score of 0.0579 for predicting HbA1c and an R2 score of 0.0157 for predicting total cholesterol. These results show that the value of deriving predictions for these two risk factors from retinal fundus images from the UK Biobank data set is limited.
☆ Hairmony: Fairness-aware hairstyle classification
We present a method for prediction of a person's hairstyle from a single image. Despite growing use cases in user digitization and enrollment for virtual experiences, available methods are limited, particularly in the range of hairstyles they can capture. Human hair is extremely diverse and lacks any universally accepted description or categorization, making this a challenging task. Most current methods rely on parametric models of hair at a strand level. These approaches, while very promising, are not yet able to represent short, frizzy, coily hair and gathered hairstyles. We instead choose a classification approach which can represent the diversity of hairstyles required for a truly robust and inclusive system. Previous classification approaches have been restricted by poorly labeled data that lacks diversity, imposing constraints on the usefulness of any resulting enrollment system. We use only synthetic data to train our models. This allows for explicit control of diversity of hairstyle attributes, hair colors, facial appearance, poses, environments and other parameters. It also produces noise-free ground-truth labels. We introduce a novel hairstyle taxonomy developed in collaboration with a diverse group of domain experts which we use to balance our training data, supervise our model, and directly measure fairness. We annotate our synthetic training data and a real evaluation dataset using this taxonomy and release both to enable comparison of future hairstyle prediction approaches. We employ an architecture based on a pre-trained feature extraction network in order to improve generalization of our method to real data and predict taxonomy attributes as an auxiliary task to improve accuracy. Results show our method to be significantly more robust for challenging hairstyles than recent parametric approaches.
☆ Look Ma, no markers: holistic performance capture without the hassle
We tackle the problem of highly-accurate, holistic performance capture for the face, body and hands simultaneously. Motion-capture technologies used in film and game production typically focus only on face, body or hand capture independently, involve complex and expensive hardware and a high degree of manual intervention from skilled operators. While machine-learning-based approaches exist to overcome these problems, they usually only support a single camera, often operate on a single part of the body, do not produce precise world-space results, and rarely generalize outside specific contexts. In this work, we introduce the first technique for marker-free, high-quality reconstruction of the complete human body, including eyes and tongue, without requiring any calibration, manual intervention or custom hardware. Our approach produces stable world-space results from arbitrary camera rigs as well as supporting varied capture environments and clothing. We achieve this through a hybrid approach that leverages machine learning models trained exclusively on synthetic data and powerful parametric models of human shape and motion. We evaluate our method on a number of body, face and hand reconstruction benchmarks and demonstrate state-of-the-art results that generalize on diverse datasets.
☆ Rician Denoising Diffusion Probabilistic Models For Sodium Breast MRI Enhancement
Sodium MRI is an imaging technique used to visualize and quantify sodium concentrations in vivo, playing a role in many biological processes and potentially aiding in breast cancer characterization. Sodium MRI, however, suffers from inherently low signal-to-noise ratios (SNR) and spatial resolution, compared with conventional proton MRI. A deep-learning method, the Denoising Diffusion Probabilistic Models (DDPM), has demonstrated success across a wide range of denoising tasks, yet struggles with sodium MRI's unique noise profile, as DDPM primarily targets Gaussian noise. DDPM can distort features when applied to sodium MRI. This paper advances the DDPM by introducing the Rician Denoising Diffusion Probabilistic Models (RDDPM) for sodium MRI denoising. RDDPM converts Rician noise to Gaussian noise at each timestep during the denoising process. The model's performance is evaluated using three non-reference image quality assessment metrics, where RDDPM consistently outperforms DDPM and other CNN-based denoising methods.
comment: 3 figures
☆ Dual-Teacher Ensemble Models with Double-Copy-Paste for 3D Semi-Supervised Medical Image Segmentation
Semi-supervised learning (SSL) techniques address the high labeling costs in 3D medical image segmentation, with the teacher-student model being a common approach. However, using an exponential moving average (EMA) in single-teacher models may cause coupling issues, where the weights of the student and teacher models become similar, limiting the teacher's ability to provide additional knowledge for the student. Dual-teacher models were introduced to address this problem but often neglected the importance of maintaining teacher model diversity, leading to coupling issues among teachers. To address the coupling issue, we incorporate a double-copy-paste (DCP) technique to enhance the diversity among the teachers. Additionally, we introduce the Staged Selective Ensemble (SSE) module, which selects different ensemble methods based on the characteristics of the samples and enables more accurate segmentation of label boundaries, thereby improving the quality of pseudo-labels. Experimental results demonstrate the effectiveness of our proposed method in 3D medical image segmentation tasks. Here is the code link: https://github.com/Fazhan-cs/DCP.
comment: 35 pages, 5 figures
☆ Spatio-Temporal Distortion Aware Omnidirectional Video Super-Resolution
Omnidirectional video (ODV) can provide an immersive experience and is widely utilized in the field of virtual reality and augmented reality. However, the restricted capturing devices and transmission bandwidth lead to the low resolution of ODVs. Video super-resolution (VSR) methods are proposed to enhance the resolution of videos, but ODV projection distortions in the application are not well addressed directly applying such methods. To achieve better super-resolution reconstruction quality, we propose a novel Spatio-Temporal Distortion Aware Network (STDAN) oriented to ODV characteristics. Specifically, a spatio-temporal distortion modulation module is introduced to improve spatial ODV projection distortions and exploit the temporal correlation according to intra and inter alignments. Next, we design a multi-frame reconstruction and fusion mechanism to refine the consistency of reconstructed ODV frames. Furthermore, we incorporate latitude-saliency adaptive maps in the loss function to concentrate on important viewpoint regions with higher texture complexity and human-watching interest. In addition, we collect a new ODV-SR dataset with various scenarios. Extensive experimental results demonstrate that the proposed STDAN achieves superior super-resolution performance on ODVs and outperforms state-of-the-art methods.
☆ LoGS: Visual Localization via Gaussian Splatting with Fewer Training Images
Visual localization involves estimating a query image's 6-DoF (degrees of freedom) camera pose, which is a fundamental component in various computer vision and robotic tasks. This paper presents LoGS, a vision-based localization pipeline utilizing the 3D Gaussian Splatting (GS) technique as scene representation. This novel representation allows high-quality novel view synthesis. During the mapping phase, structure-from-motion (SfM) is applied first, followed by the generation of a GS map. During localization, the initial position is obtained through image retrieval, local feature matching coupled with a PnP solver, and then a high-precision pose is achieved through the analysis-by-synthesis manner on the GS map. Experimental results on four large-scale datasets demonstrate the proposed approach's SoTA accuracy in estimating camera poses and robustness under challenging few-shot conditions.
comment: 8 pages
☆ NavTopo: Leveraging Topological Maps For Autonomous Navigation Of a Mobile Robot
Autonomous navigation of a mobile robot is a challenging task which requires ability of mapping, localization, path planning and path following. Conventional mapping methods build a dense metric map like an occupancy grid, which is affected by odometry error accumulation and consumes a lot of memory and computations in large environments. Another approach to mapping is the usage of topological properties, e.g. adjacency of locations in the environment. Topological maps are less prone to odometry error accumulation and high resources consumption, and also enable fast path planning because of the graph sparsity. Based on this idea, we proposed NavTopo - a full navigation pipeline based on topological map and two-level path planning. The pipeline localizes in the graph by matching neural network descriptors and 2D projections of the input point clouds, which significantly reduces memory consumption compared to metric and topological point cloud-based approaches. We test our approach in a large indoor photo-relaistic simulated environment and compare it to a metric map-based approach based on popular metric mapping method RTAB-MAP. The experimental results show that our topological approach significantly outperforms the metric one in terms of performance, keeping proper navigational efficiency.
comment: This paper is published in proceedings of the 9th International Conference "Interactive Collaborative Robotics" (ICR 2024)
☆ Online learning in motion modeling for intra-interventional image sequences MICCAI
Image monitoring and guidance during medical examinations can aid both diagnosis and treatment. However, the sampling frequency is often too low, which creates a need to estimate the missing images. We present a probabilistic motion model for sequential medical images, with the ability to both estimate motion between acquired images and forecast the motion ahead of time. The core is a low-dimensional temporal process based on a linear Gaussian state-space model with analytically tractable solutions for forecasting, simulation, and imputation of missing samples. The results, from two experiments on publicly available cardiac datasets, show reliable motion estimates and an improved forecasting performance using patient-specific adaptation by online learning.
comment: Medical Image Computing and Computer Assisted Intervention (MICCAI) 2024
☆ InvSeg: Test-Time Prompt Inversion for Semantic Segmentation
Visual-textual correlations in the attention maps derived from text-to-image diffusion models are proven beneficial to dense visual prediction tasks, e.g., semantic segmentation. However, a significant challenge arises due to the input distributional discrepancy between the context-rich sentences used for image generation and the isolated class names typically employed in semantic segmentation, hindering the diffusion models from capturing accurate visual-textual correlations. To solve this, we propose InvSeg, a test-time prompt inversion method that tackles open-vocabulary semantic segmentation by inverting image-specific visual context into text prompt embedding space, leveraging structure information derived from the diffusion model's reconstruction process to enrich text prompts so as to associate each class with a structure-consistent mask. Specifically, we introduce Contrastive Soft Clustering (CSC) to align derived masks with the image's structure information, softly selecting anchors for each class and calculating weighted distances to push inner-class pixels closer while separating inter-class pixels, thereby ensuring mask distinction and internal consistency. By incorporating sample-specific context, InvSeg learns context-rich text prompts in embedding space and achieves accurate semantic alignment across modalities. Experiments show that InvSeg achieves state-of-the-art performance on the PASCAL VOC and Context datasets. Project page: https://jylin8100.github.io/InvSegProject/.
☆ A Simple Approach to Unifying Diffusion-based Conditional Generation
Recent progress in image generation has sparked research into controlling these models through condition signals, with various methods addressing specific challenges in conditional generation. Instead of proposing another specialized technique, we introduce a simple, unified framework to handle diverse conditional generation tasks involving a specific image-condition correlation. By learning a joint distribution over a correlated image pair (e.g. image and depth) with a diffusion model, our approach enables versatile capabilities via different inference-time sampling schemes, including controllable image generation (e.g. depth to image), estimation (e.g. image to depth), signal guidance, joint generation (image & depth), and coarse control. Previous attempts at unification often introduce significant complexity through multi-stage training, architectural modification, or increased parameter counts. In contrast, our simple formulation requires a single, computationally efficient training stage, maintains the standard model input, and adds minimal learned parameters (15% of the base model). Moreover, our model supports additional capabilities like non-spatially aligned and coarse conditioning. Extensive results show that our single model can produce comparable results with specialized methods and better results than prior unified methods. We also demonstrate that multiple models can be effectively combined for multi-signal conditional generation.
comment: Project page: https://lixirui142.github.io/unicon-diffusion/
☆ CTA-Net: A CNN-Transformer Aggregation Network for Improving Multi-Scale Feature Extraction
Convolutional neural networks (CNNs) and vision transformers (ViTs) have become essential in computer vision for local and global feature extraction. However, aggregating these architectures in existing methods often results in inefficiencies. To address this, the CNN-Transformer Aggregation Network (CTA-Net) was developed. CTA-Net combines CNNs and ViTs, with transformers capturing long-range dependencies and CNNs extracting localized features. This integration enables efficient processing of detailed local and broader contextual information. CTA-Net introduces the Light Weight Multi-Scale Feature Fusion Multi-Head Self-Attention (LMF-MHSA) module for effective multi-scale feature integration with reduced parameters. Additionally, the Reverse Reconstruction CNN-Variants (RRCV) module enhances the embedding of CNNs within the transformer architecture. Extensive experiments on small-scale datasets with fewer than 100,000 samples show that CTA-Net achieves superior performance (TOP-1 Acc 86.76\%), fewer parameters (20.32M), and greater efficiency (FLOPs 2.83B), making it a highly efficient and lightweight solution for visual tasks on small-scale datasets (fewer than 100,000).
comment: 9 pages, 3 figures
☆ GS^3: Efficient Relighting with Triple Gaussian Splatting SIGGRAPH
We present a spatial and angular Gaussian based representation and a triple splatting process, for real-time, high-quality novel lighting-and-view synthesis from multi-view point-lit input images. To describe complex appearance, we employ a Lambertian plus a mixture of angular Gaussians as an effective reflectance function for each spatial Gaussian. To generate self-shadow, we splat all spatial Gaussians towards the light source to obtain shadow values, which are further refined by a small multi-layer perceptron. To compensate for other effects like global illumination, another network is trained to compute and add a per-spatial-Gaussian RGB tuple. The effectiveness of our representation is demonstrated on 30 samples with a wide variation in geometry (from solid to fluffy) and appearance (from translucent to anisotropic), as well as using different forms of input data, including rendered images of synthetic/reconstructed objects, photographs captured with a handheld camera and a flash, or from a professional lightstage. We achieve a training time of 40-70 minutes and a rendering speed of 90 fps on a single commodity GPU. Our results compare favorably with state-of-the-art techniques in terms of quality/performance. Our code and data are publicly available at https://GSrelight.github.io/.
comment: Accepted to SIGGRAPH Asia 2024. Project page: https://gsrelight.github.io/
☆ VidCompress: Memory-Enhanced Temporal Compression for Video Understanding in Large Language Models
Video-based multimodal large language models (Video-LLMs) possess significant potential for video understanding tasks. However, most Video-LLMs treat videos as a sequential set of individual frames, which results in insufficient temporal-spatial interaction that hinders fine-grained comprehension and difficulty in processing longer videos due to limited visual token capacity. To address these challenges, we propose VidCompress, a novel Video-LLM featuring memory-enhanced temporal compression. VidCompress employs a dual-compressor approach: a memory-enhanced compressor captures both short-term and long-term temporal relationships in videos and compresses the visual tokens using a multiscale transformer with a memory-cache mechanism, while a text-perceived compressor generates condensed visual tokens by utilizing Q-Former and integrating temporal contexts into query embeddings with cross attention. Experiments on several VideoQA datasets and comprehensive benchmarks demonstrate that VidCompress efficiently models complex temporal-spatial relations and significantly outperforms existing Video-LLMs.
comment: 9 pages, 4 figures
☆ MoChat: Joints-Grouped Spatio-Temporal Grounding LLM for Multi-Turn Motion Comprehension and Description
Despite continuous advancements in deep learning for understanding human motion, existing models often struggle to accurately identify action timing and specific body parts, typically supporting only single-round interaction. Such limitations in capturing fine-grained motion details reduce their effectiveness in motion understanding tasks. In this paper, we propose MoChat, a multimodal large language model capable of spatio-temporal grounding of human motion and understanding multi-turn dialogue context. To achieve these capabilities, we group the spatial information of each skeleton frame based on human anatomical structure and then apply them with Joints-Grouped Skeleton Encoder, whose outputs are combined with LLM embeddings to create spatio-aware and temporal-aware embeddings separately. Additionally, we develop a pipeline for extracting timestamps from skeleton sequences based on textual annotations, and construct multi-turn dialogues for spatially grounding. Finally, various task instructions are generated for jointly training. Experimental results demonstrate that MoChat achieves state-of-the-art performance across multiple metrics in motion understanding tasks, making it as the first model capable of fine-grained spatio-temporal grounding of human motion.
☆ MCGS: Multiview Consistency Enhancement for Sparse-View 3D Gaussian Radiance Fields
Radiance fields represented by 3D Gaussians excel at synthesizing novel views, offering both high training efficiency and fast rendering. However, with sparse input views, the lack of multi-view consistency constraints results in poorly initialized point clouds and unreliable heuristics for optimization and densification, leading to suboptimal performance. Existing methods often incorporate depth priors from dense estimation networks but overlook the inherent multi-view consistency in input images. Additionally, they rely on multi-view stereo (MVS)-based initialization, which limits the efficiency of scene representation. To overcome these challenges, we propose a view synthesis framework based on 3D Gaussian Splatting, named MCGS, enabling photorealistic scene reconstruction from sparse input views. The key innovations of MCGS in enhancing multi-view consistency are as follows: i) We introduce an initialization method by leveraging a sparse matcher combined with a random filling strategy, yielding a compact yet sufficient set of initial points. This approach enhances the initial geometry prior, promoting efficient scene representation. ii) We develop a multi-view consistency-guided progressive pruning strategy to refine the Gaussian field by strengthening consistency and eliminating low-contribution Gaussians. These modular, plug-and-play strategies enhance robustness to sparse input views, accelerate rendering, and reduce memory consumption, making MCGS a practical and efficient framework for 3D Gaussian Splatting.
☆ Augmentation-Driven Metric for Balancing Preservation and Modification in Text-Guided Image Editing
The development of vision-language and generative models has significantly advanced text-guided image editing, which seeks \textit{preservation} of core elements in the source image while implementing \textit{modifications} based on the target text. However, in the absence of evaluation metrics specifically tailored for text-guided image editing, existing metrics are limited in balancing the consideration of preservation and modification. Especially, our analysis reveals that CLIPScore, the most commonly used metric, tends to favor modification and ignore core attributes to be preserved, resulting in inaccurate evaluations. To address this problem, we propose \texttt{AugCLIP}, \black{which balances preservation and modification by estimating the representation of an ideal edited image that aligns with the target text with minimum alteration on the source image. We augment detailed textual descriptions on the source image and the target text using a multi-modal large language model, to model a hyperplane that separates CLIP space into source or target. The representation of the ideal edited image is an orthogonal projection of the source image into the hyperplane, which encapsulates the relative importance of each attribute considering the interdependent relationships.} Our extensive experiments on five benchmark datasets, encompassing a diverse range of editing scenarios, demonstrate that \texttt{AugCLIP} aligns remarkably well with human evaluation standards compared to existing metrics. The code for evaluation will be open-sourced to contribute to the community.
comment: Under review
☆ DRACO: A Denoising-Reconstruction Autoencoder for Cryo-EM
Foundation models in computer vision have demonstrated exceptional performance in zero-shot and few-shot tasks by extracting multi-purpose features from large-scale datasets through self-supervised pre-training methods. However, these models often overlook the severe corruption in cryogenic electron microscopy (cryo-EM) images by high-level noises. We introduce DRACO, a Denoising-Reconstruction Autoencoder for CryO-EM, inspired by the Noise2Noise (N2N) approach. By processing cryo-EM movies into odd and even images and treating them as independent noisy observations, we apply a denoising-reconstruction hybrid training scheme. We mask both images to create denoising and reconstruction tasks. For DRACO's pre-training, the quality of the dataset is essential, we hence build a high-quality, diverse dataset from an uncurated public database, including over 270,000 movies or micrographs. After pre-training, DRACO naturally serves as a generalizable cryo-EM image denoiser and a foundation model for various cryo-EM downstream tasks. DRACO demonstrates the best performance in denoising, micrograph curation, and particle picking tasks compared to state-of-the-art baselines. We will release the code, pre-trained models, and the curated dataset to stimulate further research.
☆ Visual-Geometric Collaborative Guidance for Affordance Learning
Perceiving potential ``action possibilities'' (\ie, affordance) regions of images and learning interactive functionalities of objects from human demonstration is a challenging task due to the diversity of human-object interactions. Prevailing affordance learning algorithms often adopt the label assignment paradigm and presume that there is a unique relationship between functional region and affordance label, yielding poor performance when adapting to unseen environments with large appearance variations. In this paper, we propose to leverage interactive affinity for affordance learning, \ie extracting interactive affinity from human-object interaction and transferring it to non-interactive objects. Interactive affinity, which represents the contacts between different parts of the human body and local regions of the target object, can provide inherent cues of interconnectivity between humans and objects, thereby reducing the ambiguity of the perceived action possibilities. To this end, we propose a visual-geometric collaborative guided affordance learning network that incorporates visual and geometric cues to excavate interactive affinity from human-object interactions jointly. Besides, a contact-driven affordance learning (CAL) dataset is constructed by collecting and labeling over 55,047 images. Experimental results demonstrate that our method outperforms the representative models regarding objective metrics and visual quality. Project: \href{https://github.com/lhc1224/VCR-Net}{github.com/lhc1224/VCR-Net}.
☆ SeaDATE: Remedy Dual-Attention Transformer with Semantic Alignment via Contrast Learning for Multimodal Object Detection
Multimodal object detection leverages diverse modal information to enhance the accuracy and robustness of detectors. By learning long-term dependencies, Transformer can effectively integrate multimodal features in the feature extraction stage, which greatly improves the performance of multimodal object detection. However, current methods merely stack Transformer-guided fusion techniques without exploring their capability to extract features at various depth layers of network, thus limiting the improvements in detection performance. In this paper, we introduce an accurate and efficient object detection method named SeaDATE. Initially, we propose a novel dual attention Feature Fusion (DTF) module that, under Transformer's guidance, integrates local and global information through a dual attention mechanism, strengthening the fusion of modal features from orthogonal perspectives using spatial and channel tokens. Meanwhile, our theoretical analysis and empirical validation demonstrate that the Transformer-guided fusion method, treating images as sequences of pixels for fusion, performs better on shallow features' detail information compared to deep semantic information. To address this, we designed a contrastive learning (CL) module aimed at learning features of multimodal samples, remedying the shortcomings of Transformer-guided fusion in extracting deep semantic features, and effectively utilizing cross-modal information. Extensive experiments and ablation studies on the FLIR, LLVIP, and M3FD datasets have proven our method to be effective, achieving state-of-the-art detection performance.
☆ SHAKTI: A 2.5 Billion Parameter Small Language Model Optimized for Edge AI and Low-Resource Environments
We introduce Shakti, a 2.5 billion parameter language model specifically optimized for resource-constrained environments such as edge devices, including smartphones, wearables, and IoT systems. Shakti combines high-performance NLP with optimized efficiency and precision, making it ideal for real-time AI applications where computational resources and memory are limited. With support for vernacular languages and domain-specific tasks, Shakti excels in industries such as healthcare, finance, and customer service. Benchmark evaluations demonstrate that Shakti performs competitively against larger models while maintaining low latency and on-device efficiency, positioning it as a leading solution for edge AI.
comment: Paper in pdf format is 11 pages and contains 4 tables
☆ Diffusion-Based Offline RL for Improved Decision-Making in Augmented ARC Task
Effective long-term strategies enable AI systems to navigate complex environments by making sequential decisions over extended horizons. Similarly, reinforcement learning (RL) agents optimize decisions across sequences to maximize rewards, even without immediate feedback. To verify that Latent Diffusion-Constrained Q-learning (LDCQ), a prominent diffusion-based offline RL method, demonstrates strong reasoning abilities in multi-step decision-making, we aimed to evaluate its performance on the Abstraction and Reasoning Corpus (ARC). However, applying offline RL methodologies to enhance strategic reasoning in AI for solving tasks in ARC is challenging due to the lack of sufficient experience data in the ARC training set. To address this limitation, we introduce an augmented offline RL dataset for ARC, called Synthesized Offline Learning Data for Abstraction and Reasoning (SOLAR), along with the SOLAR-Generator, which generates diverse trajectory data based on predefined rules. SOLAR enables the application of offline RL methods by offering sufficient experience data. We synthesized SOLAR for a simple task and used it to train an agent with the LDCQ method. Our experiments demonstrate the effectiveness of the offline RL approach on a simple ARC task, showing the agent's ability to make multi-step sequential decisions and correctly identify answer states. These results highlight the potential of the offline RL approach to enhance AI's strategic reasoning capabilities.
comment: Preprint, Under review. Comments welcome
☆ CONSULT: Contrastive Self-Supervised Learning for Few-shot Tumor Detection
Artificial intelligence aids in brain tumor detection via MRI scans, enhancing the accuracy and reducing the workload of medical professionals. However, in scenarios with extremely limited medical images, traditional deep learning approaches tend to fail due to the absence of anomalous images. Anomaly detection also suffers from ineffective feature extraction due to vague training process. Our work introduces a novel two-stage anomaly detection algorithm called CONSULT (CONtrastive Self-sUpervised Learning for few-shot Tumor detection). The first stage of CONSULT fine-tunes a pre-trained feature extractor specifically for MRI brain images, using a synthetic data generation pipeline to create tumor-like data. This process overcomes the lack of anomaly samples and enables the integration of attention mechanisms to focus on anomalous image segments. The first stage is to overcome the shortcomings of current anomaly detection in extracting features in high-variation data by incorporating Context-Aware Contrastive Learning and Self-supervised Feature Adversarial Learning. The second stage of CONSULT uses PatchCore for conventional feature extraction via the fine-tuned weights from the first stage. To summarize, we propose a self-supervised training scheme for anomaly detection, enhancing model performance and data reliability. Furthermore, our proposed contrastive loss, Tritanh Loss, stabilizes learning by offering a unique solution all while enhancing gradient flow. Finally, CONSULT achieves superior performance in few-shot brain tumor detection, demonstrating significant improvements over PatchCore by 9.4%, 12.9%, 10.2%, and 6.0% for 2, 4, 6, and 8 shots, respectively, while training exclusively on healthy images.
comment: 14 pages, 4 figures
☆ Have the VLMs Lost Confidence? A Study of Sycophancy in VLMs
In the study of LLMs, sycophancy represents a prevalent hallucination that poses significant challenges to these models. Specifically, LLMs often fail to adhere to original correct responses, instead blindly agreeing with users' opinions, even when those opinions are incorrect or malicious. However, research on sycophancy in visual language models (VLMs) has been scarce. In this work, we extend the exploration of sycophancy from LLMs to VLMs, introducing the MM-SY benchmark to evaluate this phenomenon. We present evaluation results from multiple representative models, addressing the gap in sycophancy research for VLMs. To mitigate sycophancy, we propose a synthetic dataset for training and employ methods based on prompts, supervised fine-tuning, and DPO. Our experiments demonstrate that these methods effectively alleviate sycophancy in VLMs. Additionally, we probe VLMs to assess the semantic impact of sycophancy and analyze the attention distribution of visual tokens. Our findings indicate that the ability to prevent sycophancy is predominantly observed in higher layers of the model. The lack of attention to image knowledge in these higher layers may contribute to sycophancy, and enhancing image attention at high layers proves beneficial in mitigating this issue.
☆ Open World Object Detection: A Survey
Exploring new knowledge is a fundamental human ability that can be mirrored in the development of deep neural networks, especially in the field of object detection. Open world object detection (OWOD) is an emerging area of research that adapts this principle to explore new knowledge. It focuses on recognizing and learning from objects absent from initial training sets, thereby incrementally expanding its knowledge base when new class labels are introduced. This survey paper offers a thorough review of the OWOD domain, covering essential aspects, including problem definitions, benchmark datasets, source codes, evaluation metrics, and a comparative study of existing methods. Additionally, we investigate related areas like open set recognition (OSR) and incremental learning (IL), underlining their relevance to OWOD. Finally, the paper concludes by addressing the limitations and challenges faced by current OWOD algorithms and proposes directions for future research. To our knowledge, this is the first comprehensive survey of the emerging OWOD field with over one hundred references, marking a significant step forward for object detection technology. A comprehensive source code and benchmarks are archived and concluded at https://github.com/ArminLee/OWOD Review.
☆ Scalable Indoor Novel-View Synthesis using Drone-Captured 360 Imagery with 3D Gaussian Splatting ECCV 2024
Scene reconstruction and novel-view synthesis for large, complex, multi-story, indoor scenes is a challenging and time-consuming task. Prior methods have utilized drones for data capture and radiance fields for scene reconstruction, both of which present certain challenges. First, in order to capture diverse viewpoints with the drone's front-facing camera, some approaches fly the drone in an unstable zig-zag fashion, which hinders drone-piloting and generates motion blur in the captured data. Secondly, most radiance field methods do not easily scale to arbitrarily large number of images. This paper proposes an efficient and scalable pipeline for indoor novel-view synthesis from drone-captured 360 videos using 3D Gaussian Splatting. 360 cameras capture a wide set of viewpoints, allowing for comprehensive scene capture under a simple straightforward drone trajectory. To scale our method to large scenes, we devise a divide-and-conquer strategy to automatically split the scene into smaller blocks that can be reconstructed individually and in parallel. We also propose a coarse-to-fine alignment strategy to seamlessly match these blocks together to compose the entire scene. Our experiments demonstrate marked improvement in both reconstruction quality, i.e. PSNR and SSIM, and computation time compared to prior approaches.
comment: Accepted to ECCV 2024 S3DSGR Workshop
☆ Contrastive learning of cell state dynamics in response to perturbations
We introduce DynaCLR, a self-supervised framework for modeling cell dynamics via contrastive learning of representations of time-lapse datasets. Live cell imaging of cells and organelles is widely used to analyze cellular responses to perturbations. Human annotation of dynamic cell states captured by time-lapse perturbation datasets is laborious and prone to bias. DynaCLR integrates single-cell tracking with time-aware contrastive learning to map images of cells at neighboring time points to neighboring embeddings. Mapping the morphological dynamics of cells to a temporally regularized embedding space makes the annotation, classification, clustering, or interpretation of the cell states more quantitative and efficient. We illustrate the features and applications of DynaCLR with the following experiments: analyzing the kinetics of viral infection in human cells, detecting transient changes in cell morphology due to cell division, and mapping the dynamics of organelles due to viral infection. Models trained with DynaCLR consistently achieve $>95\%$ accuracy for infection state classification, enable the detection of transient cell states and reliably embed unseen experiments. DynaCLR provides a flexible framework for comparative analysis of cell state dynamics due to perturbations, such as infection, gene knockouts, and drugs. We provide PyTorch-based implementations of the model training and inference pipeline (https://github.com/mehta-lab/viscy) and a user interface (https://github.com/czbiohub-sf/napari-iohub) for the visualization and annotation of trajectories of cells in the real space and the embedding space.
comment: 20 pages, 6 figures, 3 appendix figures, 4 videos (ancillary files)
☆ Rethinking the Role of Infrastructure in Collaborative Perception ECCV 2024
Collaborative Perception (CP) is a process in which an ego agent receives and fuses sensor information from surrounding vehicles and infrastructure to enhance its perception capability. To evaluate the need for infrastructure equipped with sensors, extensive and quantitative analysis of the role of infrastructure data in CP is crucial, yet remains underexplored. To address this gap, we first quantitatively assess the importance of infrastructure data in existing vehicle-centric CP, where the ego agent is a vehicle. Furthermore, we compare vehicle-centric CP with infra-centric CP, where the ego agent is now the infrastructure, to evaluate the effectiveness of each approach. Our results demonstrate that incorporating infrastructure data improves 3D detection accuracy by up to 10.87%, and infra-centric CP shows enhanced noise robustness and increases accuracy by up to 42.53% compared with vehicle-centric CP.
comment: Accepted by ECCV 2024 Workshop MAAS, 14 pages
☆ CLIP-DFGS: A Hard Sample Mining Method for CLIP in Generalizable Person Re-Identification
Recent advancements in pre-trained vision-language models like CLIP have shown promise in person re-identification (ReID) applications. However, their performance in generalizable person re-identification tasks remains suboptimal. The large-scale and diverse image-text pairs used in CLIP's pre-training may lead to a lack or insufficiency of certain fine-grained features. In light of these challenges, we propose a hard sample mining method called DFGS (Depth-First Graph Sampler), based on depth-first search, designed to offer sufficiently challenging samples to enhance CLIP's ability to extract fine-grained features. DFGS can be applied to both the image encoder and the text encoder in CLIP. By leveraging the powerful cross-modal learning capabilities of CLIP, we aim to apply our DFGS method to extract challenging samples and form mini-batches with high discriminative difficulty, providing the image model with more efficient and challenging samples that are difficult to distinguish, thereby enhancing the model's ability to differentiate between individuals. Our results demonstrate significant improvements over other methods, confirming the effectiveness of DFGS in providing challenging samples that enhance CLIP's performance in generalizable person re-identification.
comment: Accepted by ACM TOMM
☆ Automatically Generating Visual Hallucination Test Cases for Multimodal Large Language Models
Visual hallucination (VH) occurs when a multimodal large language model (MLLM) generates responses with incorrect visual details for prompts. Existing methods for generating VH test cases primarily rely on human annotations, typically in the form of triples: (image, question, answer). In this paper, we introduce VHExpansion, the first automated method for expanding VH test cases for MLLMs. Given an initial VH test case, VHExpansion automatically expands it by perturbing the question and answer through negation as well as modifying the image using both common and adversarial perturbations. Additionally, we propose a new evaluation metric, symmetric accuracy, which measures the proportion of correctly answered VH test-case pairs. Each pair consists of a test case and its negated counterpart. Our theoretical analysis shows that symmetric accuracy is an unbiased evaluation metric that remains unaffected by the imbalance of VH testing cases with varying answers when an MLLM is randomly guessing the answers, whereas traditional accuracy is prone to such imbalance. We apply VHExpansion to expand three VH datasets annotated manually and use these expanded datasets to benchmark seven MLLMs. Our evaluation shows that VHExpansion effectively identifies more VH test cases. Moreover, symmetric accuracy, being unbiased, leads to different conclusions about the vulnerability of MLLMs to VH compared to traditional accuracy metric. Finally, we show that fine-tuning MLLMs on the expanded VH dataset generated by VHExpansion mitigates VH more effectively than fine-tuning on the original, manually annotated dataset. Our code is available at: https://github.com/lycheeefish/VHExpansion.
☆ Learning Diffusion Model from Noisy Measurement using Principled Expectation-Maximization Method
Diffusion models have demonstrated exceptional ability in modeling complex image distributions, making them versatile plug-and-play priors for solving imaging inverse problems. However, their reliance on large-scale clean datasets for training limits their applicability in scenarios where acquiring clean data is costly or impractical. Recent approaches have attempted to learn diffusion models directly from corrupted measurements, but these methods either lack theoretical convergence guarantees or are restricted to specific types of data corruption. In this paper, we propose a principled expectation-maximization (EM) framework that iteratively learns diffusion models from noisy data with arbitrary corruption types. Our framework employs a plug-and-play Monte Carlo method to accurately estimate clean images from noisy measurements, followed by training the diffusion model using the reconstructed images. This process alternates between estimation and training until convergence. We evaluate the performance of our method across various imaging tasks, including inpainting, denoising, and deblurring. Experimental results demonstrate that our approach enables the learning of high-fidelity diffusion priors from noisy data, significantly enhancing reconstruction quality in imaging inverse problems.
☆ Ctrl-U: Robust Conditional Image Generation via Uncertainty-aware Reward Modeling
In this paper, we focus on the task of conditional image generation, where an image is synthesized according to user instructions. The critical challenge underpinning this task is ensuring both the fidelity of the generated images and their semantic alignment with the provided conditions. To tackle this issue, previous studies have employed supervised perceptual losses derived from pre-trained models, i.e., reward models, to enforce alignment between the condition and the generated result. However, we observe one inherent shortcoming: considering the diversity of synthesized images, the reward model usually provides inaccurate feedback when encountering newly generated data, which can undermine the training process. To address this limitation, we propose an uncertainty-aware reward modeling, called Ctrl-U, including uncertainty estimation and uncertainty-aware regularization, designed to reduce the adverse effects of imprecise feedback from the reward model. Given the inherent cognitive uncertainty within reward models, even images generated under identical conditions often result in a relatively large discrepancy in reward loss. Inspired by the observation, we explicitly leverage such prediction variance as an uncertainty indicator. Based on the uncertainty estimation, we regularize the model training by adaptively rectifying the reward. In particular, rewards with lower uncertainty receive higher loss weights, while those with higher uncertainty are given reduced weights to allow for larger variability. The proposed uncertainty regularization facilitates reward fine-tuning through consistency construction. Extensive experiments validate the effectiveness of our methodology in improving the controllability and generation quality, as well as its scalability across diverse conditional scenarios. Code will soon be available at https://grenoble-zhang.github.io/Ctrl-U-Page/.
comment: Preprint. Work in progress
☆ Representation Similarity: A Better Guidance of DNN Layer Sharing for Edge Computing without Training
Edge computing has emerged as an alternative to reduce transmission and processing delay and preserve privacy of the video streams. However, the ever-increasing complexity of Deep Neural Networks (DNNs) used in video-based applications (e.g. object detection) exerts pressure on memory-constrained edge devices. Model merging is proposed to reduce the DNNs' memory footprint by keeping only one copy of merged layers' weights in memory. In existing model merging techniques, (i) only architecturally identical layers can be shared; (ii) requires computationally expensive retraining in the cloud; (iii) assumes the availability of ground truth for retraining. The re-evaluation of a merged model's performance, however, requires a validation dataset with ground truth, typically runs at the cloud. Common metrics to guide the selection of shared layers include the size or computational cost of shared layers or representation size. We propose a new model merging scheme by sharing representations (i.e., outputs of layers) at the edge, guided by representation similarity S. We show that S is extremely highly correlated with merged model's accuracy with Pearson Correlation Coefficient |r| > 0.94 than other metrics, demonstrating that representation similarity can serve as a strong validation accuracy indicator without ground truth. We present our preliminary results of the newly proposed model merging scheme with identified challenges, demonstrating a promising research future direction.
comment: 3 pages, 4 figures, ACM MobiCom '24, November 18-22, 2024, Washington D.C., DC, USA
☆ TEOcc: Radar-camera Multi-modal Occupancy Prediction via Temporal Enhancement ECAI2024
As a novel 3D scene representation, semantic occupancy has gained much attention in autonomous driving. However, existing occupancy prediction methods mainly focus on designing better occupancy representations, such as tri-perspective view or neural radiance fields, while ignoring the advantages of using long-temporal information. In this paper, we propose a radar-camera multi-modal temporal enhanced occupancy prediction network, dubbed TEOcc. Our method is inspired by the success of utilizing temporal information in 3D object detection. Specifically, we introduce a temporal enhancement branch to learn temporal occupancy prediction. In this branch, we randomly discard the t-k input frame of the multi-view camera and predict its 3D occupancy by long-term and short-term temporal decoders separately with the information from other adjacent frames and multi-modal inputs. Besides, to reduce computational costs and incorporate multi-modal inputs, we specially designed 3D convolutional layers for long-term and short-term temporal decoders. Furthermore, since the lightweight occupancy prediction head is a dense classification head, we propose to use a shared occupancy prediction head for the temporal enhancement and main branches. It is worth noting that the temporal enhancement branch is only performed during training and is discarded during inference. Experiment results demonstrate that TEOcc achieves state-of-the-art occupancy prediction on nuScenes benchmarks. In addition, the proposed temporal enhancement branch is a plug-and-play module that can be easily integrated into existing occupancy prediction methods to improve the performance of occupancy prediction. The code and models will be released at https://github.com/VDIGPKU/TEOcc.
comment: Accepted by ECAI2024
☆ A CLIP-Powered Framework for Robust and Generalizable Data Selection
Large-scale datasets have been pivotal to the advancements of deep learning models in recent years, but training on such large datasets invariably incurs substantial storage and computational overhead. Meanwhile, real-world datasets often contain redundant and noisy data, imposing a negative impact on training efficiency and model performance. Data selection has shown promise in identifying the most representative samples from the entire dataset, which aims to minimize the performance gap with reduced training costs. Existing works typically rely on single-modality information to assign importance scores for individual samples, which may lead to inaccurate assessments, especially when dealing with noisy or corrupted samples. To address this limitation, we propose a novel CLIP-powered data selection framework that leverages multimodal information for more robust and generalizable sample selection. Specifically, our framework consists of three key modules-dataset adaptation, sample scoring, and selection optimization-that together harness extensive pre-trained multimodal knowledge to comprehensively assess sample influence and optimize the selection results through multi-objective optimization. Extensive experiments demonstrate that our approach consistently outperforms existing state-of-the-art baselines on various benchmark datasets. Notably, our method effectively removes noisy or damaged samples from the dataset, enabling it to achieve even higher performance with less data. This indicates that it is not only a way to accelerate training but can also improve overall data quality.
comment: 10 pages
☆ CVCP-Fusion: On Implicit Depth Estimation for 3D Bounding Box Prediction
Combining LiDAR and Camera-view data has become a common approach for 3D Object Detection. However, previous approaches combine the two input streams at a point-level, throwing away semantic information derived from camera features. In this paper we propose Cross-View Center Point-Fusion, a state-of-the-art model to perform 3D object detection by combining camera and LiDAR-derived features in the BEV space to preserve semantic density from the camera stream while incorporating spacial data from the LiDAR stream. Our architecture utilizes aspects from previously established algorithms, Cross-View Transformers and CenterPoint, and runs their backbones in parallel, allowing efficient computation for real-time processing and application. In this paper we find that while an implicitly calculated depth-estimate may be sufficiently accurate in a 2D map-view representation, explicitly calculated geometric and spacial information is needed for precise bounding box prediction in the 3D world-view space.
comment: 7 pages, 5 figures
☆ DreamSteerer: Enhancing Source Image Conditioned Editability using Personalized Diffusion Models NeurIPS 2024
Recent text-to-image personalization methods have shown great promise in teaching a diffusion model user-specified concepts given a few images for reusing the acquired concepts in a novel context. With massive efforts being dedicated to personalized generation, a promising extension is personalized editing, namely to edit an image using personalized concepts, which can provide a more precise guidance signal than traditional textual guidance. To address this, a straightforward solution is to incorporate a personalized diffusion model with a text-driven editing framework. However, such a solution often shows unsatisfactory editability on the source image. To address this, we propose DreamSteerer, a plug-in method for augmenting existing T2I personalization methods. Specifically, we enhance the source image conditioned editability of a personalized diffusion model via a novel Editability Driven Score Distillation (EDSD) objective. Moreover, we identify a mode trapping issue with EDSD, and propose a mode shifting regularization with spatial feature guided sampling to avoid such an issue. We further employ two key modifications to the Delta Denoising Score framework that enable high-fidelity local editing with personalized concepts. Extensive experiments validate that DreamSteerer can significantly improve the editability of several T2I personalization baselines while being computationally efficient.
comment: Published as a conference paper at NeurIPS 2024
☆ Adversarially Guided Stateful Defense Against Backdoor Attacks in Federated Deep Learning ACSA
Recent works have shown that Federated Learning (FL) is vulnerable to backdoor attacks. Existing defenses cluster submitted updates from clients and select the best cluster for aggregation. However, they often rely on unrealistic assumptions regarding client submissions and sampled clients population while choosing the best cluster. We show that in realistic FL settings, state-of-the-art (SOTA) defenses struggle to perform well against backdoor attacks in FL. To address this, we highlight that backdoored submissions are adversarially biased and overconfident compared to clean submissions. We, therefore, propose an Adversarially Guided Stateful Defense (AGSD) against backdoor attacks on Deep Neural Networks (DNNs) in FL scenarios. AGSD employs adversarial perturbations to a small held-out dataset to compute a novel metric, called the trust index, that guides the cluster selection without relying on any unrealistic assumptions regarding client submissions. Moreover, AGSD maintains a trust state history of each client that adaptively penalizes backdoored clients and rewards clean clients. In realistic FL settings, where SOTA defenses mostly fail to resist attacks, AGSD mostly outperforms all SOTA defenses with minimal drop in clean accuracy (5% in the worst-case compared to best accuracy) even when (a) given a very small held-out dataset -- typically AGSD assumes 50 samples (<= 0.1% of the training data) and (b) no heldout dataset is available, and out-of-distribution data is used instead. For reproducibility, our code will be openly available at: https://github.com/hassanalikhatim/AGSD.
comment: 16 pages, Accepted at ACSAC 2024
☆ Tree of Attributes Prompt Learning for Vision-Language Models
Prompt learning has proven effective in adapting vision language models for downstream tasks. However, existing methods usually append learnable prompt tokens solely with the category names to obtain textual features, which fails to fully leverage the rich context indicated in the category name. To address this issue, we propose the Tree of Attributes Prompt learning (TAP), which first instructs LLMs to generate a tree of attributes with a "concept - attribute - description" structure for each category, and then learn the hierarchy with vision and text prompt tokens. Unlike existing methods that merely augment category names with a set of unstructured descriptions, our approach essentially distills structured knowledge graphs associated with class names from LLMs. Furthermore, our approach introduces text and vision prompts designed to explicitly learn the corresponding visual attributes, effectively serving as domain experts. Additionally, the general and diverse descriptions generated based on the class names may be wrong or absent in the specific given images. To address this misalignment, we further introduce a vision-conditional pooling module to extract instance-specific text features. Extensive experimental results demonstrate that our approach outperforms state-of-the-art methods on the zero-shot base-to-novel generalization, cross-dataset transfer, as well as few-shot classification across 11 diverse datasets.
☆ Mini-Omni2: Towards Open-source GPT-4o Model with Vision, Speech and Duplex
GPT4o, an all-encompassing model, represents a milestone in the development of multi-modal large models. It can understand visual, auditory, and textual modalities, directly output audio, and support flexible duplex interaction. However, its technical framework is not open-sourced. Models from the open-source community often achieve some functionalities of GPT4o, such as visual understanding and voice dialogue. Nevertheless, training a unified model that incorporates all modalities is challenging due to the complexities of multi-modal data, intricate model architectures, and training processes. In this paper, we introduce Mini-Omni2, a visual-audio assistant capable of providing real-time, end-to-end voice responses to user video and voice queries, while also incorporating auditory capabilities. By integrating pretrained visual and auditory encoders, Mini-Omni2 maintains strong performance in individual modalities. We propose a three-stage training process to align modalities, allowing the language model to handle multi-modal inputs and outputs after training on a limited dataset. For interaction, we introduce a semantic-based interruption mechanism, enabling more flexible dialogues with users. All modeling approaches and data construction methods will be open-sourced. To the best of our knowledge, Mini-Omni2 is one of the models closest to GPT4o in functionality, and we hope it can offer valuable insights for subsequent research.
comment: 13 pages, 6 figures
☆ Multiview Scene Graph NeurIPS 2024
A proper scene representation is central to the pursuit of spatial intelligence where agents can robustly reconstruct and efficiently understand 3D scenes. A scene representation is either metric, such as landmark maps in 3D reconstruction, 3D bounding boxes in object detection, or voxel grids in occupancy prediction, or topological, such as pose graphs with loop closures in SLAM or visibility graphs in SfM. In this work, we propose to build Multiview Scene Graphs (MSG) from unposed images, representing a scene topologically with interconnected place and object nodes. The task of building MSG is challenging for existing representation learning methods since it needs to jointly address both visual place recognition, object detection, and object association from images with limited fields of view and potentially large viewpoint changes. To evaluate any method tackling this task, we developed an MSG dataset and annotation based on a public 3D dataset. We also propose an evaluation metric based on the intersection-over-union score of MSG edges. Moreover, we develop a novel baseline method built on mainstream pretrained vision models, combining visual place recognition and object association into one Transformer decoder architecture. Experiments demonstrate our method has superior performance compared to existing relevant baselines.
comment: To be published in NeurIPS 2024. Website at https://ai4ce.github.io/MSG/
☆ Synthesizing Proton-Density Fat Fraction and $R_2^*$ from 2-point Dixon MRI with Generative Machine Learning
Magnetic Resonance Imaging (MRI) is the gold standard for measuring fat and iron content non-invasively in the body via measures known as Proton Density Fat Fraction (PDFF) and $R_2^*$, respectively. However, conventional PDFF and $R_2^*$ quantification methods operate on MR images voxel-wise and require at least three measurements to estimate three quantities: water, fat, and $R_2^*$. Alternatively, the two-point Dixon MRI protocol is widely used and fast because it acquires only two measurements; however, these cannot be used to estimate three quantities voxel-wise. Leveraging the fact that neighboring voxels have similar values, we propose using a generative machine learning approach to learn PDFF and $R_2^*$ from Dixon MRI. We use paired Dixon-IDEAL data from UK Biobank in the liver and a Pix2Pix conditional GAN to demonstrate the first large-scale $R_2^*$ imputation from two-point Dixon MRIs. Using our proposed approach, we synthesize PDFF and $R_2^*$ maps that show significantly greater correlation with ground-truth than conventional voxel-wise baselines.
☆ Improving Bias in Facial Attribute Classification: A Combined Impact of KL Divergence induced Loss Function and Dual Attention
Ensuring that AI-based facial recognition systems produce fair predictions and work equally well across all demographic groups is crucial. Earlier systems often exhibited demographic bias, particularly in gender and racial classification, with lower accuracy for women and individuals with darker skin tones. To tackle this issue and promote fairness in facial recognition, researchers have introduced several bias-mitigation techniques for gender classification and related algorithms. However, many challenges remain, such as data diversity, balancing fairness with accuracy, disparity, and bias measurement. This paper presents a method using a dual attention mechanism with a pre-trained Inception-ResNet V1 model, enhanced by KL-divergence regularization and a cross-entropy loss function. This approach reduces bias while improving accuracy and computational efficiency through transfer learning. The experimental results show significant improvements in both fairness and classification accuracy, providing promising advances in addressing bias and enhancing the reliability of facial recognition systems.
comment: 15 pages, 9 figures, 5 tables
☆ Towards General Deepfake Detection with Dynamic Curriculum ICASSP 2024
Most previous deepfake detection methods bent their efforts to discriminate artifacts by end-to-end training. However, the learned networks often fail to mine the general face forgery information efficiently due to ignoring the data hardness. In this work, we propose to introduce the sample hardness into the training of deepfake detectors via the curriculum learning paradigm. Specifically, we present a novel simple yet effective strategy, named Dynamic Facial Forensic Curriculum (DFFC), which makes the model gradually focus on hard samples during the training. Firstly, we propose Dynamic Forensic Hardness (DFH) which integrates the facial quality score and instantaneous instance loss to dynamically measure sample hardness during the training. Furthermore, we present a pacing function to control the data subsets from easy to hard throughout the training process based on DFH. Comprehensive experiments show that DFFC can improve both within- and cross-dataset performance of various kinds of end-to-end deepfake detectors through a plug-and-play approach. It indicates that DFFC can help deepfake detectors learn general forgery discriminative features by effectively exploiting the information from hard samples.
comment: Received by ICASSP 2024 - 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
☆ MANet: Fine-Tuning Segment Anything Model for Multimodal Remote Sensing Semantic Segmentation
Multimodal remote sensing data, collected from a variety of sensors, provide a comprehensive and integrated perspective of the Earth's surface. By employing multimodal fusion techniques, semantic segmentation offers more detailed insights into geographic scenes compared to single-modality approaches. Building upon recent advancements in vision foundation models, particularly the Segment Anything Model (SAM), this study introduces a novel Multimodal Adapter-based Network (MANet) for multimodal remote sensing semantic segmentation. At the core of this approach is the development of a Multimodal Adapter (MMAdapter), which fine-tunes SAM's image encoder to effectively leverage the model's general knowledge for multimodal data. In addition, a pyramid-based Deep Fusion Module (DFM) is incorporated to further integrate high-level geographic features across multiple scales before decoding. This work not only introduces a novel network for multimodal fusion, but also demonstrates, for the first time, SAM's powerful generalization capabilities with Digital Surface Model (DSM) data. Experimental results on two well-established fine-resolution multimodal remote sensing datasets, ISPRS Vaihingen and ISPRS Potsdam, confirm that the proposed MANet significantly surpasses current models in the task of multimodal semantic segmentation. The source code for this work will be accessible at https://github.com/sstary/SSRS.
comment: 12 pages, 9 figures
☆ Deep unrolled primal dual network for TOF-PET list-mode image reconstruction
Time-of-flight (TOF) information provides more accurate location data for annihilation photons, thereby enhancing the quality of PET reconstruction images and reducing noise. List-mode reconstruction has a significant advantage in handling TOF information. However, current advanced TOF PET list-mode reconstruction algorithms still require improvements when dealing with low-count data. Deep learning algorithms have shown promising results in PET image reconstruction. Nevertheless, the incorporation of TOF information poses significant challenges related to the storage space required by deep learning methods, particularly for the advanced deep unrolled methods. In this study, we propose a deep unrolled primal dual network for TOF-PET list-mode reconstruction. The network is unrolled into multiple phases, with each phase comprising a dual network for list-mode domain updates and a primal network for image domain updates. We utilize CUDA for parallel acceleration and computation of the system matrix for TOF list-mode data, and we adopt a dynamic access strategy to mitigate memory consumption. Reconstructed images of different TOF resolutions and different count levels show that the proposed method outperforms the LM-OSEM, LM-EMTV, LM-SPDHG,LM-SPDHG-TV and FastPET method in both visually and quantitative analysis. These results demonstrate the potential application of deep unrolled methods for TOF-PET list-mode data and show better performance than current mainstream TOF-PET list-mode reconstruction algorithms, providing new insights for the application of deep learning methods in TOF list-mode data. The codes for this work are available at https://github.com/RickHH/LMPDnet
comment: 11 pages, 11 figures
☆ OMCAT: Omni Context Aware Transformer
Large Language Models (LLMs) have made significant strides in text generation and comprehension, with recent advancements extending into multimodal LLMs that integrate visual and audio inputs. However, these models continue to struggle with fine-grained, cross-modal temporal understanding, particularly when correlating events across audio and video streams. We address these challenges with two key contributions: a new dataset and model, called OCTAV and OMCAT respectively. OCTAV (Omni Context and Temporal Audio Video) is a novel dataset designed to capture event transitions across audio and video. Second, OMCAT (Omni Context Aware Transformer) is a powerful model that leverages RoTE (Rotary Time Embeddings), an innovative extension of RoPE, to enhance temporal grounding and computational efficiency in time-anchored tasks. Through a robust three-stage training pipeline-feature alignment, instruction tuning, and OCTAV-specific training-OMCAT excels in cross-modal temporal understanding. Our model demonstrates state-of-the-art performance on Audio-Visual Question Answering (AVQA) tasks and the OCTAV benchmark, showcasing significant gains in temporal reasoning and cross-modal alignment, as validated through comprehensive experiments and ablation studies. Our dataset and code will be made publicly available. The link to our demo page is https://om-cat.github.io.
comment: Demo page: https://om-cat.github.io
☆ SplatPose+: Real-time Image-Based Pose-Agnostic 3D Anomaly Detection
Image-based Pose-Agnostic 3D Anomaly Detection is an important task that has emerged in industrial quality control. This task seeks to find anomalies from query images of a tested object given a set of reference images of an anomaly-free object. The challenge is that the query views (a.k.a poses) are unknown and can be different from the reference views. Currently, new methods such as OmniposeAD and SplatPose have emerged to bridge the gap by synthesizing pseudo reference images at the query views for pixel-to-pixel comparison. However, none of these methods can infer in real-time, which is critical in industrial quality control for massive production. For this reason, we propose SplatPose+, which employs a hybrid representation consisting of a Structure from Motion (SfM) model for localization and a 3D Gaussian Splatting (3DGS) model for Novel View Synthesis. Although our proposed pipeline requires the computation of an additional SfM model, it offers real-time inference speeds and faster training compared to SplatPose. Quality-wise, we achieved a new SOTA on the Pose-agnostic Anomaly Detection benchmark with the Multi-Pose Anomaly Detection (MAD-SIM) dataset.
☆ WeatherDG: LLM-assisted Procedural Weather Generation for Domain-Generalized Semantic Segmentation
In this work, we propose a novel approach, namely WeatherDG, that can generate realistic, weather-diverse, and driving-screen images based on the cooperation of two foundation models, i.e, Stable Diffusion (SD) and Large Language Model (LLM). Specifically, we first fine-tune the SD with source data, aligning the content and layout of generated samples with real-world driving scenarios. Then, we propose a procedural prompt generation method based on LLM, which can enrich scenario descriptions and help SD automatically generate more diverse, detailed images. In addition, we introduce a balanced generation strategy, which encourages the SD to generate high-quality objects of tailed classes under various weather conditions, such as riders and motorcycles. This segmentation-model-agnostic method can improve the generalization ability of existing models by additionally adapting them with the generated synthetic data. Experiments on three challenging datasets show that our method can significantly improve the segmentation performance of different state-of-the-art models on target domains. Notably, in the setting of ''Cityscapes to ACDC'', our method improves the baseline HRDA by 13.9% in mIoU.
☆ nvTorchCam: An Open-source Library for Camera-Agnostic Differentiable Geometric Vision
We introduce nvTorchCam, an open-source library under the Apache 2.0 license, designed to make deep learning algorithms camera model-independent. nvTorchCam abstracts critical camera operations such as projection and unprojection, allowing developers to implement algorithms once and apply them across diverse camera models--including pinhole, fisheye, and 360 equirectangular panoramas, which are commonly used in automotive and real estate capture applications. Built on PyTorch, nvTorchCam is fully differentiable and supports GPU acceleration and batching for efficient computation. Furthermore, deep learning models trained for one camera type can be directly transferred to other camera types without requiring additional modification. In this paper, we provide an overview of nvTorchCam, its functionality, and present various code examples and diagrams to demonstrate its usage. Source code and installation instructions can be found on the nvTorchCam GitHub page at https://github.com/NVlabs/nvTorchCam.
comment: Source code and installation instructions are available at https://github.com/NVlabs/nvTorchCam
☆ V3D-SLAM: Robust RGB-D SLAM in Dynamic Environments with 3D Semantic Geometry Voting
Simultaneous localization and mapping (SLAM) in highly dynamic environments is challenging due to the correlation complexity between moving objects and the camera pose. Many methods have been proposed to deal with this problem; however, the moving properties of dynamic objects with a moving camera remain unclear. Therefore, to improve SLAM's performance, minimizing disruptive events of moving objects with a physical understanding of 3D shapes and dynamics of objects is needed. In this paper, we propose a robust method, V3D-SLAM, to remove moving objects via two lightweight re-evaluation stages, including identifying potentially moving and static objects using a spatial-reasoned Hough voting mechanism and refining static objects by detecting dynamic noise caused by intra-object motions using Chamfer distances as similarity measurements. Our experiment on the TUM RGB-D benchmark on dynamic sequences with ground-truth camera trajectories showed that our methods outperform the most recent state-of-the-art SLAM methods. Our source code is available at https://github.com/tuantdang/v3d-slam.
☆ SOE: SO(3)-Equivariant 3D MRI Encoding
Representation learning has become increasingly important, especially as powerful models have shifted towards learning latent representations before fine-tuning for downstream tasks. This approach is particularly valuable in leveraging the structural information within brain anatomy. However, a common limitation of recent models developed for MRIs is their tendency to ignore or remove geometric information, such as translation and rotation, thereby creating invariance with respect to geometric operations. We contend that incorporating knowledge about these geometric transformations into the model can significantly enhance its ability to learn more detailed anatomical information within brain structures. As a result, we propose a novel method for encoding 3D MRIs that enforces equivariance with respect to all rotations in 3D space, in other words, SO(3)-equivariance (SOE). By explicitly modeling this geometric equivariance in the representation space, we ensure that any rotational operation applied to the input image space is also reflected in the embedding representation space. This approach requires moving beyond traditional representation learning methods, as we need a representation vector space that allows for the application of the same SO(3) operation in that space. To facilitate this, we leverage the concept of vector neurons. The representation space formed by our method captures the brain's structural and anatomical information more effectively. We evaluate SOE pretrained on the structural MRIs of two public data sets with respect to the downstream task of predicting age and diagnosing Alzheimer's Disease from T1-weighted brain scans of the ADNI data set. We demonstrate that our approach not only outperforms other methods but is also robust against various degrees of rotation along different axes. The code is available at https://github.com/shizhehe/SOE-representation-learning.
☆ Learned Neural Physics Simulation for Articulated 3D Human Pose Reconstruction
We propose a novel neural network approach, LARP (Learned Articulated Rigid body Physics), to model the dynamics of articulated human motion with contact. Our goal is to develop a faster and more convenient methodological alternative to traditional physics simulators for use in computer vision tasks such as human motion reconstruction from video. To that end we introduce a training procedure and model components that support the construction of a recurrent neural architecture to accurately simulate articulated rigid body dynamics. Our neural architecture supports features typically found in traditional physics simulators, such as modeling of joint motors, variable dimensions of body parts, contact between body parts and objects, and is an order of magnitude faster than traditional systems when multiple simulations are run in parallel. To demonstrate the value of LARP we use it as a drop-in replacement for a state of the art classical non-differentiable simulator in an existing video-based reconstruction framework and show comparative or better 3D human pose reconstruction accuracy.
☆ LocoMotion: Learning Motion-Focused Video-Language Representations ACCV 2024
This paper strives for motion-focused video-language representations. Existing methods to learn video-language representations use spatial-focused data, where identifying the objects and scene is often enough to distinguish the relevant caption. We instead propose LocoMotion to learn from motion-focused captions that describe the movement and temporal progression of local object motions. We achieve this by adding synthetic motions to videos and using the parameters of these motions to generate corresponding captions. Furthermore, we propose verb-variation paraphrasing to increase the caption variety and learn the link between primitive motions and high-level verbs. With this, we are able to learn a motion-focused video-language representation. Experiments demonstrate our approach is effective for a variety of downstream tasks, particularly when limited data is available for fine-tuning. Code is available: https://hazeldoughty.github.io/Papers/LocoMotion/
comment: ACCV 2024
☆ Beyond Labels: A Self-Supervised Framework with Masked Autoencoders and Random Cropping for Breast Cancer Subtype Classification
This work contributes to breast cancer sub-type classification using histopathological images. We utilize masked autoencoders (MAEs) to learn a self-supervised embedding tailored for computer vision tasks in this domain. This embedding captures informative representations of histopathological data, facilitating feature learning without extensive labeled datasets. During pre-training, we investigate employing a random crop technique to generate a large dataset from WSIs automatically. Additionally, we assess the performance of linear probes for multi-class classification tasks of cancer sub-types using the representations learnt by the MAE. Our approach aims to achieve strong performance on downstream tasks by leveraging the complementary strengths of ViTs and autoencoders. We evaluate our model's performance on the BRACS dataset and compare it with existing benchmarks.
☆ DDIL: Improved Diffusion Distillation With Imitation Learning
Diffusion models excel at generative modeling (e.g., text-to-image) but sampling requires multiple denoising network passes, limiting practicality. Efforts such as progressive distillation or consistency distillation have shown promise by reducing the number of passes at the expense of quality of the generated samples. In this work we identify co-variate shift as one of reason for poor performance of multi-step distilled models from compounding error at inference time. To address co-variate shift, we formulate diffusion distillation within imitation learning (DDIL) framework and enhance training distribution for distilling diffusion models on both data distribution (forward diffusion) and student induced distributions (backward diffusion). Training on data distribution helps to diversify the generations by preserving marginal data distribution and training on student distribution addresses compounding error by correcting covariate shift. In addition, we adopt reflected diffusion formulation for distillation and demonstrate improved performance, stable training across different distillation methods. We show that DDIL consistency improves on baseline algorithms of progressive distillation (PD), Latent consistency models (LCM) and Distribution Matching Distillation (DMD2).
☆ Integrating Artificial Intelligence Models and Synthetic Image Data for Enhanced Asset Inspection and Defect Identification
In the past utilities relied on in-field inspections to identify asset defects. Recently, utilities have started using drone-based inspections to enhance the field-inspection process. We consider a vast repository of drone images, providing a wealth of information about asset health and potential issues. However, making the collected imagery data useful for automated defect detection requires significant manual labeling effort. We propose a novel solution that combines synthetic asset defect images with manually labeled drone images. This solution has several benefits: improves performance of defect detection, reduces the number of hours spent on manual labeling, and enables the capability to generate realistic images of rare defects where not enough real-world data is available. We employ a workflow that combines 3D modeling tools such as Maya and Unreal Engine to create photorealistic 3D models and 2D renderings of defective assets and their surroundings. These synthetic images are then integrated into our training pipeline augmenting the real data. This study implements an end-to-end Artificial Intelligence solution to detect assets and asset defects from the combined imagery repository. The unique contribution of this research lies in the application of advanced computer vision models and the generation of photorealistic 3D renderings of defective assets, aiming to transform the asset inspection process. Our asset detection model has achieved an accuracy of 92 percent, we achieved a performance lift of 67 percent when introducing approximately 2,000 synthetic images of 2k resolution. In our tests, the defect detection model achieved an accuracy of 73 percent across two batches of images. Our analysis demonstrated that synthetic data can be successfully used in place of real-world manually labeled data to train defect detection model.
☆ CtrlSynth: Controllable Image Text Synthesis for Data-Efficient Multimodal Learning
Pretraining robust vision or multimodal foundation models (e.g., CLIP) relies on large-scale datasets that may be noisy, potentially misaligned, and have long-tail distributions. Previous works have shown promising results in augmenting datasets by generating synthetic samples. However, they only support domain-specific ad hoc use cases (e.g., either image or text only, but not both), and are limited in data diversity due to a lack of fine-grained control over the synthesis process. In this paper, we design a \emph{controllable} image-text synthesis pipeline, CtrlSynth, for data-efficient and robust multimodal learning. The key idea is to decompose the visual semantics of an image into basic elements, apply user-specified control policies (e.g., remove, add, or replace operations), and recompose them to synthesize images or texts. The decompose and recompose feature in CtrlSynth allows users to control data synthesis in a fine-grained manner by defining customized control policies to manipulate the basic elements. CtrlSynth leverages the capabilities of pretrained foundation models such as large language models or diffusion models to reason and recompose basic elements such that synthetic samples are natural and composed in diverse ways. CtrlSynth is a closed-loop, training-free, and modular framework, making it easy to support different pretrained models. With extensive experiments on 31 datasets spanning different vision and vision-language tasks, we show that CtrlSynth substantially improves zero-shot classification, image-text retrieval, and compositional reasoning performance of CLIP models.
☆ Dual-frame Fluid Motion Estimation with Test-time Optimization and Zero-divergence Loss NeurIPS 2024
3D particle tracking velocimetry (PTV) is a key technique for analyzing turbulent flow, one of the most challenging computational problems of our century. At the core of 3D PTV is the dual-frame fluid motion estimation algorithm, which tracks particles across two consecutive frames. Recently, deep learning-based methods have achieved impressive accuracy in dual-frame fluid motion estimation; however, they heavily depend on large volumes of labeled data. In this paper, we introduce a new method that is completely self-supervised and notably outperforms its fully-supervised counterparts while requiring only 1% of the training samples (without labels) used by previous methods. Our method features a novel zero-divergence loss that is specific to the domain of turbulent flow. Inspired by the success of splat operation in high-dimensional filtering and random fields, we propose a splat-based implementation for this loss which is both efficient and effective. The self-supervised nature of our method naturally supports test-time optimization, leading to the development of a tailored Dynamic Velocimetry Enhancer (DVE) module. We demonstrate that strong cross-domain robustness is achieved through test-time optimization on unseen leave-one-out synthetic domains and real physical/biological domains. Code, data and models are available at https://github.com/Forrest-110/FluidMotionNet.
comment: Accepted by NeurIPS 2024
♻ ☆ TemporalBench: Benchmarking Fine-grained Temporal Understanding for Multimodal Video Models
Understanding fine-grained temporal dynamics is crucial for multimodal video comprehension and generation. Due to the lack of fine-grained temporal annotations, existing video benchmarks mostly resemble static image benchmarks and are incompetent at evaluating models for temporal understanding. In this paper, we introduce TemporalBench, a new benchmark dedicated to evaluating fine-grained temporal understanding in videos. TemporalBench consists of ~10K video question-answer pairs, derived from ~2K high-quality human annotations detailing the temporal dynamics in video clips. As a result, our benchmark provides a unique testbed for evaluating various temporal understanding and reasoning abilities such as action frequency, motion magnitude, event order, etc. Moreover, it enables evaluations on various tasks like both video question answering and captioning, both short and long video understanding, as well as different models such as multimodal video embedding models and text generation models. Results show that state-of-the-art models like GPT-4o achieve only 38.5% question answering accuracy on TemporalBench, demonstrating a significant gap (~30%) between humans and AI in temporal understanding. Furthermore, we notice a critical pitfall for multi-choice QA where LLMs can detect the subtle changes in negative captions and find a centralized description as a cue for its prediction, where we propose Multiple Binary Accuracy (MBA) to correct such bias. We hope that TemporalBench can foster research on improving models' temporal reasoning capabilities. Both dataset and evaluation code will be made available.
comment: Project Page: https://temporalbench.github.io/
♻ ☆ Advancements in Road Lane Mapping: Comparative Fine-Tuning Analysis of Deep Learning-based Semantic Segmentation Methods Using Aerial Imagery
This research addresses the need for high-definition (HD) maps for autonomous vehicles (AVs), focusing on road lane information derived from aerial imagery. While Earth observation data offers valuable resources for map creation, specialized models for road lane extraction are still underdeveloped in remote sensing. In this study, we perform an extensive comparison of twelve foundational deep learning-based semantic segmentation models for road lane marking extraction from high-definition remote sensing images, assessing their performance under transfer learning with partially labeled datasets. These models were fine-tuned on the partially labeled Waterloo Urban Scene dataset, and pre-trained on the SkyScapes dataset, simulating a likely scenario of real-life model deployment under partial labeling. We observed and assessed the fine-tuning performance and overall performance. Models showed significant performance improvements after fine-tuning, with mean IoU scores ranging from 33.56% to 76.11%, and recall ranging from 66.0% to 98.96%. Transformer-based models outperformed convolutional neural networks, emphasizing the importance of model pre-training and fine-tuning in enhancing HD map development for AV navigation.
♻ ☆ VIA: Unified Spatiotemporal Video Adaptation Framework for Global and Local Video Editing
Video editing is a cornerstone of digital media, from entertainment and education to professional communication. However, previous methods often overlook the necessity of comprehensively understanding both global and local contexts, leading to inaccurate and inconsistent edits in the spatiotemporal dimension, especially for long videos. In this paper, we introduce VIA, a unified spatiotemporal Video Adaptation framework for global and local video editing, pushing the limits of consistently editing minute-long videos. First, to ensure local consistency within individual frames, we designed test-time editing adaptation to adapt a pre-trained image editing model for improving consistency between potential editing directions and the text instruction, and adapt masked latent variables for precise local control. Furthermore, to maintain global consistency over the video sequence, we introduce spatiotemporal adaptation that recursively gather consistent attention variables in key frames and strategically applies them across the whole sequence to realize the editing effects. Extensive experiments demonstrate that, compared to baseline methods, our VIA approach produces edits that are more faithful to the source videos, more coherent in the spatiotemporal context, and more precise in local control. More importantly, we show that VIA can achieve consistent long video editing in minutes, unlocking the potential for advanced video editing tasks over long video sequences.
comment: 19 pages, 14 figures
♻ ☆ Active Label Refinement for Robust Training of Imbalanced Medical Image Classification Tasks in the Presence of High Label Noise MICCAI 2024
The robustness of supervised deep learning-based medical image classification is significantly undermined by label noise. Although several methods have been proposed to enhance classification performance in the presence of noisy labels, they face some challenges: 1) a struggle with class-imbalanced datasets, leading to the frequent overlooking of minority classes as noisy samples; 2) a singular focus on maximizing performance using noisy datasets, without incorporating experts-in-the-loop for actively cleaning the noisy labels. To mitigate these challenges, we propose a two-phase approach that combines Learning with Noisy Labels (LNL) and active learning. This approach not only improves the robustness of medical image classification in the presence of noisy labels, but also iteratively improves the quality of the dataset by relabeling the important incorrect labels, under a limited annotation budget. Furthermore, we introduce a novel Variance of Gradients approach in LNL phase, which complements the loss-based sample selection by also sampling under-represented samples. Using two imbalanced noisy medical classification datasets, we demonstrate that that our proposed technique is superior to its predecessors at handling class imbalance by not misidentifying clean samples from minority classes as mostly noisy samples.
comment: Accepted at MICCAI 2024
♻ ☆ ImageFolder: Autoregressive Image Generation with Folded Tokens
Image tokenizers are crucial for visual generative models, e.g., diffusion models (DMs) and autoregressive (AR) models, as they construct the latent representation for modeling. Increasing token length is a common approach to improve the image reconstruction quality. However, tokenizers with longer token lengths are not guaranteed to achieve better generation quality. There exists a trade-off between reconstruction and generation quality regarding token length. In this paper, we investigate the impact of token length on both image reconstruction and generation and provide a flexible solution to the tradeoff. We propose ImageFolder, a semantic tokenizer that provides spatially aligned image tokens that can be folded during autoregressive modeling to improve both generation efficiency and quality. To enhance the representative capability without increasing token length, we leverage dual-branch product quantization to capture different contexts of images. Specifically, semantic regularization is introduced in one branch to encourage compacted semantic information while another branch is designed to capture the remaining pixel-level details. Extensive experiments demonstrate the superior quality of image generation and shorter token length with ImageFolder tokenizer.
comment: Code: https://github.com/lxa9867/ImageFolder
♻ ☆ U-MedSAM: Uncertainty-aware MedSAM for Medical Image Segmentation
Medical Image Foundation Models have proven to be powerful tools for mask prediction across various datasets. However, accurately assessing the uncertainty of their predictions remains a significant challenge. To address this, we propose a new model, U-MedSAM, which integrates the MedSAM model with an uncertainty-aware loss function and the Sharpness-Aware Minimization (SharpMin) optimizer. The uncertainty-aware loss function automatically combines region-based, distribution-based, and pixel-based loss designs to enhance segmentation accuracy and robustness. SharpMin improves generalization by finding flat minima in the loss landscape, thereby reducing overfitting. Our method was evaluated in the CVPR24 MedSAM on Laptop challenge, where U-MedSAM demonstrated promising performance.
comment: arXiv admin note: text overlap with arXiv:2405.17496
♻ ☆ Augmentation-aware Self-supervised Learning with Conditioned Projector NeurIPS 2023
Self-supervised learning (SSL) is a powerful technique for learning from unlabeled data. By learning to remain invariant to applied data augmentations, methods such as SimCLR and MoCo can reach quality on par with supervised approaches. However, this invariance may be detrimental for solving downstream tasks that depend on traits affected by augmentations used during pretraining, such as color. In this paper, we propose to foster sensitivity to such characteristics in the representation space by modifying the projector network, a common component of self-supervised architectures. Specifically, we supplement the projector with information about augmentations applied to images. For the projector to take advantage of this auxiliary conditioning when solving the SSL task, the feature extractor learns to preserve the augmentation information in its representations. Our approach, coined Conditional Augmentation-aware Self-supervised Learning (CASSLE), is directly applicable to typical joint-embedding SSL methods regardless of their objective functions. Moreover, it does not require major changes in the network architecture or prior knowledge of downstream tasks. In addition to an analysis of sensitivity towards different data augmentations, we conduct a series of experiments, which show that CASSLE improves over various SSL methods, reaching state-of-the-art performance in multiple downstream tasks.
comment: A short version of this paper appeared at the NeurIPS 2023 Workshop: Self-Supervised Learning - Theory and Practice (https://sslneurips23.github.io). The full paper was published (OA) in Knowledge-Based Systems (https://www.sciencedirect.com/science/article/pii/S0950705124012061)
♻ ☆ LongHalQA: Long-Context Hallucination Evaluation for MultiModal Large Language Models
Hallucination, a phenomenon where multimodal large language models~(MLLMs) tend to generate textual responses that are plausible but unaligned with the image, has become one major hurdle in various MLLM-related applications. Several benchmarks have been created to gauge the hallucination levels of MLLMs, by either raising discriminative questions about the existence of objects or introducing LLM evaluators to score the generated text from MLLMs. However, the discriminative data largely involve simple questions that are not aligned with real-world text, while the generative data involve LLM evaluators that are computationally intensive and unstable due to their inherent randomness. We propose LongHalQA, an LLM-free hallucination benchmark that comprises 6K long and complex hallucination text. LongHalQA is featured by GPT4V-generated hallucinatory data that are well aligned with real-world scenarios, including object/image descriptions and multi-round conversations with 14/130 words and 189 words, respectively, on average. It introduces two new tasks, hallucination discrimination and hallucination completion, unifying both discriminative and generative evaluations in a single multiple-choice-question form and leading to more reliable and efficient evaluations without the need for LLM evaluators. Further, we propose an advanced pipeline that greatly facilitates the construction of future hallucination benchmarks with long and complex questions and descriptions. Extensive experiments over multiple recent MLLMs reveal various new challenges when they are handling hallucinations with long and complex textual data. Dataset and evaluation code are available at https://github.com/hanqiu-hq/LongHalQA.
♻ ☆ MimicTalk: Mimicking a personalized and expressive 3D talking face in minutes NeurIPS 2024
Talking face generation (TFG) aims to animate a target identity's face to create realistic talking videos. Personalized TFG is a variant that emphasizes the perceptual identity similarity of the synthesized result (from the perspective of appearance and talking style). While previous works typically solve this problem by learning an individual neural radiance field (NeRF) for each identity to implicitly store its static and dynamic information, we find it inefficient and non-generalized due to the per-identity-per-training framework and the limited training data. To this end, we propose MimicTalk, the first attempt that exploits the rich knowledge from a NeRF-based person-agnostic generic model for improving the efficiency and robustness of personalized TFG. To be specific, (1) we first come up with a person-agnostic 3D TFG model as the base model and propose to adapt it into a specific identity; (2) we propose a static-dynamic-hybrid adaptation pipeline to help the model learn the personalized static appearance and facial dynamic features; (3) To generate the facial motion of the personalized talking style, we propose an in-context stylized audio-to-motion model that mimics the implicit talking style provided in the reference video without information loss by an explicit style representation. The adaptation process to an unseen identity can be performed in 15 minutes, which is 47 times faster than previous person-dependent methods. Experiments show that our MimicTalk surpasses previous baselines regarding video quality, efficiency, and expressiveness. Source code and video samples are available at https://mimictalk.github.io .
comment: Accepted by NeurIPS 2024
♻ ☆ Learning Truncated Causal History Model for Video Restoration NeurIPS 2024
One key challenge to video restoration is to model the transition dynamics of video frames governed by motion. In this work, we propose TURTLE to learn the truncated causal history model for efficient and high-performing video restoration. Unlike traditional methods that process a range of contextual frames in parallel, TURTLE enhances efficiency by storing and summarizing a truncated history of the input frame latent representation into an evolving historical state. This is achieved through a sophisticated similarity-based retrieval mechanism that implicitly accounts for inter-frame motion and alignment. The causal design in TURTLE enables recurrence in inference through state-memorized historical features while allowing parallel training by sampling truncated video clips. We report new state-of-the-art results on a multitude of video restoration benchmark tasks, including video desnowing, nighttime video deraining, video raindrops and rain streak removal, video super-resolution, real-world and synthetic video deblurring, and blind video denoising while reducing the computational cost compared to existing best contextual methods on all these tasks.
comment: Accepted to NeurIPS 2024. 24 pages
♻ ☆ MedSyn: Text-guided Anatomy-aware Synthesis of High-Fidelity 3D CT Images
This paper introduces an innovative methodology for producing high-quality 3D lung CT images guided by textual information. While diffusion-based generative models are increasingly used in medical imaging, current state-of-the-art approaches are limited to low-resolution outputs and underutilize radiology reports' abundant information. The radiology reports can enhance the generation process by providing additional guidance and offering fine-grained control over the synthesis of images. Nevertheless, expanding text-guided generation to high-resolution 3D images poses significant memory and anatomical detail-preserving challenges. Addressing the memory issue, we introduce a hierarchical scheme that uses a modified UNet architecture. We start by synthesizing low-resolution images conditioned on the text, serving as a foundation for subsequent generators for complete volumetric data. To ensure the anatomical plausibility of the generated samples, we provide further guidance by generating vascular, airway, and lobular segmentation masks in conjunction with the CT images. The model demonstrates the capability to use textual input and segmentation tasks to generate synthesized images. The results of comparative assessments indicate that our approach exhibits superior performance compared to the most advanced models based on GAN and diffusion techniques, especially in accurately retaining crucial anatomical features such as fissure lines, airways, and vascular structures. This innovation introduces novel possibilities. This study focuses on two main objectives: (1) the development of a method for creating images based on textual prompts and anatomical components, and (2) the capability to generate new images conditioning on anatomical elements. The advancements in image generation can be applied to enhance numerous downstream tasks.
♻ ☆ LeOCLR: Leveraging Original Images for Contrastive Learning of Visual Representations
Contrastive instance discrimination methods outperform supervised learning in downstream tasks such as image classification and object detection. However, these methods rely heavily on data augmentation during representation learning, which can lead to suboptimal results if not implemented carefully. A common augmentation technique in contrastive learning is random cropping followed by resizing. This can degrade the quality of representation learning when the two random crops contain distinct semantic content. To tackle this issue, we introduce LeOCLR (Leveraging Original Images for Contrastive Learning of Visual Representations), a framework that employs a novel instance discrimination approach and an adapted loss function. This method prevents the loss of important semantic features caused by mapping different object parts during representation learning. Our experiments demonstrate that LeOCLR consistently improves representation learning across various datasets, outperforming baseline models. For instance, LeOCLR surpasses MoCo-v2 by 5.1% on ImageNet-1K in linear evaluation and outperforms several other methods on transfer learning and object detection tasks.
comment: 15 pages, 5 figures, 9 tables - accepted at TMLR 10/2024
♻ ☆ POPoS: Improving Efficient and Robust Facial Landmark Detection with Parallel Optimal Position Search
Achieving a balance between accuracy and efficiency is a critical challenge in facial landmark detection (FLD). This paper introduces the Parallel Optimal Position Search (POPoS), a high-precision encoding-decoding framework designed to address the fundamental limitations of traditional FLD methods. POPoS employs three key innovations: (1) Pseudo-range multilateration is utilized to correct heatmap errors, enhancing the precision of landmark localization. By integrating multiple anchor points, this approach minimizes the impact of individual heatmap inaccuracies, leading to robust overall positioning. (2) To improve the pseudo-range accuracy of selected anchor points, a new loss function, named multilateration anchor loss, is proposed. This loss function effectively enhances the accuracy of the distance map, mitigates the risk of local optima, and ensures optimal solutions. (3) A single-step parallel computation algorithm is introduced, significantly enhancing computational efficiency and reducing processing time. Comprehensive evaluations across five benchmark datasets demonstrate that POPoS consistently outperforms existing methods, particularly excelling in low-resolution scenarios with minimal computational overhead. These features establish POPoS as a highly efficient and accurate tool for FLD, with broad applicability in real-world scenarios. The code is available at https://github.com/teslatasy/PoPoS
comment: The experimental setup and results require further modifications and improvements to ensure the accuracy and quality of the paper. Therefore, we are requesting to withdraw the submission
♻ ☆ Mitigating Backdoor Attack by Injecting Proactive Defensive Backdoor NeurIPS 2024
Data-poisoning backdoor attacks are serious security threats to machine learning models, where an adversary can manipulate the training dataset to inject backdoors into models. In this paper, we focus on in-training backdoor defense, aiming to train a clean model even when the dataset may be potentially poisoned. Unlike most existing methods that primarily detect and remove/unlearn suspicious samples to mitigate malicious backdoor attacks, we propose a novel defense approach called PDB (Proactive Defensive Backdoor). Specifically, PDB leverages the home-field advantage of defenders by proactively injecting a defensive backdoor into the model during training. Taking advantage of controlling the training process, the defensive backdoor is designed to suppress the malicious backdoor effectively while remaining secret to attackers. In addition, we introduce a reversible mapping to determine the defensive target label. During inference, PDB embeds a defensive trigger in the inputs and reverses the model's prediction, suppressing malicious backdoor and ensuring the model's utility on the original task. Experimental results across various datasets and models demonstrate that our approach achieves state-of-the-art defense performance against a wide range of backdoor attacks. The code is available at https://github.com/shawkui/Proactive_Defensive_Backdoor.
comment: Accepted by NeurIPS 2024. 32 pages, 7 figures, 28 tables
♻ ☆ Estimating the distribution of numerosity and non-numerical visual magnitudes in natural scenes using computer vision
Humans share with many animal species the ability to perceive and approximately represent the number of objects in visual scenes. This ability improves throughout childhood, suggesting that learning and development play a key role in shaping our number sense. This hypothesis is further supported by computational investigations based on deep learning, which have shown that numerosity perception can spontaneously emerge in neural networks that learn the statistical structure of images with a varying number of items. However, neural network models are usually trained using synthetic datasets that might not faithfully reflect the statistical structure of natural environments, and there is also growing interest in using more ecological visual stimuli to investigate numerosity perception in humans. In this work, we exploit recent advances in computer vision algorithms to design and implement an original pipeline that can be used to estimate the distribution of numerosity and non-numerical magnitudes in large-scale datasets containing thousands of real images depicting objects in daily life situations. We show that in natural visual scenes the frequency of appearance of different numerosities follows a power law distribution. Moreover, we show that the correlational structure for numerosity and continuous magnitudes is stable across datasets and scene types (homogeneous vs. heterogeneous object sets). We suggest that considering such "ecological" pattern of covariance is important to understand the influence of non-numerical visual cues on numerosity judgements.
♻ ☆ Evaluating Image Hallucination in Text-to-Image Generation with Question-Answering
Despite the impressive success of text-to-image (TTI) generation models, existing studies overlook the issue of whether these models accurately convey factual information. In this paper, we focus on the problem of image hallucination, where images created by generation models fail to faithfully depict factual content. To address this, we introduce I-HallA (Image Hallucination evaluation with Question Answering), a novel automated evaluation metric that measures the factuality of generated images through visual question answering (VQA). We also introduce I-HallA v1.0, a curated benchmark dataset for this purpose. As part of this process, we develop a pipeline that generates high-quality question-answer pairs using multiple GPT-4 Omni-based agents, with human judgments to ensure accuracy. Our evaluation protocols measure image hallucination by testing if images from existing text-to-image models can correctly respond to these questions. The I-HallA v1.0 dataset comprises 1.2K diverse image-text pairs across nine categories with 1,000 rigorously curated questions covering various compositional challenges. We evaluate five text-to-image models using I-HallA and reveal that these state-of-the-art models often fail to accurately convey factual information. Moreover, we validate the reliability of our metric by demonstrating a strong Spearman correlation (rho=0.95) with human judgments. We believe our benchmark dataset and metric can serve as a foundation for developing factually accurate text-to-image generation models.
comment: 20 pages
♻ ☆ Teaching AI Agents to Search with Reflective-MCTS and Exploratory Learning
Autonomous agents have demonstrated significant potential in automating complex multistep decision-making tasks. However, even state-of-the-art vision-language models (VLMs), such as GPT-4o, still fall short of human-level performance, particularly in intricate web environments and long-horizon planning tasks. To address these limitations, we present Reflective Monte Carlo Tree Search (R-MCTS) and Exploratory Learning to build o1-like models for agentic applications. We first introduce R-MCTS, a novel test-time algorithm designed to enhance the ability of AI agents to explore decision space on the fly. R-MCTS extends traditional MCTS by 1) incorporating contrastive reflection, allowing agents to learn from past interactions and dynamically improve their search efficiency; and 2) using multi-agent debate to provide reliable state evaluation. Next, we introduce Exploratory Learning, a novel learning strategy to teach agents to search at inference time without relying on any external search algorithms. On the challenging VisualWebArena benchmark, our GPT-4o-based R-MCTS agent achieves a 6% to 30% relative improvement across various tasks compared to the previous state-of-the-art. Additionally, we show that the experience gained from test-time search can be effectively transferred back to GPT-4o via fine-tuning. After Exploratory Learning, GPT-4o 1) demonstrates the ability to explore the environment, evaluate a state, and backtrack to viable ones when it detects that the current state cannot lead to success, and 2) matches 87% of R-MCTS's performance while using significantly less compute. Notably, our work demonstrates the compute scaling properties in both training - data collection with R-MCTS - and testing time. These results suggest a promising research direction to enhance VLMs' reasoning and planning capabilities for agentic applications via test-time search and self-learning.
♻ ☆ Shelf-Supervised Cross-Modal Pre-Training for 3D Object Detection
State-of-the-art 3D object detectors are often trained on massive labeled datasets. However, annotating 3D bounding boxes remains prohibitively expensive and time-consuming, particularly for LiDAR. Instead, recent works demonstrate that self-supervised pre-training with unlabeled data can improve detection accuracy with limited labels. Contemporary methods adapt best-practices for self-supervised learning from the image domain to point clouds (such as contrastive learning). However, publicly available 3D datasets are considerably smaller and less diverse than those used for image-based self-supervised learning, limiting their effectiveness. We do note, however, that such 3D data is naturally collected in a multimodal fashion, often paired with images. Rather than pre-training with only self-supervised objectives, we argue that it is better to bootstrap point cloud representations using image-based foundation models trained on internet-scale data. Specifically, we propose a shelf-supervised approach (e.g. supervised with off-the-shelf image foundation models) for generating zero-shot 3D bounding boxes from paired RGB and LiDAR data. Pre-training 3D detectors with such pseudo-labels yields significantly better semi-supervised detection accuracy than prior self-supervised pretext tasks. Importantly, we show that image-based shelf-supervision is helpful for training LiDAR-only, RGB-only and multi-modal (RGB + LiDAR) detectors. We demonstrate the effectiveness of our approach on nuScenes and WOD, significantly improving over prior work in limited data settings. Our code is available at https://github.com/meharkhurana03/cm3d
comment: The first two authors contributed equally. This work has been accepted to the Conference on Robot Learning (CoRL) 2024
♻ ☆ LLM-wrapper: Black-Box Semantic-Aware Adaptation of Vision-Language Models for Referring Expression Comprehension ECCV 2024
Vision Language Models (VLMs) have demonstrated remarkable capabilities in various open-vocabulary tasks, yet their zero-shot performance lags behind task-specific finetuned models, particularly in complex tasks like Referring Expression Comprehension (REC). Fine-tuning usually requires 'white-box' access to the model's architecture and weights, which is not always feasible due to proprietary or privacy concerns. In this work, we propose LLM-wrapper, a method for 'black-box' adaptation of VLMs for the REC task using Large Language Models (LLMs). LLM-wrapper capitalizes on the reasoning abilities of LLMs, improved with a light fine-tuning, to select the most relevant bounding box matching the referring expression, from candidates generated by a zero-shot black-box VLM. Our approach offers several advantages: it enables the adaptation of closed-source models without needing access to their internal workings, it is versatile as it works with any VLM, it transfers to new VLMs, and it allows for the adaptation of an ensemble of VLMs. We evaluate LLM-wrapper on multiple datasets using different VLMs and LLMs, demonstrating significant performance improvements and highlighting the versatility of our method. While LLM-wrapper is not meant to directly compete with standard white-box fine-tuning, it offers a practical and effective alternative for black-box VLM adaptation. The code will be open-sourced.
comment: v1 at EVAL-FoMo workshop, ECCV 2024
♻ ☆ GSE: Group-wise Sparse and Explainable Adversarial Attacks
Sparse adversarial attacks fool deep neural networks (DNNs) through minimal pixel perturbations, often regularized by the $\ell_0$ norm. Recent efforts have replaced this norm with a structural sparsity regularizer, such as the nuclear group norm, to craft group-wise sparse adversarial attacks. The resulting perturbations are thus explainable and hold significant practical relevance, shedding light on an even greater vulnerability of DNNs. However, crafting such attacks poses an optimization challenge, as it involves computing norms for groups of pixels within a non-convex objective. We address this by presenting a two-phase algorithm that generates group-wise sparse attacks within semantically meaningful areas of an image. Initially, we optimize a quasinorm adversarial loss using the $1/2-$quasinorm proximal operator tailored for non-convex programming. Subsequently, the algorithm transitions to a projected Nesterov's accelerated gradient descent with $2-$norm regularization applied to perturbation magnitudes. Rigorous evaluations on CIFAR-10 and ImageNet datasets demonstrate a remarkable increase in group-wise sparsity, e.g., $50.9\%$ on CIFAR-10 and $38.4\%$ on ImageNet (average case, targeted attack). This performance improvement is accompanied by significantly faster computation times, improved explainability, and a $100\%$ attack success rate.
♻ ☆ Bridging The Gap between Low-rank and Orthogonal Adaptation via Householder Reflection Adaptation
While following different technical routes, both low-rank and orthogonal adaptation techniques can efficiently adapt large-scale pre-training models in specific tasks or domains based on a small piece of trainable parameters. In this study, we bridge the gap between these two techniques, proposing a simple but effective adaptation method based on Householder reflections. Given a pre-trained model, our method fine-tunes its layers by multiplying each frozen weight matrix with an orthogonal matrix constructed by a chain of learnable Householder reflections (HRs). This HR-based orthogonal fine-tuning is equivalent to an adaptive low-rank adaptation. Moreover, we show that the orthogonality of the reflection planes corresponding to the HRs impacts the model capacity and regularity. The analysis motivates us to regularize the orthogonality of the HRs, leading to different implementations of the proposed Householder reflection adaptation (HRA) method. Compared with state-of-the-art methods, HRA achieves superior performance with fewer learnable parameters when adapting large language models and conditional image generators. The code of the experiments is available at \url{https://github.com/DaShenZi721/HRA}, and the method has been merged into the \href{https://github.com/huggingface/peft}{PEFT} package.
♻ ☆ Capture Artifacts via Progressive Disentangling and Purifying Blended Identities for Deepfake Detection
The Deepfake technology has raised serious concerns regarding privacy breaches and trust issues. To tackle these challenges, Deepfake detection technology has emerged. Current methods over-rely on the global feature space, which contains redundant information independent of the artifacts. As a result, existing Deepfake detection techniques suffer performance degradation when encountering unknown datasets. To reduce information redundancy, the current methods use disentanglement techniques to roughly separate the fake faces into artifacts and content information. However, these methods lack a solid disentanglement foundation and cannot guarantee the reliability of their disentangling process. To address these issues, a Deepfake detection method based on progressive disentangling and purifying blended identities is innovatively proposed in this paper. Based on the artifact generation mechanism, the coarse- and fine-grained strategies are combined to ensure the reliability of the disentanglement method. Our method aims to more accurately capture and separate artifact features in fake faces. Specifically, we first perform the coarse-grained disentangling on fake faces to obtain a pair of blended identities that require no additional annotation to distinguish between source face and target face. Then, the artifact features from each identity are separated to achieve fine-grained disentanglement. To obtain pure identity information and artifacts, an Identity-Artifact Correlation Compression module (IACC) is designed based on the information bottleneck theory, effectively reducing the potential correlation between identity information and artifacts. Additionally, an Identity-Artifact Separation Contrast Loss is designed to enhance the independence of artifact features post-disentangling. Finally, the classifier only focuses on pure artifact features to achieve a generalized Deepfake detector.
comment: TCSVT(Under Review)
♻ ☆ Finding the Subjective Truth: Collecting 2 Million Votes for Comprehensive Gen-AI Model Evaluation
Efficiently evaluating the performance of text-to-image models is difficult as it inherently requires subjective judgment and human preference, making it hard to compare different models and quantify the state of the art. Leveraging Rapidata's technology, we present an efficient annotation framework that sources human feedback from a diverse, global pool of annotators. Our study collected over 2 million annotations across 4,512 images, evaluating four prominent models (DALL-E 3, Flux.1, MidJourney, and Stable Diffusion) on style preference, coherence, and text-to-image alignment. We demonstrate that our approach makes it feasible to comprehensively rank image generation models based on a vast pool of annotators and show that the diverse annotator demographics reflect the world population, significantly decreasing the risk of biases.
♻ ☆ MotionLCM: Real-time Controllable Motion Generation via Latent Consistency Model ECCV 2024
This work introduces MotionLCM, extending controllable motion generation to a real-time level. Existing methods for spatial-temporal control in text-conditioned motion generation suffer from significant runtime inefficiency. To address this issue, we first propose the motion latent consistency model (MotionLCM) for motion generation, building upon the latent diffusion model. By adopting one-step (or few-step) inference, we further improve the runtime efficiency of the motion latent diffusion model for motion generation. To ensure effective controllability, we incorporate a motion ControlNet within the latent space of MotionLCM and enable explicit control signals (e.g., initial poses) in the vanilla motion space to control the generation process directly, similar to controlling other latent-free diffusion models for motion generation. By employing these techniques, our approach can generate human motions with text and control signals in real-time. Experimental results demonstrate the remarkable generation and controlling capabilities of MotionLCM while maintaining real-time runtime efficiency.
comment: MotionLCM project version 1.0 (ECCV 2024)
♻ ☆ A3D: Does Diffusion Dream about 3D Alignment?
We tackle the problem of text-driven 3D generation from a geometry alignment perspective. Given a set of text prompts, we aim to generate a collection of objects with semantically corresponding parts aligned across them. Recent methods based on Score Distillation have succeeded in distilling the knowledge from 2D diffusion models to high-quality representations of the 3D objects. These methods handle multiple text queries separately, and therefore the resulting objects have a high variability in object pose and structure. However, in some applications, such as 3D asset design, it may be desirable to obtain a set of objects aligned with each other. In order to achieve the alignment of the corresponding parts of the generated objects, we propose to embed these objects into a common latent space and optimize the continuous transitions between these objects. We enforce two kinds of properties of these transitions: smoothness of the transition and plausibility of the intermediate objects along the transition. We demonstrate that both of these properties are essential for good alignment. We provide several practical scenarios that benefit from alignment between the objects, including 3D editing and object hybridization, and experimentally demonstrate the effectiveness of our method. \href{https://voyleg.github.io/a3d/}{voyleg.github.io/a3d}
♻ ☆ ReinDiffuse: Crafting Physically Plausible Motions with Reinforced Diffusion Model WACV 2025
Generating human motion from textual descriptions is a challenging task. Existing methods either struggle with physical credibility or are limited by the complexities of physics simulations. In this paper, we present \emph{ReinDiffuse} that combines reinforcement learning with motion diffusion model to generate physically credible human motions that align with textual descriptions. Our method adapts Motion Diffusion Model to output a parameterized distribution of actions, making them compatible with reinforcement learning paradigms. We employ reinforcement learning with the objective of maximizing physically plausible rewards to optimize motion generation for physical fidelity. Our approach outperforms existing state-of-the-art models on two major datasets, HumanML3D and KIT-ML, achieving significant improvements in physical plausibility and motion quality. Project: https://reindiffuse.github.io/
comment: Accepted by WACV 2025 in Round 1
♻ ☆ ActionAtlas: A VideoQA Benchmark for Domain-specialized Action Recognition
Our world is full of varied actions and moves across specialized domains that we, as humans, strive to identify and understand. Within any single domain, actions can often appear quite similar, making it challenging for deep models to distinguish them accurately. To evaluate the effectiveness of multimodal foundation models in helping us recognize such actions, we present ActionAtlas v1.0, a multiple-choice video question answering benchmark featuring short videos across various sports. Each video in the dataset is paired with a question and four or five choices. The question pinpoints specific individuals, asking which choice "best" describes their action within a certain temporal context. Overall, the dataset includes 934 videos showcasing 580 unique actions across 56 sports, with a total of 1896 actions within choices. Unlike most existing video question answering benchmarks that only cover simplistic actions, often identifiable from a single frame, ActionAtlas focuses on intricate movements and rigorously tests the model's capability to discern subtle differences between moves that look similar within each domain. We evaluate open and proprietary foundation models on this benchmark, finding that the best model, GPT-4o, achieves a maximum accuracy of 45.52%. Meanwhile, Non-expert crowd workers, provided with action description for each choice, achieve 61.64% accuracy, where random chance is approximately 21%. Our findings with state-of-the-art models indicate that having a high frame sampling rate is important for accurately recognizing actions in ActionAtlas, a feature that some leading proprietary video models, such as Gemini, do not include in their default configuration.
♻ ☆ Agent Planning with World Knowledge Model NeurIPS 2024
Recent endeavors towards directly using large language models (LLMs) as agent models to execute interactive planning tasks have shown commendable results. Despite their achievements, however, they still struggle with brainless trial-and-error in global planning and generating hallucinatory actions in local planning due to their poor understanding of the ``real'' physical world. Imitating humans' mental world knowledge model which provides global prior knowledge before the task and maintains local dynamic knowledge during the task, in this paper, we introduce parametric World Knowledge Model (WKM) to facilitate agent planning. Concretely, we steer the agent model to self-synthesize knowledge from both expert and sampled trajectories. Then we develop WKM, providing prior task knowledge to guide the global planning and dynamic state knowledge to assist the local planning. Experimental results on three complex real-world simulated datasets with three state-of-the-art open-source LLMs, Mistral-7B, Gemma-7B, and Llama-3-8B, demonstrate that our method can achieve superior performance compared to various strong baselines. Besides, we analyze to illustrate that our WKM can effectively alleviate the blind trial-and-error and hallucinatory action issues, providing strong support for the agent's understanding of the world. Other interesting findings include: 1) our instance-level task knowledge can generalize better to unseen tasks, 2) weak WKM can guide strong agent model planning, and 3) unified WKM training has promising potential for further development. The code is available at https://github.com/zjunlp/WKM.
comment: NeurIPS 2024
♻ ☆ Enhancing Gait Video Analysis in Neurodegenerative Diseases by Knowledge Augmentation in Vision Language Model MICCAI 2024
We present a knowledge augmentation strategy for assessing the diagnostic groups and gait impairment from monocular gait videos. Based on a large-scale pre-trained Vision Language Model (VLM), our model learns and improves visual, textual, and numerical representations of patient gait videos, through a collective learning across three distinct modalities: gait videos, class-specific descriptions, and numerical gait parameters. Our specific contributions are two-fold: First, we adopt a knowledge-aware prompt tuning strategy to utilize the class-specific medical description in guiding the text prompt learning. Second, we integrate the paired gait parameters in the form of numerical texts to enhance the numeracy of the textual representation. Results demonstrate that our model not only significantly outperforms state-of-the-art methods in video-based classification tasks but also adeptly decodes the learned class-specific text features into natural language descriptions using the vocabulary of quantitative gait parameters. The code and the model will be made available at our project page: https://lisqzqng.github.io/GaitAnalysisVLM/.
comment: MICCAI 2024
♻ ☆ ODES: Domain Adaptation with Expert Guidance for Online Medical Image Segmentation
Unsupervised domain adaptive segmentation typically relies on self-training using pseudo labels predicted by a pre-trained network on an unlabeled target dataset. However, the noisy nature of such pseudo-labels presents a major bottleneck in adapting a network to the distribution shift between source and target datasets. This challenge is exaggerated when the network encounters an incoming data stream in online fashion, where the network is constrained to adapt to incoming streams of target domain data in exactly one round of forward and backward passes. In this scenario, relying solely on inaccurate pseudo-labels can lead to low-quality segmentation, which is detrimental to medical image analysis where accuracy and precision are of utmost priority. We hypothesize that a small amount of pixel-level annotation obtained from an expert can address this problem, thereby enhancing the performance of domain adaptation of online streaming data, even in the absence of dedicated training data. We call our method ODES: Domain Adaptation with Expert Guidance for Online Medical Image Segmentation that adapts to each incoming data batch in an online setup, incorporating feedback from an expert through active learning. Through active learning, the most informative pixels in each image can be selected for expert annotation. However, the acquisition of pixel-level annotations across all images in a batch often leads to redundant information while increasing temporal overhead in online learning. To reduce the annotation acquisition time and make the adaptation process more online-friendly, we further propose a novel image-pruning strategy that selects the most useful subset of images from the current batch for active learning. Our proposed approach outperforms existing online adaptation approaches and produces competitive results compared to offline domain adaptive active learning methods.
♻ ☆ Tora: Trajectory-oriented Diffusion Transformer for Video Generation
Recent advancements in Diffusion Transformer (DiT) have demonstrated remarkable proficiency in producing high-quality video content. Nonetheless, the potential of transformer-based diffusion models for effectively generating videos with controllable motion remains an area of limited exploration. This paper introduces Tora, the first trajectory-oriented DiT framework that concurrently integrates textual, visual, and trajectory conditions, thereby enabling scalable video generation with effective motion guidance. Specifically, Tora consists of a Trajectory Extractor(TE), a Spatial-Temporal DiT, and a Motion-guidance Fuser(MGF). The TE encodes arbitrary trajectories into hierarchical spacetime motion patches with a 3D video compression network. The MGF integrates the motion patches into the DiT blocks to generate consistent videos that accurately follow designated trajectories. Our design aligns seamlessly with DiT's scalability, allowing precise control of video content's dynamics with diverse durations, aspect ratios, and resolutions. Extensive experiments demonstrate Tora's excellence in achieving high motion fidelity, while also meticulously simulating the intricate movement of the physical world. Code is available at: https://github.com/alibaba/Tora.
♻ ☆ Fine-Grained Spatiotemporal Motion Alignment for Contrastive Video Representation Learning ACM MM 2023
As the most essential property in a video, motion information is critical to a robust and generalized video representation. To inject motion dynamics, recent works have adopted frame difference as the source of motion information in video contrastive learning, considering the trade-off between quality and cost. However, existing works align motion features at the instance level, which suffers from spatial and temporal weak alignment across modalities. In this paper, we present a \textbf{Fi}ne-grained \textbf{M}otion \textbf{A}lignment (FIMA) framework, capable of introducing well-aligned and significant motion information. Specifically, we first develop a dense contrastive learning framework in the spatiotemporal domain to generate pixel-level motion supervision. Then, we design a motion decoder and a foreground sampling strategy to eliminate the weak alignments in terms of time and space. Moreover, a frame-level motion contrastive loss is presented to improve the temporal diversity of the motion features. Extensive experiments demonstrate that the representations learned by FIMA possess great motion-awareness capabilities and achieve state-of-the-art or competitive results on downstream tasks across UCF101, HMDB51, and Diving48 datasets. Code is available at \url{https://github.com/ZMHH-H/FIMA}.
comment: ACM MM 2023 Camera Ready
♻ ☆ MicroSSIM: Improved Structural Similarity for Comparing Microscopy Data ECCV 24
Microscopy is routinely used to image biological structures of interest. Due to imaging constraints, acquired images, also called as micrographs, are typically low-SNR and contain noise. Over the last few years, regression-based tasks like unsupervised denoising and splitting have found utility in working with such noisy micrographs. For evaluation, Structural Similarity (SSIM) is one of the most popular measures used in the field. For such tasks, the best evaluation would be when both low-SNR noisy images and corresponding high-SNR clean images are obtained directly from a microscope. However, due to the following three peculiar properties of the microscopy data, we observe that SSIM is not well suited to this data regime: (a) high-SNR micrographs have higher intensity pixels as compared to low-SNR micrographs, (b) high-SNR micrographs have higher intensity pixels than found in natural images, images for which SSIM was developed, and (c) a digitally configurable offset is added by the detector present inside the microscope which affects the SSIM value. We show that SSIM components behave unexpectedly when the prediction generated from low-SNR input is compared with the corresponding high-SNR data. We explain this by introducing the phenomenon of saturation, where SSIM components become less sensitive to (dis)similarity between the images. We propose an intuitive way to quantify this, which explains the observed SSIM behavior. We introduce MicroSSIM, a variant of SSIM, which overcomes the above-discussed issues. We justify the soundness and utility of MicroSSIM using theoretical and empirical arguments and show the utility of MicroSSIM on two tasks: unsupervised denoising and joint image splitting with unsupervised denoising. Since our formulation can be applied to a broad family of SSIM-based measures, we also introduce MicroMS3IM, a microscopy-specific variation of MS-SSIM.
comment: Accepted at BIC workshop, ECCV 24
♻ ☆ A Taxonomy of Miscompressions: Preparing Image Forensics for Neural Compression
Neural compression has the potential to revolutionize lossy image compression. Based on generative models, recent schemes achieve unprecedented compression rates at high perceptual quality but compromise semantic fidelity. Details of decompressed images may appear optically flawless but semantically different from the originals, making compression errors difficult or impossible to detect. We explore the problem space and propose a provisional taxonomy of miscompressions. It defines three types of 'what happens' and has a binary 'high impact' flag indicating miscompressions that alter symbols. We discuss how the taxonomy can facilitate risk communication and research into mitigations.
comment: 6 pages, 6 figures
♻ ☆ ChangeMinds: Multi-task Framework for Detecting and Describing Changes in Remote Sensing
Recent advancements in Remote Sensing (RS) for Change Detection (CD) and Change Captioning (CC) have seen substantial success by adopting deep learning techniques. Despite these advances, existing methods often handle CD and CC tasks independently, leading to inefficiencies from the absence of synergistic processing. In this paper, we present ChangeMinds, a novel unified multi-task framework that concurrently optimizes CD and CC processes within a single, end-to-end model. We propose the change-aware long short-term memory module (ChangeLSTM) to effectively capture complex spatiotemporal dynamics from extracted bi-temporal deep features, enabling the generation of universal change-aware representations that effectively serve both CC and CD tasks. Furthermore, we introduce a multi-task predictor with a cross-attention mechanism that enhances the interaction between image and text features, promoting efficient simultaneous learning and processing for both tasks. Extensive evaluations on the LEVIR-MCI dataset, alongside other standard benchmarks, show that ChangeMinds surpasses existing methods in multi-task learning settings and markedly improves performance in individual CD and CC tasks. Codes and pre-trained models will be available online.
♻ ☆ HAIR: Hypernetworks-based All-in-One Image Restoration
Image restoration aims to recover a high-quality clean image from its degraded version. Recent progress in image restoration has demonstrated the effectiveness of All-in-One image restoration models in addressing various unknown degradations simultaneously. However, these existing methods typically utilize the same parameters to tackle images with different types of degradation, forcing the model to balance the performance between different tasks and limiting its performance on each task. To alleviate this issue, we propose HAIR, a Hypernetworks-based All-in-One Image Restoration plug-and-play method that generates parameters based on the input image and thus makes the model to adapt to specific degradation dynamically. Specifically, HAIR consists of two main components, i.e., Classifier and Hyper Selecting Net (HSN). The Classifier is a simple image classification network used to generate a Global Information Vector (GIV) that contains the degradation information of the input image, and the HSN is a simple fully-connected neural network that receives the GIV and outputs parameters for the corresponding modules. Extensive experiments demonstrate that HAIR can significantly improve the performance of existing image restoration models in a plug-and-play manner, both in single-task and All-in-One settings. Notably, our proposed model Res-HAIR, which integrates HAIR into the well-known Restormer, can obtain superior or comparable performance compared with current state-of-the-art methods. Moreover, we theoretically demonstrate that to achieve a given small enough error, our proposed HAIR requires fewer parameters in contrast to mainstream embedding-based All-in-One methods. The code is available at https://github.com/toummHus/HAIR.
♻ ☆ AIC MLLM: Autonomous Interactive Correction MLLM for Robust Robotic Manipulation
The ability to reflect on and correct failures is crucial for robotic systems to interact stably with real-life objects.Observing the generalization and reasoning capabilities of Multimodal Large Language Models (MLLMs), previous approaches have aimed to utilize these models to enhance robotic systems accordingly.However, these methods typically focus on high-level planning corrections using an additional MLLM, with limited utilization of failed samples to correct low-level contact poses which is particularly prone to occur during articulated object manipulation.To address this gap, we propose an Autonomous Interactive Correction (AIC) MLLM, which makes use of previous low-level interaction experiences to correct SE(3) pose predictions for articulated object. Specifically, AIC MLLM is initially fine-tuned to acquire both pose prediction and feedback prompt comprehension abilities.We design two types of prompt instructions for interactions with objects: 1) visual masks to highlight unmovable parts for position correction, and 2) textual descriptions to indicate potential directions for rotation correction. During inference, a Feedback Information Extraction module is introduced to recognize the failure cause, allowing AIC MLLM to adaptively correct the pose prediction using the corresponding prompts.To further enhance manipulation stability, we devise a Test Time Adaptation strategy that enables AIC MLLM to better adapt to the current scene configuration.Finally, extensive experiments are conducted in both simulated and real-world environments to evaluate the proposed method. The results demonstrate that our AIC MLLM can efficiently correct failure samples by leveraging interaction experience prompts.Our project website is https://sites.google.com/view/aic-mllm.
♻ ☆ RAUCA: A Novel Physical Adversarial Attack on Vehicle Detectors via Robust and Accurate Camouflage Generation ICML
Adversarial camouflage is a widely used physical attack against vehicle detectors for its superiority in multi-view attack performance. One promising approach involves using differentiable neural renderers to facilitate adversarial camouflage optimization through gradient back-propagation. However, existing methods often struggle to capture environmental characteristics during the rendering process or produce adversarial textures that can precisely map to the target vehicle, resulting in suboptimal attack performance. Moreover, these approaches neglect diverse weather conditions, reducing the efficacy of generated camouflage across varying weather scenarios. To tackle these challenges, we propose a robust and accurate camouflage generation method, namely RAUCA. The core of RAUCA is a novel neural rendering component, Neural Renderer Plus (NRP), which can accurately project vehicle textures and render images with environmental characteristics such as lighting and weather. In addition, we integrate a multi-weather dataset for camouflage generation, leveraging the NRP to enhance the attack robustness. Experimental results on six popular object detectors show that RAUCA consistently outperforms existing methods in both simulation and real-world settings.
comment: 12 pages. In Proceedings of the Forty-first International Conference on Machine Learning (ICML), Vienna, Austria, July 21-27, 2024
♻ ☆ 4-LEGS: 4D Language Embedded Gaussian Splatting
The emergence of neural representations has revolutionized our means for digitally viewing a wide range of 3D scenes, enabling the synthesis of photorealistic images rendered from novel views. Recently, several techniques have been proposed for connecting these low-level representations with the high-level semantics understanding embodied within the scene. These methods elevate the rich semantic understanding from 2D imagery to 3D representations, distilling high-dimensional spatial features onto 3D space. In our work, we are interested in connecting language with a dynamic modeling of the world. We show how to lift spatio-temporal features to a 4D representation based on 3D Gaussian Splatting. This enables an interactive interface where the user can spatiotemporally localize events in the video from text prompts. We demonstrate our system on public 3D video datasets of people and animals performing various actions.
comment: Project webpage: https://tau-vailab.github.io/4-LEGS/
♻ ☆ AnimateLCM: Accelerating the Animation of Personalized Diffusion Models and Adapters with Decoupled Consistency Learning SIGGRAPH
This paper introduces an effective method for computation-efficient personalized style video generation without requiring access to any personalized video data. It reduces the necessary generation time of similarly sized video diffusion models from 25 seconds to around 1 second while maintaining the same level of performance. The method's effectiveness lies in its dual-level decoupling learning approach: 1) separating the learning of video style from video generation acceleration, which allows for personalized style video generation without any personalized style video data, and 2) separating the acceleration of image generation from the acceleration of video motion generation, enhancing training efficiency and mitigating the negative effects of low-quality video data.
comment: Accepted as a Short Paper by SIGGRAPH ASIA 2024 Technical Communications. This is a short version of the original work. Project Page: https://animatelcm.github.io/
♻ ☆ WEAR: An Outdoor Sports Dataset for Wearable and Egocentric Activity Recognition
Research has shown the complementarity of camera- and inertial-based data for modeling human activities, yet datasets with both egocentric video and inertial-based sensor data remain scarce. In this paper, we introduce WEAR, an outdoor sports dataset for both vision- and inertial-based human activity recognition (HAR). Data from 22 participants performing a total of 18 different workout activities was collected with synchronized inertial (acceleration) and camera (egocentric video) data recorded at 11 different outside locations. WEAR provides a challenging prediction scenario in changing outdoor environments using a sensor placement, in line with recent trends in real-world applications. Benchmark results show that through our sensor placement, each modality interestingly offers complementary strengths and weaknesses in their prediction performance. Further, in light of the recent success of single-stage Temporal Action Localization (TAL) models, we demonstrate their versatility of not only being trained using visual data, but also using raw inertial data and being capable to fuse both modalities by means of simple concatenation. The dataset and code to reproduce experiments is publicly available via: mariusbock.github.io/wear/.
comment: accepted at IMWUT; 21 pages, 8 figures, 2 tables
♻ ☆ MetaEarth: A Generative Foundation Model for Global-Scale Remote Sensing Image Generation
The recent advancement of generative foundational models has ushered in a new era of image generation in the realm of natural images, revolutionizing art design, entertainment, environment simulation, and beyond. Despite producing high-quality samples, existing methods are constrained to generating images of scenes at a limited scale. In this paper, we present MetaEarth, a generative foundation model that breaks the barrier by scaling image generation to a global level, exploring the creation of worldwide, multi-resolution, unbounded, and virtually limitless remote sensing images. In MetaEarth, we propose a resolution-guided self-cascading generative framework, which enables the generating of images at any region with a wide range of geographical resolutions. To achieve unbounded and arbitrary-sized image generation, we design a novel noise sampling strategy for denoising diffusion models by analyzing the generation conditions and initial noise. To train MetaEarth, we construct a large dataset comprising multi-resolution optical remote sensing images with geographical information. Experiments have demonstrated the powerful capabilities of our method in generating global-scale images. Additionally, the MetaEarth serves as a data engine that can provide high-quality and rich training data for downstream tasks. Our model opens up new possibilities for constructing generative world models by simulating Earth visuals from an innovative overhead perspective.
comment: Project page: https://jiupinjia.github.io/metaearth/
♻ ☆ Improving Siamese Based Trackers with Light or No Training through Multiple Templates and Temporal Network
High computational power and significant time are usually needed to train a deep learning based tracker on large datasets. Depending on many factors, training might not always be an option. In this paper, we propose a framework with two ideas on Siamese-based trackers. (i) Extending number of templates in a way that removes the need to retrain the network and (ii) a lightweight temporal network with a novel architecture focusing on both local and global information that can be used independently from trackers. Most Siamese-based trackers only rely on the first frame as the ground truth for objects and struggle when the target's appearance changes significantly in subsequent frames in presence of similar distractors. Some trackers use multiple templates which mostly rely on constant thresholds to update, or they replace those templates that have low similarity scores only with more similar ones. Unlike previous works, we use adaptive thresholds that update the bag with similar templates as well as those templates which are slightly diverse. Adaptive thresholds also cause an overall improvement over constant ones. In addition, mixing feature maps obtained by each template in the last stage of networks removes the need to retrain trackers. Our proposed lightweight temporal network, CombiNet, learns the path history of different objects using only object coordinates and predicts target's potential location in the next frame. It is tracker independent and applying it on new trackers does not need further training. By implementing these ideas, trackers' performance improved on all datasets tested on, including LaSOT, LaSOT extension, TrackingNet, OTB100, OTB50, UAV123 and UAV20L. Experiments indicate the proposed framework works well with both convolutional and transformer-based trackers. The official python code for this paper will be publicly available upon publication.
♻ ☆ On Representation of 3D Rotation in the Context of Deep Learning ICCV
This paper investigates various methods of representing 3D rotations and their impact on the learning process of deep neural networks. We evaluated the performance of ResNet18 networks for 3D rotation estimation using several rotation representations and loss functions on both synthetic and real data. The real datasets contained 3D scans of industrial bins, while the synthetic datasets included views of a simple asymmetric object rendered under different rotations. On synthetic data, we also assessed the effects of different rotation distributions within the training and test sets, as well as the impact of the object's texture. In line with previous research, we found that networks using the continuous 5D and 6D representations performed better than the discontinuous ones.
comment: Accepted at International Conference on Computer Vision and Graphics ICCVG 2024. The proceedings of the conference will be published in Lecture Notes in Networks and Systems (LNNS), Springer
♻ ☆ Spectral-GS: Taming 3D Gaussian Splatting with Spectral Entropy
Recently, 3D Gaussian Splatting (3D-GS) has achieved impressive results in novel view synthesis, demonstrating high fidelity and efficiency. However, it easily exhibits needle-like artifacts, especially when increasing the sampling rate. Mip-Splatting tries to remove these artifacts with a 3D smoothing filter for frequency constraints and a 2D Mip filter for approximated supersampling. Unfortunately, it tends to produce over-blurred results, and sometimes needle-like Gaussians still persist. Our spectral analysis of the covariance matrix during optimization and densification reveals that current 3D-GS lacks shape awareness, relying instead on spectral radius and view positional gradients to determine splitting. As a result, needle-like Gaussians with small positional gradients and low spectral entropy fail to split and overfit high-frequency details. Furthermore, both the filters used in 3D-GS and Mip-Splatting reduce the spectral entropy and increase the condition number during zooming in to synthesize novel view, causing view inconsistencies and more pronounced artifacts. Our Spectral-GS, based on spectral analysis, introduces 3D shape-aware splitting and 2D view-consistent filtering strategies, effectively addressing these issues, enhancing 3D-GS's capability to represent high-frequency details without noticeable artifacts, and achieving high-quality photorealistic rendering.
♻ ☆ Advancing Supervised Local Learning Beyond Classification with Long-term Feature Bank
Local learning offers an alternative to traditional end-to-end back-propagation in deep neural networks, significantly reducing GPU memory usage. While local learning has shown promise in image classification tasks, its application to other visual tasks remains limited. This limitation arises primarily from two factors: 1) architectures tailored for classification are often not transferable to other tasks, leading to a lack of reusability of task-specific knowledge; 2) the absence of cross-scale feature communication results in degraded performance in tasks such as object detection and super-resolution. To address these challenges, we propose the Memory-augmented Auxiliary Network (MAN), which introduces a simplified design principle and incorporates a feature bank to enhance cross-task adaptability and communication. This work represents the first successful application of local learning methods beyond classification, demonstrating that MAN not only conserves GPU memory but also achieves performance on par with end-to-end approaches across multiple datasets for various visual tasks.
♻ ☆ Back to the Color: Learning Depth to Specific Color Transformation for Unsupervised Depth Estimation
Virtual engines can generate dense depth maps for various synthetic scenes, making them invaluable for training depth estimation models. However, discrepancies between synthetic and real-world colors pose significant challenges for depth estimation in real-world scenes, especially in complex and uncertain environments encountered in unsupervised monocular depth estimation tasks. To address this issue, we propose Back2Color, a framework that predicts realistic colors from depth using a model trained on real-world data, thus transforming synthetic colors into their real-world counterparts. Additionally, we introduce the Syn-Real CutMix method for joint training with both real-world unsupervised and synthetic supervised depth samples, enhancing monocular depth estimation performance in real-world scenes. Furthermore, to mitigate the impact of non-rigid motions on depth estimation, we present an auto-learning uncertainty temporal-spatial fusion method (Auto-UTSF), which leverages the strengths of unsupervised learning in both temporal and spatial dimensions. We also designed VADepth, based on the Vision Attention Network, which offers lower computational complexity and higher accuracy than transformers. Our Back2Color framework achieves state-of-the-art performance on the Kitti dataset, as evidenced by improvements in performance metrics and the production of fine-grained details. This is particularly evident on more challenging datasets such as Cityscapes for unsupervised depth estimation.
♻ ☆ Faster Diffusion: Rethinking the Role of the Encoder for Diffusion Model Inference NeurIPS 2024
One of the main drawback of diffusion models is the slow inference time for image generation. Among the most successful approaches to addressing this problem are distillation methods. However, these methods require considerable computational resources. In this paper, we take another approach to diffusion model acceleration. We conduct a comprehensive study of the UNet encoder and empirically analyze the encoder features. This provides insights regarding their changes during the inference process. In particular, we find that encoder features change minimally, whereas the decoder features exhibit substantial variations across different time-steps. This insight motivates us to omit encoder computation at certain adjacent time-steps and reuse encoder features of previous time-steps as input to the decoder in multiple time-steps. Importantly, this allows us to perform decoder computation in parallel, further accelerating the denoising process. Additionally, we introduce a prior noise injection method to improve the texture details in the generated image. Besides the standard text-to-image task, we also validate our approach on other tasks: text-to-video, personalized generation and reference-guided generation. Without utilizing any knowledge distillation technique, our approach accelerates both the Stable Diffusion (SD) and DeepFloyd-IF model sampling by 41$\%$ and 24$\%$ respectively, and DiT model sampling by 34$\%$, while maintaining high-quality generation performance.
comment: NeurIPS 2024
♻ ☆ LiveXiv -- A Multi-Modal Live Benchmark Based on Arxiv Papers Content
The large-scale training of multi-modal models on data scraped from the web has shown outstanding utility in infusing these models with the required world knowledge to perform effectively on multiple downstream tasks. However, one downside of scraping data from the web can be the potential sacrifice of the benchmarks on which the abilities of these models are often evaluated. To safeguard against test data contamination and to truly test the abilities of these foundation models we propose LiveXiv: A scalable evolving live benchmark based on scientific ArXiv papers. LiveXiv accesses domain-specific manuscripts at any given timestamp and proposes to automatically generate visual question-answer pairs (VQA). This is done without any human-in-the-loop, using the multi-modal content in the manuscripts, like graphs, charts, and tables. Moreover, we introduce an efficient evaluation approach that estimates the performance of all models on the evolving benchmark using evaluations of only a subset of models. This significantly reduces the overall evaluation cost. We benchmark multiple open and proprietary Large Multi-modal Models (LMMs) on the first version of our benchmark, showing its challenging nature and exposing the models true abilities, avoiding contamination. Lastly, in our commitment to high quality, we have collected and evaluated a manually verified subset. By comparing its overall results to our automatic annotations, we have found that the performance variance is indeed minimal (<2.5%). Our dataset is available online on HuggingFace, and our code will be available here.
♻ ☆ MMAR: Towards Lossless Multi-Modal Auto-Regressive Probabilistic Modeling
Recent advancements in multi-modal large language models have propelled the development of joint probabilistic models capable of both image understanding and generation. However, we have identified that recent methods inevitably suffer from loss of image information during understanding task, due to either image discretization or diffusion denoising steps. To address this issue, we propose a novel Multi-Modal Auto-Regressive (MMAR) probabilistic modeling framework. Unlike discretization line of method, MMAR takes in continuous-valued image tokens to avoid information loss. Differing from diffusion-based approaches, we disentangle the diffusion process from auto-regressive backbone model by employing a light-weight diffusion head on top each auto-regressed image patch embedding. In this way, when the model transits from image generation to understanding through text generation, the backbone model's hidden representation of the image is not limited to the last denoising step. To successfully train our method, we also propose a theoretically proven technique that addresses the numerical stability issue and a training strategy that balances the generation and understanding task goals. Through extensive evaluations on 18 image understanding benchmarks, MMAR demonstrates much more superior performance than other joint multi-modal models, matching the method that employs pretrained CLIP vision encoder, meanwhile being able to generate high quality images at the same time. We also showed that our method is scalable with larger data and model size.
♻ ☆ Parameterize Structure with Differentiable Template for 3D Shape Generation
Structural representation is crucial for reconstructing and generating editable 3D shapes with part semantics. Recent 3D shape generation works employ complicated networks and structure definitions relying on hierarchical annotations and pay less attention to the details inside parts. In this paper, we propose the method that parameterizes the shared structure in the same category using a differentiable template and corresponding fixed-length parameters. Specific parameters are fed into the template to calculate cuboids that indicate a concrete shape. We utilize the boundaries of three-view drawings of each cuboid to further describe the inside details. Shapes are represented with the parameters and three-view details inside cuboids, from which the SDF can be calculated to recover the object. Benefiting from our fixed-length parameters and three-view details, our networks for reconstruction and generation are simple and effective to learn the latent space. Our method can reconstruct or generate diverse shapes with complicated details, and interpolate them smoothly. Extensive evaluations demonstrate the superiority of our method on reconstruction from point cloud, generation, and interpolation.
♻ ☆ On the Adversarial Risk of Test Time Adaptation: An Investigation into Realistic Test-Time Data Poisoning
Test-time adaptation (TTA) updates the model weights during the inference stage using testing data to enhance generalization. However, this practice exposes TTA to adversarial risks. Existing studies have shown that when TTA is updated with crafted adversarial test samples, also known as test-time poisoned data, the performance on benign samples can deteriorate. Nonetheless, the perceived adversarial risk may be overstated if the poisoned data is generated under overly strong assumptions. In this work, we first review realistic assumptions for test-time data poisoning, including white-box versus grey-box attacks, access to benign data, attack budget, and more. We then propose an effective and realistic attack method that better produces poisoned samples without access to benign samples, and derive an effective in-distribution attack objective. We also design two TTA-aware attack objectives. Our benchmarks of existing attack methods reveal that the TTA methods are more robust than previously believed. In addition, we analyze effective defense strategies to help develop adversarially robust TTA methods.
comment: 19 pages, 4 figures, 8 tables
♻ ☆ Towards Defining an Efficient and Expandable File Format for AI-Generated Contents
Recently, AI-generated content (AIGC) has gained significant traction due to its powerful creation capability. However, the storage and transmission of large amounts of high-quality AIGC images inevitably pose new challenges for recent file formats. To overcome this, we define a new file format for AIGC images, named AIGIF, enabling ultra-low bitrate coding of AIGC images. Unlike compressing AIGC images intuitively with pixel-wise space as existing file formats, AIGIF instead compresses the generation syntax. This raises a crucial question: Which generation syntax elements, e.g., text prompt, device configuration, etc, are necessary for compression/transmission? To answer this question, we systematically investigate the effects of three essential factors: platform, generative model, and data configuration. We experimentally find that a well-designed composable bitstream structure incorporating the above three factors can achieve an impressive compression ratio of even up to 1/10,000 while still ensuring high fidelity. We also introduce an expandable syntax in AIGIF to support the extension of the most advanced generation models to be developed in the future.
♻ ☆ SANA: Efficient High-Resolution Image Synthesis with Linear Diffusion Transformers
We introduce Sana, a text-to-image framework that can efficiently generate images up to 4096$\times$4096 resolution. Sana can synthesize high-resolution, high-quality images with strong text-image alignment at a remarkably fast speed, deployable on laptop GPU. Core designs include: (1) Deep compression autoencoder: unlike traditional AEs, which compress images only 8$\times$, we trained an AE that can compress images 32$\times$, effectively reducing the number of latent tokens. (2) Linear DiT: we replace all vanilla attention in DiT with linear attention, which is more efficient at high resolutions without sacrificing quality. (3) Decoder-only text encoder: we replaced T5 with modern decoder-only small LLM as the text encoder and designed complex human instruction with in-context learning to enhance the image-text alignment. (4) Efficient training and sampling: we propose Flow-DPM-Solver to reduce sampling steps, with efficient caption labeling and selection to accelerate convergence. As a result, Sana-0.6B is very competitive with modern giant diffusion model (e.g. Flux-12B), being 20 times smaller and 100+ times faster in measured throughput. Moreover, Sana-0.6B can be deployed on a 16GB laptop GPU, taking less than 1 second to generate a 1024$\times$1024 resolution image. Sana enables content creation at low cost. Code and model will be publicly released.
comment: Technical Report
♻ ☆ EVF-SAM: Early Vision-Language Fusion for Text-Prompted Segment Anything Model
Segment Anything Model (SAM) has attracted widespread attention for its superior interactive segmentation capabilities with visual prompts while lacking further exploration of text prompts. In this paper, we empirically investigate what text prompt encoders (e.g., CLIP or LLM) are good for adapting SAM for referring expression segmentation and introduce the Early Vision-language Fusion-based SAM (EVF-SAM). EVF-SAM is a simple yet effective referring segmentation method which exploits multimodal prompts (i.e., image and text) and comprises a pre-trained vision-language model to generate referring prompts and a SAM model for segmentation. Surprisingly, we observe that: (1) multimodal prompts and (2) vision-language models with early fusion (e.g., BEIT-3) are beneficial for prompting SAM for accurate referring segmentation. Our experiments show that the proposed EVF-SAM based on BEIT-3 can obtain state-of-the-art performance on RefCOCO/+/g for referring expression segmentation and demonstrate the superiority of prompting SAM with early vision-language fusion. In addition, the proposed EVF-SAM with 1.32B parameters achieves remarkably higher performance while reducing nearly 82% of parameters compared to previous SAM methods based on large multimodal models.
comment: Preprint. Update: (1) better performance and (2) versatile segmentation. Code and models are available at: https://github.com/hustvl/EVF-SAM
♻ ☆ Swarm Intelligence in Geo-Localization: A Multi-Agent Large Vision-Language Model Collaborative Framework
Visual geo-localization demands in-depth knowledge and advanced reasoning skills to associate images with precise real-world geographic locations. Existing image database retrieval methods are limited by the impracticality of storing sufficient visual records of global landmarks. Recently, Large Vision-Language Models (LVLMs) have demonstrated the capability of geo-localization through Visual Question Answering (VQA), enabling a solution that does not require external geo-tagged image records. However, the performance of a single LVLM is still limited by its intrinsic knowledge and reasoning capabilities. To address these challenges, we introduce smileGeo, a novel visual geo-localization framework that leverages multiple Internet-enabled LVLM agents operating within an agent-based architecture. By facilitating inter-agent communication, smileGeo integrates the inherent knowledge of these agents with additional retrieved information, enhancing the ability to effectively localize images. Furthermore, our framework incorporates a dynamic learning strategy that optimizes agent communication, reducing redundant interactions and enhancing overall system efficiency. To validate the effectiveness of the proposed framework, we conducted experiments on three different datasets, and the results show that our approach significantly outperforms current state-of-the-art methods. The source code is available at https://anonymous.4open.science/r/ViusalGeoLocalization-F8F5.
comment: resubmit to www2025
♻ ☆ Ensemble of ConvNeXt V2 and MaxViT for Long-Tailed CXR Classification with View-Based Aggregation MICCAI
In this work, we present our solution for the MICCAI 2024 CXR-LT challenge, achieving 4th place in Subtask 2 and 5th in Subtask 1. We leveraged an ensemble of ConvNeXt V2 and MaxViT models, pretrained on an external chest X-ray dataset, to address the long-tailed distribution of chest findings. The proposed method combines state-of-the-art image classification techniques, asymmetric loss for handling class imbalance, and view-based prediction aggregation to enhance classification performance. Through experiments, we demonstrate the advantages of our approach in improving both detection accuracy and the handling of the long-tailed distribution in CXR findings. The code is available at https://github.com/yamagishi0824/cxrlt24-multiview-pp.
comment: Solution paper for MICCAI CXR-LT 2024 challenge. 4th place in Subtask 2, 5th in Subtask 1
♻ ☆ A Training-free Sub-quadratic Cost Transformer Model Serving Framework With Hierarchically Pruned Attention
In modern large language models (LLMs), increasing the context length is crucial for improving comprehension and coherence in long-context, multi-modal, and retrieval-augmented language generation. While many recent transformer models attempt to extend their context length over a million tokens, they remain impractical due to the quadratic time and space complexities. Although recent works on linear and sparse attention mechanisms can achieve this goal, their real-world applicability is often limited by the need to re-train from scratch and significantly worse performance. In response, we propose a novel approach, Hierarchically Pruned Attention (HiP), which reduces the time complexity of the attention mechanism to $O(T \log T)$ and the space complexity to $O(T)$, where $T$ is the sequence length. We notice a pattern in the attention scores of pretrained LLMs where tokens close together tend to have similar scores, which we call ``attention locality''. Based on this observation, we utilize a novel tree-search-like algorithm that estimates the top-$k$ key tokens for a given query on the fly, which is mathematically guaranteed to have better performance than random attention pruning. In addition to improving the time complexity of the attention mechanism, we further optimize GPU memory usage by implementing KV cache offloading, which stores only $O(\log T)$ tokens on the GPU while maintaining similar decoding throughput. Experiments on benchmarks show that HiP, with its training-free nature, significantly reduces both prefill and decoding latencies, as well as memory usage, while maintaining high-quality generation with minimal degradation. HiP enables pretrained LLMs to scale up to millions of tokens on commodity GPUs, potentially unlocking long-context LLM applications previously deemed infeasible.
comment: 44 pages
♻ ☆ Diffusion Models in 3D Vision: A Survey
In recent years, 3D vision has become a crucial field within computer vision, powering a wide range of applications such as autonomous driving, robotics, augmented reality (AR), and medical imaging. This field relies on the accurate perception, understanding, and reconstruction of 3D scenes from 2D data sources like images and videos. Diffusion models, originally designed for 2D generative tasks, offer the potential for more flexible, probabilistic approaches that can better capture the variability and uncertainty present in real-world 3D data. However, traditional methods often struggle with efficiency and scalability. In this paper, we review the state-of-the-art approaches that leverage diffusion models for 3D visual tasks, including but not limited to 3D object generation, shape completion, point cloud reconstruction, and scene understanding. We provide an in-depth discussion of the underlying mathematical principles of diffusion models, outlining their forward and reverse processes, as well as the various architectural advancements that enable these models to work with 3D datasets. We also discuss the key challenges in applying diffusion models to 3D vision, such as handling occlusions and varying point densities, and the computational demands of high-dimensional data. Finally, we discuss potential solutions, including improving computational efficiency, enhancing multimodal fusion, and exploring the use of large-scale pretraining for better generalization across 3D tasks. This paper serves as a foundation for future exploration and development in this rapidly evolving field.
♻ ☆ MonoMAE: Enhancing Monocular 3D Detection through Depth-Aware Masked Autoencoders NeurIPS 2024
Monocular 3D object detection aims for precise 3D localization and identification of objects from a single-view image. Despite its recent progress, it often struggles while handling pervasive object occlusions that tend to complicate and degrade the prediction of object dimensions, depths, and orientations. We design MonoMAE, a monocular 3D detector inspired by Masked Autoencoders that addresses the object occlusion issue by masking and reconstructing objects in the feature space. MonoMAE consists of two novel designs. The first is depth-aware masking that selectively masks certain parts of non-occluded object queries in the feature space for simulating occluded object queries for network training. It masks non-occluded object queries by balancing the masked and preserved query portions adaptively according to the depth information. The second is lightweight query completion that works with the depth-aware masking to learn to reconstruct and complete the masked object queries. With the proposed object occlusion and completion, MonoMAE learns enriched 3D representations that achieve superior monocular 3D detection performance qualitatively and quantitatively for both occluded and non-occluded objects. Additionally, MonoMAE learns generalizable representations that can work well in new domains.
comment: NeurIPS 2024
♻ ☆ Dual-Modeling Decouple Distillation for Unsupervised Anomaly Detection ACM MM '24
Knowledge distillation based on student-teacher network is one of the mainstream solution paradigms for the challenging unsupervised Anomaly Detection task, utilizing the difference in representation capabilities of the teacher and student networks to implement anomaly localization. However, over-generalization of the student network to the teacher network may lead to negligible differences in representation capabilities of anomaly, thus affecting the detection effectiveness. Existing methods address the possible over-generalization by using differentiated students and teachers from the structural perspective or explicitly expanding distilled information from the content perspective, which inevitably result in an increased likelihood of underfitting of the student network and poor anomaly detection capabilities in anomaly center or edge. In this paper, we propose Dual-Modeling Decouple Distillation (DMDD) for the unsupervised anomaly detection. In DMDD, a Decouple Student-Teacher Network is proposed to decouple the initial student features into normality and abnormality features. We further introduce Dual-Modeling Distillation based on normal-anomaly image pairs, fitting normality features of anomalous image and the teacher features of the corresponding normal image, widening the distance between abnormality features and the teacher features in anomalous regions. Synthesizing these two distillation ideas, we achieve anomaly detection which focuses on both edge and center of anomaly. Finally, a Multi-perception Segmentation Network is proposed to achieve focused anomaly map fusion based on multiple attention. Experimental results on MVTec AD show that DMDD surpasses SOTA localization performance of previous knowledge distillation-based methods, reaching 98.85% on pixel-level AUC and 96.13% on PRO.
comment: 10 pages, 8 figures, Accepted to ACM MM '24
♻ ☆ Pose-dIVE: Pose-Diversified Augmentation with Diffusion Model for Person Re-Identification
Person re-identification (Re-ID) often faces challenges due to variations in human poses and camera viewpoints, which significantly affect the appearance of individuals across images. Existing datasets frequently lack diversity and scalability in these aspects, hindering the generalization of Re-ID models to new camera systems. We propose Pose-dIVE, a novel data augmentation approach that incorporates sparse and underrepresented human pose and camera viewpoint examples into the training data, addressing the limited diversity in the original training data distribution. Our objective is to augment the training dataset to enable existing Re-ID models to learn features unbiased by human pose and camera viewpoint variations. To achieve this, we leverage the knowledge of pre-trained large-scale diffusion models. By conditioning the diffusion model on both the human pose and camera viewpoint concurrently through the SMPL model, we generate training data with diverse human poses and camera viewpoints. Experimental results demonstrate the effectiveness of our method in addressing human pose bias and enhancing the generalizability of Re-ID models compared to other data augmentation-based Re-ID approaches.
♻ ☆ Eyes Closed, Safety On: Protecting Multimodal LLMs via Image-to-Text Transformation ECCV2024
Multimodal large language models (MLLMs) have shown impressive reasoning abilities. However, they are also more vulnerable to jailbreak attacks than their LLM predecessors. Although still capable of detecting the unsafe responses, we observe that safety mechanisms of the pre-aligned LLMs in MLLMs can be easily bypassed with the introduction of image features. To construct robust MLLMs, we propose ECSO (Eyes Closed, Safety On), a novel training-free protecting approach that exploits the inherent safety awareness of MLLMs, and generates safer responses via adaptively transforming unsafe images into texts to activate the intrinsic safety mechanism of pre-aligned LLMs in MLLMs. Experiments on five state-of-the-art (SoTA) MLLMs demonstrate that ECSO enhances model safety significantly (e.g.,, 37.6% improvement on the MM-SafetyBench (SD+OCR) and 71.3% on VLSafe with LLaVA-1.5-7B), while consistently maintaining utility results on common MLLM benchmarks. Furthermore, we show that ECSO can be used as a data engine to generate supervised-finetuning (SFT) data for MLLM alignment without extra human intervention.
comment: ECCV2024 (Project Page: https://gyhdog99.github.io/projects/ecso/)
♻ ☆ DriVLMe: Enhancing LLM-based Autonomous Driving Agents with Embodied and Social Experiences IROS
Recent advancements in foundation models (FMs) have unlocked new prospects in autonomous driving, yet the experimental settings of these studies are preliminary, over-simplified, and fail to capture the complexity of real-world driving scenarios in human environments. It remains under-explored whether FM agents can handle long-horizon navigation tasks with free-from dialogue and deal with unexpected situations caused by environmental dynamics or task changes. To explore the capabilities and boundaries of FMs faced with the challenges above, we introduce DriVLMe, a video-language-model-based agent to facilitate natural and effective communication between humans and autonomous vehicles that perceive the environment and navigate. We develop DriVLMe from both embodied experiences in a simulated environment and social experiences from real human dialogue. While DriVLMe demonstrates competitive performance in both open-loop benchmarks and closed-loop human studies, we reveal several limitations and challenges, including unacceptable inference time, imbalanced training data, limited visual understanding, challenges with multi-turn interactions, simplified language generation from robotic experiences, and difficulties in handling on-the-fly unexpected situations like environmental dynamics and task changes.
comment: 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
♻ ☆ Triple Point Masking
Existing 3D mask learning methods encounter performance bottlenecks under limited data, and our objective is to overcome this limitation. In this paper, we introduce a triple point masking scheme, named TPM, which serves as a scalable framework for pre-training of masked autoencoders to achieve multi-mask learning for 3D point clouds. Specifically, we augment the baselines with two additional mask choices (i.e., medium mask and low mask) as our core insight is that the recovery process of an object can manifest in diverse ways. Previous high-masking schemes focus on capturing the global representation but lack the fine-grained recovery capability, so that the generated pre-trained weights tend to play a limited role in the fine-tuning process. With the support of the proposed TPM, available methods can exhibit more flexible and accurate completion capabilities, enabling the potential autoencoder in the pre-training stage to consider multiple representations of a single 3D object. In addition, an SVM-guided weight selection module is proposed to fill the encoder parameters for downstream networks with the optimal weight during the fine-tuning stage, maximizing linear accuracy and facilitating the acquisition of intricate representations for new objects. Extensive experiments show that the four baselines equipped with the proposed TPM achieve comprehensive performance improvements on various downstream tasks. Our code and models are available at https://github.com/liujia99/TPM.
♻ ☆ Edit-Your-Motion: Space-Time Diffusion Decoupling Learning for Video Motion Editing
Existing diffusion-based methods have achieved impressive results in human motion editing. However, these methods often exhibit significant ghosting and body distortion in unseen in-the-wild cases. In this paper, we introduce Edit-Your-Motion, a video motion editing method that tackles these challenges through one-shot fine-tuning on unseen cases. Specifically, firstly, we utilized DDIM inversion to initialize the noise, preserving the appearance of the source video and designed a lightweight motion attention adapter module to enhance motion fidelity. DDIM inversion aims to obtain the implicit representations by estimating the prediction noise from the source video, which serves as a starting point for the sampling process, ensuring the appearance consistency between the source and edited videos. The Motion Attention Module (MA) enhances the model's motion editing ability by resolving the conflict between the skeleton features and the appearance features. Secondly, to effectively decouple motion and appearance of source video, we design a spatio-temporal two-stage learning strategy (STL). In the first stage, we focus on learning temporal features of human motion and propose recurrent causal attention (RCA) to ensure consistency between video frames. In the second stage, we shift focus on learning the appearance features of the source video. With Edit-Your-Motion, users can edit the motion of humans in the source video, creating more engaging and diverse content. Extensive qualitative and quantitative experiments, along with user preference studies, show that Edit-Your-Motion outperforms other methods.
♻ ☆ Harmonic Mobile Manipulation
Recent advancements in robotics have enabled robots to navigate complex scenes or manipulate diverse objects independently. However, robots are still impotent in many household tasks requiring coordinated behaviors such as opening doors. The factorization of navigation and manipulation, while effective for some tasks, fails in scenarios requiring coordinated actions. To address this challenge, we introduce, HarmonicMM, an end-to-end learning method that optimizes both navigation and manipulation, showing notable improvement over existing techniques in everyday tasks. This approach is validated in simulated and real-world environments and adapts to novel unseen settings without additional tuning. Our contributions include a new benchmark for mobile manipulation and the successful deployment with only RGB visual observation in a real unseen apartment, demonstrating the potential for practical indoor robot deployment in daily life. More results are on our project site: https://rchalyang.github.io/HarmonicMM/
comment: More results are on our project site: https://rchalyang.github.io/HarmonicMM/
♻ ☆ Fast Encoding and Decoding for Implicit Video Representation ECCV 2024
Despite the abundant availability and content richness for video data, its high-dimensionality poses challenges for video research. Recent advancements have explored the implicit representation for videos using neural networks, demonstrating strong performance in applications such as video compression and enhancement. However, the prolonged encoding time remains a persistent challenge for video Implicit Neural Representations (INRs). In this paper, we focus on improving the speed of video encoding and decoding within implicit representations. We introduce two key components: NeRV-Enc, a transformer-based hyper-network for fast encoding; and NeRV-Dec, a parallel decoder for efficient video loading. NeRV-Enc achieves an impressive speed-up of $\mathbf{10^4\times}$ by eliminating gradient-based optimization. Meanwhile, NeRV-Dec simplifies video decoding, outperforming conventional codecs with a loading speed $\mathbf{11\times}$ faster, and surpassing RAM loading with pre-decoded videos ($\mathbf{2.5\times}$ faster while being $\mathbf{65\times}$ smaller in size).
comment: ECCV 2024. Project page at https://haochen-rye.github.io/FastNeRV/, code will be at https://github.com/haochen-rye/FastNeRV
♻ ☆ DDR: Exploiting Deep Degradation Response as Flexible Image Descriptor NeurIPS
Image deep features extracted by pre-trained networks are known to contain rich and informative representations. In this paper, we present Deep Degradation Response (DDR), a method to quantify changes in image deep features under varying degradation conditions. Specifically, our approach facilitates flexible and adaptive degradation, enabling the controlled synthesis of image degradation through text-driven prompts. Extensive evaluations demonstrate the versatility of DDR as an image descriptor, with strong correlations observed with key image attributes such as complexity, colorfulness, sharpness, and overall quality. Moreover, we demonstrate the efficacy of DDR across a spectrum of applications. It excels as a blind image quality assessment metric, outperforming existing methodologies across multiple datasets. Additionally, DDR serves as an effective unsupervised learning objective in image restoration tasks, yielding notable advancements in image deblurring and single-image super-resolution. Our code is available at: https://github.com/eezkni/DDR
comment: Accepted to Advances in Neural Information Processing Systems (NeurIPS) 2024
♻ ☆ M3Bench: Benchmarking Whole-body Motion Generation for Mobile Manipulation in 3D Scenes
We propose M^3Bench, a new benchmark of whole-body motion generation for mobile manipulation tasks. Given a 3D scene context, M^3Bench requires an embodied agent to understand its configuration, environmental constraints and task objectives, then generate coordinated whole-body motion trajectories for object rearrangement tasks. M^3Bench features 30k object rearrangement tasks across 119 diverse scenes, providing expert demonstrations generated by our newly developed M^3BenchMaker. This automatic data generation tool produces coordinated whole-body motion trajectories from high-level task instructions, requiring only basic scene and robot information. Our benchmark incorporates various task splits to assess generalization across different dimensions and leverages realistic physics simulation for trajectory evaluation. Through extensive experimental analyses, we reveal that state-of-the-art models still struggle with coordinated base-arm motion while adhering to environment-context and task-specific constraints, highlighting the need to develop new models that address this gap. Through M^3Bench, we aim to facilitate future robotics research towards more adaptive and capable mobile manipulation in diverse, real-world environments.
comment: Code and data set will be released after acceptance
♻ ☆ Pubic Symphysis-Fetal Head Segmentation Network Using BiFormer Attention Mechanism and Multipath Dilated Convolution
Pubic symphysis-fetal head segmentation in transperineal ultrasound images plays a critical role for the assessment of fetal head descent and progression. Existing transformer segmentation methods based on sparse attention mechanism use handcrafted static patterns, which leads to great differences in terms of segmentation performance on specific datasets. To address this issue, we introduce a dynamic, query-aware sparse attention mechanism for ultrasound image segmentation. Specifically, we propose a novel method, named BRAU-Net to solve the pubic symphysis-fetal head segmentation task in this paper. The method adopts a U-Net-like encoder-decoder architecture with bi-level routing attention and skip connections, which effectively learns local-global semantic information. In addition, we propose an inverted bottleneck patch expanding (IBPE) module to reduce information loss while performing up-sampling operations. The proposed BRAU-Net is evaluated on FH-PS-AoP and HC18 datasets. The results demonstrate that our method could achieve excellent segmentation results. The code is available on GitHub.
comment: MMM2025;Camera-ready Version;The code is available at https://github.com/Caipengzhou/BRAU-Net
♻ ☆ FiT: Flexible Vision Transformer for Diffusion Model
Nature is infinitely resolution-free. In the context of this reality, existing diffusion models, such as Diffusion Transformers, often face challenges when processing image resolutions outside of their trained domain. To overcome this limitation, we present the Flexible Vision Transformer (FiT), a transformer architecture specifically designed for generating images with unrestricted resolutions and aspect ratios. Unlike traditional methods that perceive images as static-resolution grids, FiT conceptualizes images as sequences of dynamically-sized tokens. This perspective enables a flexible training strategy that effortlessly adapts to diverse aspect ratios during both training and inference phases, thus promoting resolution generalization and eliminating biases induced by image cropping. Enhanced by a meticulously adjusted network structure and the integration of training-free extrapolation techniques, FiT exhibits remarkable flexibility in resolution extrapolation generation. Comprehensive experiments demonstrate the exceptional performance of FiT across a broad range of resolutions, showcasing its effectiveness both within and beyond its training resolution distribution. Repository available at https://github.com/whlzy/FiT.
♻ ☆ MambaIR: A Simple Baseline for Image Restoration with State-Space Model ECCV2024
Recent years have seen significant advancements in image restoration, largely attributed to the development of modern deep neural networks, such as CNNs and Transformers. However, existing restoration backbones often face the dilemma between global receptive fields and efficient computation, hindering their application in practice. Recently, the Selective Structured State Space Model, especially the improved version Mamba, has shown great potential for long-range dependency modeling with linear complexity, which offers a way to resolve the above dilemma. However, the standard Mamba still faces certain challenges in low-level vision such as local pixel forgetting and channel redundancy. In this work, we introduce a simple but effective baseline, named MambaIR, which introduces both local enhancement and channel attention to improve the vanilla Mamba. In this way, our MambaIR takes advantage of the local pixel similarity and reduces the channel redundancy. Extensive experiments demonstrate the superiority of our method, for example, MambaIR outperforms SwinIR by up to 0.45dB on image SR, using similar computational cost but with a global receptive field. Code is available at \url{https://github.com/csguoh/MambaIR}.
comment: Accepted by ECCV2024
♻ ☆ Measuring Agreeableness Bias in Multimodal Models
This paper examines a phenomenon in multimodal language models where pre-marked options in question images can significantly influence model responses. Our study employs a systematic methodology to investigate this effect: we present models with images of multiple-choice questions, which they initially answer correctly, then expose the same model to versions with pre-marked options. Our findings reveal a significant shift in the models' responses towards the pre-marked option, even when it contradicts their answers in the neutral settings. Comprehensive evaluations demonstrate that this agreeableness bias is a consistent and quantifiable behavior across various model architectures. These results show potential limitations in the reliability of these models when processing images with pre-marked options, raising important questions about their application in critical decision-making contexts where such visual cues might be present.
♻ ☆ I-Max: Maximize the Resolution Potential of Pre-trained Rectified Flow Transformers with Projected Flow
Rectified Flow Transformers (RFTs) offer superior training and inference efficiency, making them likely the most viable direction for scaling up diffusion models. However, progress in generation resolution has been relatively slow due to data quality and training costs. Tuning-free resolution extrapolation presents an alternative, but current methods often reduce generative stability, limiting practical application. In this paper, we review existing resolution extrapolation methods and introduce the I-Max framework to maximize the resolution potential of Text-to-Image RFTs. I-Max features: (i) a novel Projected Flow strategy for stable extrapolation and (ii) an advanced inference toolkit for generalizing model knowledge to higher resolutions. Experiments with Lumina-Next-2K and Flux.1-dev demonstrate I-Max's ability to enhance stability in resolution extrapolation and show that it can bring image detail emergence and artifact correction, confirming the practical value of tuning-free resolution extrapolation.
♻ ☆ Optimizing Waste Management with Advanced Object Detection for Garbage Classification
Garbage production and littering are persistent global issues that pose significant environmental challenges. Despite large-scale efforts to manage waste through collection and sorting, existing approaches remain inefficient, leading to inadequate recycling and disposal. Therefore, developing advanced AI-based systems is less labor intensive approach for addressing the growing waste problem more effectively. These models can be applied to sorting systems or possibly waste collection robots that may produced in the future. AI models have grown significantly at identifying objects through object detection. This paper reviews the implementation of AI models for classifying trash through object detection, specifically focusing on using YOLO V5 for training and testing. The study demonstrates how YOLO V5 can effectively identify various types of waste, including plastic, paper, glass, metal, cardboard, and biodegradables.
comment: 8 pages, 8 figures
♻ ☆ RL-I2IT: Image-to-Image Translation with Deep Reinforcement Learning
Most existing Image-to-Image Translation (I2IT) methods generate images in a single run of a deep learning (DL) model. However, designing such a single-step model is always challenging, requiring a huge number of parameters and easily falling into bad global minimums and overfitting. In this work, we reformulate I2IT as a step-wise decision-making problem via deep reinforcement learning (DRL) and propose a novel framework that performs RL-based I2IT (RL-I2IT). The key feature in the RL-I2IT framework is to decompose a monolithic learning process into small steps with a lightweight model to progressively transform a source image successively to a target image. Considering that it is challenging to handle high dimensional continuous state and action spaces in the conventional RL framework, we introduce meta policy with a new concept Plan to the standard Actor-Critic model, which is of a lower dimension than the original image and can facilitate the actor to generate a tractable high dimensional action. In the RL-I2IT framework, we also employ a task-specific auxiliary learning strategy to stabilize the training process and improve the performance of the corresponding task. Experiments on several I2IT tasks demonstrate the effectiveness and robustness of the proposed method when facing high-dimensional continuous action space problems. Our implementation of the RL-I2IT framework is available at https://github.com/Algolzw/SPAC-Deformable-Registration.
♻ ☆ When ControlNet Meets Inexplicit Masks: A Case Study of ControlNet on its Contour-following Ability
ControlNet excels at creating content that closely matches precise contours in user-provided masks. However, when these masks contain noise, as a frequent occurrence with non-expert users, the output would include unwanted artifacts. This paper first highlights the crucial role of controlling the impact of these inexplicit masks with diverse deterioration levels through in-depth analysis. Subsequently, to enhance controllability with inexplicit masks, an advanced Shape-aware ControlNet consisting of a deterioration estimator and a shape-prior modulation block is devised. The deterioration estimator assesses the deterioration factor of the provided masks. Then this factor is utilized in the modulation block to adaptively modulate the model's contour-following ability, which helps it dismiss the noise part in the inexplicit masks. Extensive experiments prove its effectiveness in encouraging ControlNet to interpret inaccurate spatial conditions robustly rather than blindly following the given contours, suitable for diverse kinds of conditions. We showcase application scenarios like modifying shape priors and composable shape-controllable generation. Codes are available at github.
comment: Accepted by ACM-MM 2024
♻ ☆ Learning-based Axial Video Motion Magnification ECCV
Video motion magnification amplifies invisible small motions to be perceptible, which provides humans with a spatially dense and holistic understanding of small motions in the scene of interest. This is based on the premise that magnifying small motions enhances the legibility of motions. In the real world, however, vibrating objects often possess convoluted systems that have complex natural frequencies, modes, and directions. Existing motion magnification often fails to improve legibility since the intricate motions still retain complex characteristics even after being magnified, which may distract us from analyzing them. In this work, we focus on improving legibility by proposing a new concept, axial motion magnification, which magnifies decomposed motions along the user-specified direction. Axial motion magnification can be applied to various applications where motions of specific axes are critical, by providing simplified and easily readable motion information. To achieve this, we propose a novel Motion Separation Module that enables to disentangle and magnify the motion representation along axes of interest. Furthermore, we build a new synthetic training dataset for the axial motion magnification task. Our proposed method improves the legibility of resulting motions along certain axes by adding a new feature: user controllability. Axial motion magnification is a more generalized concept; thus, our method can be directly adapted to the generic motion magnification and achieves favorable performance against competing methods.
comment: Presented at European Conference on Computer Vision (ECCV) 2024. Project page (code&dataset): https://axial-momag.github.io/axial-momag/
♻ ☆ SpatialRGPT: Grounded Spatial Reasoning in Vision Language Models NeurIPS 2024
Vision Language Models (VLMs) have demonstrated remarkable performance in 2D vision and language tasks. However, their ability to reason about spatial arrangements remains limited. In this work, we introduce Spatial Region GPT (SpatialRGPT) to enhance VLMs' spatial perception and reasoning capabilities. SpatialRGPT advances VLMs' spatial understanding through two key innovations: (1) a data curation pipeline that enables effective learning of regional representation from 3D scene graphs, and (2) a flexible plugin module for integrating depth information into the visual encoder of existing VLMs. During inference, when provided with user-specified region proposals, SpatialRGPT can accurately perceive their relative directions and distances. Additionally, we propose SpatialRGBT-Bench, a benchmark with ground-truth 3D annotations encompassing indoor, outdoor, and simulated environments, for evaluating 3D spatial cognition in VLMs. Our results demonstrate that SpatialRGPT significantly enhances performance in spatial reasoning tasks, both with and without local region prompts. The model also exhibits strong generalization capabilities, effectively reasoning about complex spatial relations and functioning as a region-aware dense reward annotator for robotic tasks. Code, dataset, and benchmark are released at https://www.anjiecheng.me/SpatialRGPT
comment: NeurIPS 2024, Project Page: https://www.anjiecheng.me/SpatialRGPT
♻ ☆ Simultaneous Tri-Modal Medical Image Fusion and Super-Resolution using Conditional Diffusion Model MICCAI 2024
In clinical practice, tri-modal medical image fusion, compared to the existing dual-modal technique, can provide a more comprehensive view of the lesions, aiding physicians in evaluating the disease's shape, location, and biological activity. However, due to the limitations of imaging equipment and considerations for patient safety, the quality of medical images is usually limited, leading to sub-optimal fusion performance, and affecting the depth of image analysis by the physician. Thus, there is an urgent need for a technology that can both enhance image resolution and integrate multi-modal information. Although current image processing methods can effectively address image fusion and super-resolution individually, solving both problems synchronously remains extremely challenging. In this paper, we propose TFS-Diff, a simultaneously realize tri-modal medical image fusion and super-resolution model. Specially, TFS-Diff is based on the diffusion model generation of a random iterative denoising process. We also develop a simple objective function and the proposed fusion super-resolution loss, effectively evaluates the uncertainty in the fusion and ensures the stability of the optimization process. And the channel attention module is proposed to effectively integrate key information from different modalities for clinical diagnosis, avoiding information loss caused by multiple image processing. Extensive experiments on public Harvard datasets show that TFS-Diff significantly surpass the existing state-of-the-art methods in both quantitative and visual evaluations. Code is available at https://github.com/XylonXu01/TFS-Diff.
comment: Accepted by MICCAI 2024
♻ ☆ Dynamic Open Vocabulary Enhanced Safe-landing with Intelligence (DOVESEI) IROS 2023
This work targets what we consider to be the foundational step for urban airborne robots, a safe landing. Our attention is directed toward what we deem the most crucial aspect of the safe landing perception stack: segmentation. We present a streamlined reactive UAV system that employs visual servoing by harnessing the capabilities of open vocabulary image segmentation. This approach can adapt to various scenarios with minimal adjustments, bypassing the necessity for extensive data accumulation for refining internal models, thanks to its open vocabulary methodology. Given the limitations imposed by local authorities, our primary focus centers on operations originating from altitudes of 100 meters. This choice is deliberate, as numerous preceding works have dealt with altitudes up to 30 meters, aligning with the capabilities of small stereo cameras. Consequently, we leave the remaining 20m to be navigated using conventional 3D path planning methods. Utilizing monocular cameras and image segmentation, our findings demonstrate the system's capability to successfully execute landing maneuvers at altitudes as low as 20 meters. However, this approach is vulnerable to intermittent and occasionally abrupt fluctuations in the segmentation between frames in a video stream. To address this challenge, we enhance the image segmentation output by introducing what we call a dynamic focus: a masking mechanism that self adjusts according to the current landing stage. This dynamic focus guides the control system to avoid regions beyond the drone's safety radius projected onto the ground, thus mitigating the problems with fluctuations. Through the implementation of this supplementary layer, our experiments have reached improvements in the landing success rate of almost tenfold when compared to global segmentation. All the source code is open source and available online (github.com/MISTLab/DOVESEI).
comment: IROS 2023 The Last-Mile Robotics Workshop
♻ ☆ CNG-SFDA:Clean-and-Noisy Region Guided Online-Offline Source-Free Domain Adaptation ACCV 2024
Domain shift occurs when training (source) and test (target) data diverge in their distribution. Source-Free Domain Adaptation (SFDA) addresses this domain shift problem, aiming to adopt a trained model on the source domain to the target domain in a scenario where only a well-trained source model and unlabeled target data are available. In this scenario, handling false labels in the target domain is crucial because they negatively impact the model performance. To deal with this problem, we propose to update cluster prototypes (i.e., centroid of each sample cluster) and their structure in the target domain formulated by the source model in online manners. In the feature space, samples in different regions have different pseudo-label distribution characteristics affected by the cluster prototypes, and we adopt distinct training strategies for these samples by defining clean and noisy regions: we selectively train the target with clean pseudo-labels in the clean region, whereas we introduce mix-up inputs representing intermediate features between clean and noisy regions to increase the compactness of the cluster. We conducted extensive experiments on multiple datasets in online/offline SFDA settings, whose results demonstrate that our method, CNG-SFDA, achieves state-of-the-art for most cases. Code is available at https://github.com/hyeonwoocho7/CNG-SFDA.
comment: 14 pages, 5 figures, ACCV 2024 Camera-Ready Version
♻ ☆ Nellie: Automated organelle segmentation, tracking, and hierarchical feature extraction in 2D/3D live-cell microscopy
The analysis of dynamic organelles remains a formidable challenge, though key to understanding biological processes. We introduce Nellie, an automated and unbiased user-friendly pipeline for segmentation, tracking, and feature extraction of diverse intracellular structures. Nellie adapts to image metadata, eliminating user input. Nellie's preprocessing pipeline enhances structural contrast on multiple intracellular scales allowing for robust hierarchical segmentation of sub-organellar regions. Internal motion capture markers are generated and tracked via a radius-adaptive pattern matching scheme, and used as guides for sub-voxel flow interpolation. Nellie extracts a plethora of features at multiple hierarchical levels for deep and customizable analysis. Nellie features a point-and-click Napari-based GUI that allows for code-free operation and visualization, while its modular open-source codebase invites extension by experienced users. We demonstrate Nellie's wide variety of use cases with three examples: unmixing multiple organelles from a single channel using feature-based classification, training an unsupervised graph autoencoder on mitochondrial multi-mesh graphs to quantify latent space embedding changes following ionomycin treatment, and performing in-depth characterization and comparison of endoplasmic reticulum networks across different cell types and temporal frames.
comment: 134 pages, 6 main figures, 6 extended figures, 8 supplementary figures; for associated code, see https://github.com/aelefebv/nellie
♻ ☆ Exploring Invariance in Images through One-way Wave Equations
In this paper, we empirically reveal an invariance over images-images share a set of one-way wave equations with latent speeds. Each image is uniquely associated with a solution to these wave equations, allowing for its reconstruction with high fidelity from an initial condition. We demonstrate it using an intuitive encoder-decoder framework where each image is encoded into its corresponding initial condition (a single vector). Subsequently, the initial condition undergoes a specialized decoder, transforming the one-way wave equations into a first-order norm + linear autoregressive process. This process propagates the initial condition along the x and y directions, generating a high-resolution feature map (up to the image resolution), followed by a few convolutional layers to reconstruct image pixels. The revealed invariance, rooted in the shared wave equations, offers a fresh perspective for comprehending images, establishing a promising avenue for further exploration.
comment: This is an improvement version, which fuses some parts from the preliminary work arXiv:2305.16319
♻ ☆ Towards Rationality in Language and Multimodal Agents: A Survey
Rationality is the quality of being guided by reason, characterized by decision-making that aligns with evidence and logical principles. It plays a crucial role in reliable problem-solving by ensuring well-grounded and consistent solutions. While large language models (LLMs) have made significant progress in generating human-like text, they still exhibit limitations such as bounded knowledge space and inconsistent outputs. In response, recent efforts have shifted toward developing multimodal and multi-agent systems, as well as integrating modules like external tools, programming codes, symbolic reasoners, utility function, and conformal risk controls rather than relying solely on a single LLM for decision-making. This paper surveys the state-of-the-art advancements in language and multimodal agents, evaluates how they contribute to make intelligent agents more rational, and identifies open challenges and future research directions. We maintain an open repository at https://github.com/bowen-upenn/Agent_Rationality.
comment: We maintain an open repository at https://github.com/bowen-upenn/Agent_Rationality
♻ ☆ On the power of data augmentation for head pose estimation CVPR
Deep learning has been impressively successful in the last decade in predicting human head poses from monocular images. However, for in-the-wild inputs the research community relies predominantly on a single training set, 300W-LP, of semisynthetic nature without many alternatives. This paper focuses on gradual extension and improvement of the data to explore the performance achievable with augmentation and synthesis strategies further. Modeling-wise a novel multitask head/loss design which includes uncertainty estimation is proposed. Overall, the thus obtained models are small, efficient, suitable for full 6 DoF pose estimation, and exhibit very competitive accuracy.
comment: CVPR version. Added evaluation on BIWI. Plenty of writing changes
♻ ☆ Vision transformers in domain adaptation and domain generalization: a study of robustness
Deep learning models are often evaluated in scenarios where the data distribution is different from those used in the training and validation phases. The discrepancy presents a challenge for accurately predicting the performance of models once deployed on the target distribution. Domain adaptation and generalization are widely recognized as effective strategies for addressing such shifts, thereby ensuring reliable performance. The recent promising results in applying vision transformers in computer vision tasks, coupled with advancements in self-attention mechanisms, have demonstrated their significant potential for robustness and generalization in handling distribution shifts. Motivated by the increased interest from the research community, our paper investigates the deployment of vision transformers in domain adaptation and domain generalization scenarios. For domain adaptation methods, we categorize research into feature-level, instance-level, model-level adaptations, and hybrid approaches, along with other categorizations with respect to diverse strategies for enhancing domain adaptation. Similarly, for domain generalization, we categorize research into multi-domain learning, meta-learning, regularization techniques, and data augmentation strategies. We further classify diverse strategies in research, underscoring the various approaches researchers have taken to address distribution shifts by integrating vision transformers. The inclusion of comprehensive tables summarizing these categories is a distinct feature of our work, offering valuable insights for researchers. These findings highlight the versatility of vision transformers in managing distribution shifts, crucial for real-world applications, especially in critical safety and decision-making scenarios.
♻ ☆ How to Determine the Preferred Image Distribution of a Black-Box Vision-Language Model? NeurIPS 2024
Large foundation models have revolutionized the field, yet challenges remain in optimizing multi-modal models for specialized visual tasks. We propose a novel, generalizable methodology to identify preferred image distributions for black-box Vision-Language Models (VLMs) by measuring output consistency across varied input prompts. Applying this to different rendering types of 3D objects, we demonstrate its efficacy across various domains requiring precise interpretation of complex structures, with a focus on Computer-Aided Design (CAD) as an exemplar field. We further refine VLM outputs using in-context learning with human feedback, significantly enhancing explanation quality. To address the lack of benchmarks in specialized domains, we introduce CAD-VQA, a new dataset for evaluating VLMs on CAD-related visual question answering tasks. Our evaluation of state-of-the-art VLMs on CAD-VQA establishes baseline performance levels, providing a framework for advancing VLM capabilities in complex visual reasoning tasks across various fields requiring expert-level visual interpretation. We release the dataset and evaluation codes at \url{https://github.com/asgsaeid/cad_vqa}.
comment: Accepted to NeurIPS 2024, Safe Generative AI
♻ ☆ Medical Graph RAG: Towards Safe Medical Large Language Model via Graph Retrieval-Augmented Generation
We introduce a novel graph-based Retrieval-Augmented Generation (RAG) framework specifically designed for the medical domain, called \textbf{MedGraphRAG}, aimed at enhancing Large Language Model (LLM) capabilities for generating evidence-based medical responses, thereby improving safety and reliability when handling private medical data. Graph-based RAG (GraphRAG) leverages LLMs to organize RAG data into graphs, showing strong potential for gaining holistic insights from long-form documents. However, its standard implementation is overly complex for general use and lacks the ability to generate evidence-based responses, limiting its effectiveness in the medical field. To extend the capabilities of GraphRAG to the medical domain, we propose unique Triple Graph Construction and U-Retrieval techniques over it. In our graph construction, we create a triple-linked structure that connects user documents to credible medical sources and controlled vocabularies. In the retrieval process, we propose U-Retrieval which combines Top-down Precise Retrieval with Bottom-up Response Refinement to balance global context awareness with precise indexing. These effort enable both source information retrieval and comprehensive response generation. Our approach is validated on 9 medical Q\&A benchmarks, 2 health fact-checking benchmarks, and one collected dataset testing long-form generation. The results show that MedGraphRAG consistently outperforms state-of-the-art models across all benchmarks, while also ensuring that responses include credible source documentation and definitions. Our code is released at: https://github.com/MedicineToken/Medical-Graph-RAG.
Information Retrieval 18
☆ GaVaMoE: Gaussian-Variational Gated Mixture of Experts for Explainable Recommendation
Large language model-based explainable recommendation (LLM-based ER) systems show promise in generating human-like explanations for recommendations. However, they face challenges in modeling user-item collaborative preferences, personalizing explanations, and handling sparse user-item interactions. To address these issues, we propose GaVaMoE, a novel Gaussian-Variational Gated Mixture of Experts framework for explainable recommendation. GaVaMoE introduces two key components: (1) a rating reconstruction module that employs Variational Autoencoder (VAE) with a Gaussian Mixture Model (GMM) to capture complex user-item collaborative preferences, serving as a pre-trained multi-gating mechanism; and (2) a set of fine-grained expert models coupled with the multi-gating mechanism for generating highly personalized explanations. The VAE component models latent factors in user-item interactions, while the GMM clusters users with similar behaviors. Each cluster corresponds to a gate in the multi-gating mechanism, routing user-item pairs to appropriate expert models. This architecture enables GaVaMoE to generate tailored explanations for specific user types and preferences, mitigating data sparsity by leveraging user similarities. Extensive experiments on three real-world datasets demonstrate that GaVaMoE significantly outperforms existing methods in explanation quality, personalization, and consistency. Notably, GaVaMoE exhibits robust performance in scenarios with sparse user-item interactions, maintaining high-quality explanations even for users with limited historical data.
☆ Adaptive Coordinators and Prompts on Heterogeneous Graphs for Cross-Domain Recommendations
In the online digital world, users frequently engage with diverse items across multiple domains (e.g., e-commerce platforms, streaming services, and social media networks), forming complex heterogeneous interaction graphs. Leveraging this multi-domain information can undoubtedly enhance the performance of recommendation systems by providing more comprehensive user insights and alleviating data sparsity in individual domains. However, integrating multi-domain knowledge for the cross-domain recommendation is very hard due to inherent disparities in user behavior and item characteristics and the risk of negative transfer, where irrelevant or conflicting information from the source domains adversely impacts the target domain's performance. To address these challenges, we offer HAGO, a novel framework with $\textbf{H}$eterogeneous $\textbf{A}$daptive $\textbf{G}$raph co$\textbf{O}$rdinators, which dynamically integrate multi-domain graphs into a cohesive structure by adaptively adjusting the connections between coordinators and multi-domain graph nodes, thereby enhancing beneficial inter-domain interactions while mitigating negative transfer effects. Additionally, we develop a universal multi-domain graph pre-training strategy alongside HAGO to collaboratively learn high-quality node representations across domains. To effectively transfer the learned multi-domain knowledge to the target domain, we design an effective graph prompting method, which incorporates pre-trained embeddings with learnable prompts for the recommendation task. Our framework is compatible with various graph-based models and pre-training techniques, demonstrating broad applicability and effectiveness. Further experimental results show that our solutions outperform state-of-the-art methods in multi-domain recommendation scenarios and highlight their potential for real-world applications.
comment: Under review
☆ CoActionGraphRec: Sequential Multi-Interest Recommendations Using Co-Action Graphs
There are unique challenges to developing item recommender systems for e-commerce platforms like eBay due to sparse data and diverse user interests. While rich user-item interactions are important, eBay's data sparsity exceeds other e-commerce sites by an order of magnitude. To address this challenge, we propose CoActionGraphRec (CAGR), a text based two-tower deep learning model (Item Tower and User Tower) utilizing co-action graph layers. In order to enhance user and item representations, a graph-based solution tailored to eBay's environment is utilized. For the Item Tower, we represent each item using its co-action items to capture collaborative signals in a co-action graph that is fully leveraged by the graph neural network component. For the User Tower, we build a fully connected graph of each user's behavior sequence, with edges encoding pairwise relationships. Furthermore, an explicit interaction module learns representations capturing behavior interactions. Extensive offline and online A/B test experiments demonstrate the effectiveness of our proposed approach and results show improved performance over state-of-the-art methods on key metrics.
☆ LR-SQL: A Supervised Fine-Tuning Method for Text2SQL Tasks under Low-Resource Scenarios
Large language models revolutionize Text2SQL through supervised fine-tuning, yet a crucial limitation is overlooked: the complexity of databases leads to an increased context length, consequently resulting in higher GPU memory demands for model fine-tuning. To address this issue, we propose LR-SQL. LR-SQL comprises two supervised fine-tuning models: the schema\_link model and the SQL\_generation model, with the schema\_link model serving as the focal point for streamlining the overall process. During the fine-tuning of the schema\_link model, LR-SQL breaks down the complete database into flexible combinations of tables with adjustable quantities, enabling the model to learn the relationships within the entire database from these dispersed slices. Furthermore, to enhance the model's ability to perceive the relationships among various discrete slices during inference, LR-SQL trains the model's Chain-of-Thought capability for this task. Experimental results demonstrate that LR-SQL can reduce the total GPU memory usage by 40\% compared to existing fine-tuning methods, while only losing 2\% of table prediction accuracy in schema\_link task. For the overall Text2SQL task, the Execution Accuracy decrease by 0.6\%.Our project is now available on https://github.com/hongWin/LR-SQL
comment: 12pages, 4 figures,submitting to a journal
☆ Enhance Graph Alignment for Large Language Models
Graph-structured data is prevalent in the real world. Recently, due to the powerful emergent capabilities, Large Language Models (LLMs) have shown promising performance in modeling graphs. The key to effectively applying LLMs on graphs is converting graph data into a format LLMs can comprehend. Graph-to-token approaches are popular in enabling LLMs to process graph information. They transform graphs into sequences of tokens and align them with text tokens through instruction tuning, where self-supervised instruction tuning helps LLMs acquire general knowledge about graphs, and supervised fine-tuning specializes LLMs for the downstream tasks on graphs. Despite their initial success, we find that existing methods have a misalignment between self-supervised tasks and supervised downstream tasks, resulting in negative transfer from self-supervised fine-tuning to downstream tasks. To address these issues, we propose Graph Alignment Large Language Models (GALLM) to benefit from aligned task templates. In the self-supervised tuning stage, we introduce a novel text matching task using templates aligned with downstream tasks. In the task-specific tuning stage, we propose two category prompt methods that learn supervision information from additional explanation with further aligned templates. Experimental evaluations on four datasets demonstrate substantial improvements in supervised learning, multi-dataset generalizability, and particularly in zero-shot capability, highlighting the model's potential as a graph foundation model.
comment: Under review
☆ Reducing Labeling Costs in Sentiment Analysis via Semi-Supervised Learning
Labeling datasets is a noteworthy challenge in machine learning, both in terms of cost and time. This research, however, leverages an efficient answer. By exploring label propagation in semi-supervised learning, we can significantly reduce the number of labels required compared to traditional methods. We employ a transductive label propagation method based on the manifold assumption for text classification. Our approach utilizes a graph-based method to generate pseudo-labels for unlabeled data for the text classification task, which are then used to train deep neural networks. By extending labels based on cosine proximity within a nearest neighbor graph from network embeddings, we combine unlabeled data into supervised learning, thereby reducing labeling costs. Based on previous successes in other domains, this study builds and evaluates this approach's effectiveness in sentiment analysis, presenting insights into semi-supervised learning.
comment: 12 pages, 7 figures, accepted at the 2024 8th International Conference on Natural Language Processing and Information Retrieval (NLPIR 2024), Okayama, Japan, 2024
☆ Sequential LLM Framework for Fashion Recommendation
The fashion industry is one of the leading domains in the global e-commerce sector, prompting major online retailers to employ recommendation systems for product suggestions and customer convenience. While recommendation systems have been widely studied, most are designed for general e-commerce problems and struggle with the unique challenges of the fashion domain. To address these issues, we propose a sequential fashion recommendation framework that leverages a pre-trained large language model (LLM) enhanced with recommendation-specific prompts. Our framework employs parameter-efficient fine-tuning with extensive fashion data and introduces a novel mix-up-based retrieval technique for translating text into relevant product suggestions. Extensive experiments show our proposed framework significantly enhances fashion recommendation performance.
☆ On the Capacity of Citation Generation by Large Language Models
Retrieval-augmented generation (RAG) appears as a promising method to alleviate the "hallucination" problem in large language models (LLMs), since it can incorporate external traceable resources for response generation. The essence of RAG in combating the hallucination issue lies in accurately attributing claims in responses to the corresponding retrieved documents. However, most of existing works focus on improving the quality of generated responses from the LLM, while largely overlooked its ability to attribute sources accurately. In this study, we conduct a systematic analysis about the capabilities of LLMs in generating citations within response generation, and further introduce a novel method to enhance their citation generation abilities. Specifically, we evaluate both the correctness and citation quality for seven widely-used LLMs on two benchmark datasets. Meanwhile, we introduce new citation evaluation metrics to eliminate the over-penalization of unnecessary and excessive citations in existing metrics. Furthermore, we propose a Generate-then-Refine method that completes relevant citations and removes irrelevant ones without altering the response text. The results on WebGLM-QA, ASQA and ELI5 datasets show that our method substantially improves the quality of citations in responses generated by LLMs.
comment: Accepted by CCIR 2024
☆ Optimizing Encoder-Only Transformers for Session-Based Recommendation Systems
Session-based recommendation is the task of predicting the next item a user will interact with, often without access to historical user data. In this work, we introduce Sequential Masked Modeling, a novel approach for encoder-only transformer architectures to tackle the challenges of single-session recommendation. Our method combines data augmentation through window sliding with a unique penultimate token masking strategy to capture sequential dependencies more effectively. By enhancing how transformers handle session data, Sequential Masked Modeling significantly improves next-item prediction performance. We evaluate our approach on three widely-used datasets, Yoochoose 1/64, Diginetica, and Tmall, comparing it to state-of-the-art single-session, cross-session, and multi-relation approaches. The results demonstrate that our Transformer-SMM models consistently outperform all models that rely on the same amount of information, while even rivaling methods that have access to more extensive user history. This study highlights the potential of encoder-only transformers in session-based recommendation and opens the door for further improvements.
☆ The Moral Case for Using Language Model Agents for Recommendation
Our information and communication environment has fallen short of the ideals that networked global communication might have served. Identifying all the causes of its pathologies is difficult, but existing recommender systems very likely play a contributing role. In this paper, which draws on the normative tools of philosophy of computing, informed by empirical and technical insights from natural language processing and recommender systems, we make the moral case for an alternative approach. We argue that existing recommenders incentivise mass surveillance, concentrate power, fall prey to narrow behaviourism, and compromise user agency. Rather than just trying to avoid algorithms entirely, or to make incremental improvements to the current paradigm, researchers and engineers should explore an alternative paradigm: the use of language model (LM) agents to source and curate content that matches users' preferences and values, expressed in natural language. The use of LM agents for recommendation poses its own challenges, including those related to candidate generation, computational efficiency, preference modelling, and prompt injection. Nonetheless, if implemented successfully LM agents could: guide us through the digital public sphere without relying on mass surveillance; shift power away from platforms towards users; optimise for what matters instead of just for behavioural proxies; and scaffold our agency instead of undermining it.
♻ ☆ GOVERN: Gradient Orientation Vote Ensemble for Multi-Teacher Reinforced Distillation EMNLP 2024
Pre-trained language models have become an integral component of question-answering systems, achieving remarkable performance. However, for practical deployment, it is crucial to perform knowledge distillation to maintain high performance while operating under computational constraints. In this paper, we address a key question: given the importance of unsupervised distillation for student model performance, how can knowledge from multiple teacher models be effectively ensemble during this stage without the guidance of labels? We propose a novel algorithm, GOVERN, to tackle this issue. GOVERN has demonstrated significant improvements in both offline and online experiments, enabling the student model to achieve results comparable to that of teacher ensembles. Our experiments show that GOVERN remarkably requires a mere 1\% of the ensemble method's inference budget to achieve 99.5\% of performance. The proposed algorithm has been successfully deployed in a real-world commercial question-answering system, demonstrating its real-world applicability.
comment: Accepted by EMNLP 2024 Industry Track
♻ ☆ DISCO: A Hierarchical Disentangled Cognitive Diagnosis Framework for Interpretable Job Recommendation ICDM 2024
The rapid development of online recruitment platforms has created unprecedented opportunities for job seekers while concurrently posing the significant challenge of quickly and accurately pinpointing positions that align with their skills and preferences. Job recommendation systems have significantly alleviated the extensive search burden for job seekers by optimizing user engagement metrics, such as clicks and applications, thus achieving notable success. In recent years, a substantial amount of research has been devoted to developing effective job recommendation models, primarily focusing on text-matching based and behavior modeling based methods. While these approaches have realized impressive outcomes, it is imperative to note that research on the explainability of recruitment recommendations remains profoundly unexplored. To this end, in this paper, we propose DISCO, a hierarchical Disentanglement based Cognitive diagnosis framework, aimed at flexibly accommodating the underlying representation learning model for effective and interpretable job recommendations. Specifically, we first design a hierarchical representation disentangling module to explicitly mine the hierarchical skill-related factors implied in hidden representations of job seekers and jobs. Subsequently, we propose level-aware association modeling to enhance information communication and robust representation learning both inter- and intra-level, which consists of the interlevel knowledge influence module and the level-wise contrastive learning. Finally, we devise an interaction diagnosis module incorporating a neural diagnosis function for effectively modeling the multi-level recruitment interaction process between job seekers and jobs, which introduces the cognitive measurement theory.
comment: Accepted by ICDM 2024. 10 pages
♻ ☆ Curriculum-scheduled Knowledge Distillation from Multiple Pre-trained Teachers for Multi-domain Sequential Recommendation
Pre-trained recommendation models (PRMs) have received increasing interest recently. However, their intrinsically heterogeneous model structure, huge model size and computation cost hinder their adoptions in practical recommender systems. Hence, it is highly essential to explore how to use different pre-trained recommendation models efficiently in real-world systems. In this paper, we propose a novel curriculum-scheduled knowledge distillation from multiple pre-trained teachers for multi-domain sequential recommendation, called CKD-MDSR, which takes full advantages of different PRMs as multiple teacher models to boost a small student recommendation model, integrating the knowledge across multiple domains from PRMs. Specifically, CKD-MDSR first adopts curriculum-scheduled user behavior sequence sampling and distills informative knowledge jointly from the representative PRMs such as UniSRec and Recformer. Then, the knowledge from the above PRMs are selectively integrated into the student model in consideration of their confidence and consistency. Finally, we verify the proposed method on multi-domain sequential recommendation and further demonstrate its universality with multiple types of student models, including feature interaction and graph based recommendation models. Extensive experiments on five real-world datasets demonstrate the effectiveness and efficiency of CKD-MDSR, which can be viewed as an efficient shortcut using PRMs in real-world systems.
♻ ☆ DIRAS: Efficient LLM Annotation of Document Relevance in Retrieval Augmented Generation
Retrieval Augmented Generation (RAG) is widely employed to ground responses to queries on domain-specific documents. But do RAG implementations leave out important information when answering queries that need an integrated analysis of information (e.g., Tell me good news in the stock market today.)? To address these concerns, RAG developers need to annotate information retrieval (IR) data for their domain of interest, which is challenging because (1) domain-specific queries usually need nuanced definitions of relevance beyond shallow semantic relevance; and (2) human or GPT-4 annotation is costly and cannot cover all (query, document) pairs (i.e., annotation selection bias), thus harming the effectiveness in evaluating IR recall. To address these challenges, we propose DIRAS (Domain-specific Information Retrieval Annotation with Scalability), a manual-annotation-free schema that fine-tunes open-sourced LLMs to consider nuanced relevance definition and annotate (partial) relevance labels with calibrated relevance scores. Extensive evaluation shows that DIRAS enables smaller (8B) LLMs to achieve GPT-4-level performance on annotating and ranking unseen (query, document) pairs, and is helpful for real-world RAG development. All code, LLM generations, and human annotations can be found in \url{https://github.com/EdisonNi-hku/DIRAS}.
♻ ☆ CoRA: Collaborative Information Perception by Large Language Model's Weights for Recommendation
Involving collaborative information in Large Language Models (LLMs) is a promising technique for adapting LLMs for recommendation. Existing methods achieve this by concatenating collaborative features with text tokens into a unified sequence input and then fine-tuning to align these features with LLM's input space. Although effective, in this work, we identify two limitations when adapting LLMs to recommendation tasks, which hinder the integration of general knowledge and collaborative information, resulting in sub-optimal recommendation performance. (1) Fine-tuning LLM with recommendation data can undermine its inherent world knowledge and fundamental competencies, which are crucial for interpreting and inferring recommendation text. (2) Incorporating collaborative features into textual prompts disrupts the semantics of the original prompts, preventing LLM from generating appropriate outputs. In this paper, we propose a new paradigm, CoRA (an acronym for Collaborative LoRA), with a collaborative weights generator. Rather than input space alignment, this method aligns collaborative information with LLM's parameter space, representing them as incremental weights to update LLM's output. This way, LLM perceives collaborative information without altering its general knowledge and text inference capabilities. Specifically, we employ a collaborative filtering model to extract user and item embeddings, converting them into collaborative weights with low-rank properties through the collaborative weights generator. We then merge the collaborative weights into LLM's weights, enabling LLM to perceive the collaborative signals and generate personalized recommendations without fine-tuning or extra collaborative tokens in prompts. Extensive experiments confirm that CoRA effectively integrates collaborative information into LLM, enhancing recommendation performance.
♻ ☆ Recommenadation aided Caching using Combinatorial Multi-armed Bandits
We study content caching with recommendations in a wireless network where the users are connected through a base station equipped with a finite-capacity cache. We assume a fixed set of contents with unknown user preferences and content popularities. The base station can cache a subset of the contents and can also recommend subsets of the contents to different users in order to encourage them to request the recommended contents. Recommendations, depending on their acceptability, can thus be used to increase cache hits. We first assume that the users' recommendation acceptabilities are known and formulate the cache hit optimization problem as a combinatorial multi-armed bandit (CMAB). We propose a UCB-based algorithm to decide which contents to cache and recommend and provide an upper bound on the regret of this algorithm. Subsequently, we consider a more general scenario where the users' recommendation acceptabilities are also unknown and propose another UCB-based algorithm that learns these as well. We numerically demonstrate the performance of our algorithms and compare these to state-of-the-art algorithms.
♻ ☆ Spectral-Based Graph Neural Networks for Complementary Item Recommendation AAAI-24
Modeling complementary relationships greatly helps recommender systems to accurately and promptly recommend the subsequent items when one item is purchased. Unlike traditional similar relationships, items with complementary relationships may be purchased successively (such as iPhone and Airpods Pro), and they not only share relevance but also exhibit dissimilarity. Since the two attributes are opposites, modeling complementary relationships is challenging. Previous attempts to exploit these relationships have either ignored or oversimplified the dissimilarity attribute, resulting in ineffective modeling and an inability to balance the two attributes. Since Graph Neural Networks (GNNs) can capture the relevance and dissimilarity between nodes in the spectral domain, we can leverage spectral-based GNNs to effectively understand and model complementary relationships. In this study, we present a novel approach called Spectral-based Complementary Graph Neural Networks (SComGNN) that utilizes the spectral properties of complementary item graphs. We make the first observation that complementary relationships consist of low-frequency and mid-frequency components, corresponding to the relevance and dissimilarity attributes, respectively. Based on this spectral observation, we design spectral graph convolutional networks with low-pass and mid-pass filters to capture the low-frequency and mid-frequency components. Additionally, we propose a two-stage attention mechanism to adaptively integrate and balance the two attributes. Experimental results on four e-commerce datasets demonstrate the effectiveness of our model, with SComGNN significantly outperforming existing baseline models.
comment: Accepted by AAAI-24
♻ ☆ DyVo: Dynamic Vocabularies for Learned Sparse Retrieval with Entities
Learned Sparse Retrieval (LSR) models use vocabularies from pre-trained transformers, which often split entities into nonsensical fragments. Splitting entities can reduce retrieval accuracy and limits the model's ability to incorporate up-to-date world knowledge not included in the training data. In this work, we enhance the LSR vocabulary with Wikipedia concepts and entities, enabling the model to resolve ambiguities more effectively and stay current with evolving knowledge. Central to our approach is a Dynamic Vocabulary (DyVo) head, which leverages existing entity embeddings and an entity retrieval component that identifies entities relevant to a query or document. We use the DyVo head to generate entity weights, which are then merged with word piece weights to create joint representations for efficient indexing and retrieval using an inverted index. In experiments across three entity-rich document ranking datasets, the resulting DyVo model substantially outperforms state-of-the-art baselines.
comment: https://github.com/thongnt99/DyVo
Machine Learning 200
☆ MoH: Multi-Head Attention as Mixture-of-Head Attention
In this work, we upgrade the multi-head attention mechanism, the core of the Transformer model, to improve efficiency while maintaining or surpassing the previous accuracy level. We show that multi-head attention can be expressed in the summation form. Drawing on the insight that not all attention heads hold equal significance, we propose Mixture-of-Head attention (MoH), a new architecture that treats attention heads as experts in the Mixture-of-Experts (MoE) mechanism. MoH has two significant advantages: First, MoH enables each token to select the appropriate attention heads, enhancing inference efficiency without compromising accuracy or increasing the number of parameters. Second, MoH replaces the standard summation in multi-head attention with a weighted summation, introducing flexibility to the attention mechanism and unlocking extra performance potential. Extensive experiments on ViT, DiT, and LLMs demonstrate that MoH outperforms multi-head attention by using only 50%-90% of the attention heads. Moreover, we demonstrate that pre-trained multi-head attention models, such as LLaMA3-8B, can be further continue-tuned into our MoH models. Notably, MoH-LLaMA3-8B achieves an average accuracy of 64.0% across 14 benchmarks, outperforming LLaMA3-8B by 2.4% by utilizing only 75% of the attention heads. We believe the proposed MoH is a promising alternative to multi-head attention and provides a strong foundation for developing advanced and efficient attention-based models.
comment: 23 pages, code: https://github.com/SkyworkAI/MoH
☆ A Hitchhiker's Guide to Scaling Law Estimation
Scaling laws predict the loss of a target machine learning model by extrapolating from easier-to-train models with fewer parameters or smaller training sets. This provides an efficient way for practitioners and researchers alike to compare pretraining decisions involving optimizers, datasets, and model architectures. Despite the widespread use of scaling laws to model the dynamics of language model training, there has been little work on understanding how to best estimate and interpret them. We collect (and release) a large-scale dataset containing losses and downstream evaluations for 485 previously published pretrained models. We use these to estimate more than 1000 scaling laws, then derive a set of best practices for estimating scaling laws in new model families. We find that fitting scaling laws to intermediate checkpoints of training runs (and not just their final losses) substantially improves accuracy, and that -- all else equal -- estimates of performance are generally most accurate when derived from other models of similar sizes. However, because there is a significant degree of variability across model seeds, training multiple small models is sometimes more useful than training a single large one. Moreover, while different model families differ scaling behavior, they are often similar enough that a target model's behavior can be predicted from a single model with the same architecture, along with scaling parameter estimates derived from other model families.
☆ Mitigating Suboptimality of Deterministic Policy Gradients in Complex Q-functions
In reinforcement learning, off-policy actor-critic approaches like DDPG and TD3 are based on the deterministic policy gradient. Herein, the Q-function is trained from off-policy environment data and the actor (policy) is trained to maximize the Q-function via gradient ascent. We observe that in complex tasks like dexterous manipulation and restricted locomotion, the Q-value is a complex function of action, having several local optima or discontinuities. This poses a challenge for gradient ascent to traverse and makes the actor prone to get stuck at local optima. To address this, we introduce a new actor architecture that combines two simple insights: (i) use multiple actors and evaluate the Q-value maximizing action, and (ii) learn surrogates to the Q-function that are simpler to optimize with gradient-based methods. We evaluate tasks such as restricted locomotion, dexterous manipulation, and large discrete-action space recommender systems and show that our actor finds optimal actions more frequently and outperforms alternate actor architectures.
☆ Bayesian Experimental Design via Contrastive Diffusions
Bayesian Optimal Experimental Design (BOED) is a powerful tool to reduce the cost of running a sequence of experiments. When based on the Expected Information Gain (EIG), design optimization corresponds to the maximization of some intractable expected {\it contrast} between prior and posterior distributions. Scaling this maximization to high dimensional and complex settings has been an issue due to BOED inherent computational complexity. In this work, we introduce an {\it expected posterior} distribution with cost-effective sampling properties and provide a tractable access to the EIG contrast maximization via a new EIG gradient expression. Diffusion-based samplers are used to compute the dynamics of the expected posterior and ideas from bi-level optimization are leveraged to derive an efficient joint sampling-optimization loop, without resorting to lower bound approximations of the EIG. The resulting efficiency gain allows to extend BOED to the well-tested generative capabilities of diffusion models. By incorporating generative models into the BOED framework, we expand its scope and its use in scenarios that were previously impractical. Numerical experiments and comparison with state-of-the-art methods show the potential of the approach.
☆ Adaptive Data Optimization: Dynamic Sample Selection with Scaling Laws
The composition of pretraining data is a key determinant of foundation models' performance, but there is no standard guideline for allocating a limited computational budget across different data sources. Most current approaches either rely on extensive experiments with smaller models or dynamic data adjustments that also require proxy models, both of which significantly increase the workflow complexity and computational overhead. In this paper, we introduce Adaptive Data Optimization (ADO), an algorithm that optimizes data distributions in an online fashion, concurrent with model training. Unlike existing techniques, ADO does not require external knowledge, proxy models, or modifications to the model update. Instead, ADO uses per-domain scaling laws to estimate the learning potential of each domain during training and adjusts the data mixture accordingly, making it more scalable and easier to integrate. Experiments demonstrate that ADO can achieve comparable or better performance than prior methods while maintaining computational efficiency across different computation scales, offering a practical solution for dynamically adjusting data distribution without sacrificing flexibility or increasing costs. Beyond its practical benefits, ADO also provides a new perspective on data collection strategies via scaling laws.
comment: 21 pages, 10 figures
☆ Improving Long-Text Alignment for Text-to-Image Diffusion Models
The rapid advancement of text-to-image (T2I) diffusion models has enabled them to generate unprecedented results from given texts. However, as text inputs become longer, existing encoding methods like CLIP face limitations, and aligning the generated images with long texts becomes challenging. To tackle these issues, we propose LongAlign, which includes a segment-level encoding method for processing long texts and a decomposed preference optimization method for effective alignment training. For segment-level encoding, long texts are divided into multiple segments and processed separately. This method overcomes the maximum input length limits of pretrained encoding models. For preference optimization, we provide decomposed CLIP-based preference models to fine-tune diffusion models. Specifically, to utilize CLIP-based preference models for T2I alignment, we delve into their scoring mechanisms and find that the preference scores can be decomposed into two components: a text-relevant part that measures T2I alignment and a text-irrelevant part that assesses other visual aspects of human preference. Additionally, we find that the text-irrelevant part contributes to a common overfitting problem during fine-tuning. To address this, we propose a reweighting strategy that assigns different weights to these two components, thereby reducing overfitting and enhancing alignment. After fine-tuning $512 \times 512$ Stable Diffusion (SD) v1.5 for about 20 hours using our method, the fine-tuned SD outperforms stronger foundation models in T2I alignment, such as PixArt-$\alpha$ and Kandinsky v2.2. The code is available at https://github.com/luping-liu/LongAlign.
☆ Regional Ocean Forecasting with Hierarchical Graph Neural Networks NeurIPS 2024
Accurate ocean forecasting systems are vital for understanding marine dynamics, which play a crucial role in environmental management and climate adaptation strategies. Traditional numerical solvers, while effective, are computationally expensive and time-consuming. Recent advancements in machine learning have revolutionized weather forecasting, offering fast and energy-efficient alternatives. Building on these advancements, we introduce SeaCast, a neural network designed for high-resolution, medium-range ocean forecasting. SeaCast employs a graph-based framework to effectively handle the complex geometry of ocean grids and integrates external forcing data tailored to the regional ocean context. Our approach is validated through experiments at a high spatial resolution using the operational numerical model of the Mediterranean Sea provided by the Copernicus Marine Service, along with both numerical and data-driven atmospheric forcings.
comment: 28 pages, 35 figures. Accepted to the Tackling Climate Change with Machine Learning workshop at NeurIPS 2024
☆ FoundTS: Comprehensive and Unified Benchmarking of Foundation Models for Time Series Forecasting
Time Series Forecasting (TSF) is key functionality in numerous fields, including in finance, weather services, and energy management. While TSF methods are emerging these days, many of them require domain-specific data collection and model training and struggle with poor generalization performance on new domains. Foundation models aim to overcome this limitation. Pre-trained on large-scale language or time series data, they exhibit promising inferencing capabilities in new or unseen data. This has spurred a surge in new TSF foundation models. We propose a new benchmark, FoundTS, to enable thorough and fair evaluation and comparison of such models. FoundTS covers a variety of TSF foundation models, including those based on large language models and those pretrained on time series. Next, FoundTS supports different forecasting strategies, including zero-shot, few-shot, and full-shot, thereby facilitating more thorough evaluations. Finally, FoundTS offers a pipeline that standardizes evaluation processes such as dataset splitting, loading, normalization, and few-shot sampling, thereby facilitating fair evaluations. Building on this, we report on an extensive evaluation of TSF foundation models on a broad range of datasets from diverse domains and with different statistical characteristics. Specifically, we identify pros and cons and inherent limitations of existing foundation models, and we identify directions for future model design. We make our code and datasets available at https://anonymous.4open.science/r/FoundTS-C2B0.
☆ OKAMI: Teaching Humanoid Robots Manipulation Skills through Single Video Imitation
We study the problem of teaching humanoid robots manipulation skills by imitating from single video demonstrations. We introduce OKAMI, a method that generates a manipulation plan from a single RGB-D video and derives a policy for execution. At the heart of our approach is object-aware retargeting, which enables the humanoid robot to mimic the human motions in an RGB-D video while adjusting to different object locations during deployment. OKAMI uses open-world vision models to identify task-relevant objects and retarget the body motions and hand poses separately. Our experiments show that OKAMI achieves strong generalizations across varying visual and spatial conditions, outperforming the state-of-the-art baseline on open-world imitation from observation. Furthermore, OKAMI rollout trajectories are leveraged to train closed-loop visuomotor policies, which achieve an average success rate of 79.2% without the need for labor-intensive teleoperation. More videos can be found on our website https://ut-austin-rpl.github.io/OKAMI/.
comment: Accepted for oral presentation at 8th Annual Conference on Robot Learning. Project website: https://ut-austin-rpl.github.io/OKAMI/
☆ Selection-p: Self-Supervised Task-Agnostic Prompt Compression for Faithfulness and Transferability EMNLP 2024
Large Language Models (LLMs) have demonstrated impressive capabilities in a wide range of natural language processing tasks when leveraging in-context learning. To mitigate the additional computational and financial costs associated with in-context learning, several prompt compression methods have been proposed to compress the in-context learning prompts. Despite their success, these methods face challenges with transferability due to model-specific compression, or rely on external training data, such as GPT-4. In this paper, we investigate the ability of LLMs to develop a unified compression method that discretizes uninformative tokens, utilizing a self-supervised pre-training technique. By introducing a small number of parameters during the continual pre-training, the proposed Selection-p produces a probability for each input token, indicating whether to preserve or discard it. Experiments show Selection-p achieves state-of-the-art performance across numerous classification tasks, achieving compression rates of up to 10 times while experiencing only a marginal 0.8% decrease in performance. Moreover, it exhibits superior transferability to different models compared to prior work. Additionally, we further analyze how Selection-p helps maintain performance on in-context learning with long contexts.
comment: 14 pages, 5 figures, 10 tables, EMNLP 2024 Findings
☆ G-Designer: Architecting Multi-agent Communication Topologies via Graph Neural Networks
Recent advancements in large language model (LLM)-based agents have demonstrated that collective intelligence can significantly surpass the capabilities of individual agents, primarily due to well-crafted inter-agent communication topologies. Despite the diverse and high-performing designs available, practitioners often face confusion when selecting the most effective pipeline for their specific task: \textit{Which topology is the best choice for my task, avoiding unnecessary communication token overhead while ensuring high-quality solution?} In response to this dilemma, we introduce G-Designer, an adaptive, efficient, and robust solution for multi-agent deployment, which dynamically designs task-aware, customized communication topologies. Specifically, G-Designer models the multi-agent system as a multi-agent network, leveraging a variational graph auto-encoder to encode both the nodes (agents) and a task-specific virtual node, and decodes a task-adaptive and high-performing communication topology. Extensive experiments on six benchmarks showcase that G-Designer is: \textbf{(1) high-performing}, achieving superior results on MMLU with accuracy at $84.50\%$ and on HumanEval with pass@1 at $89.90\%$; \textbf{(2) task-adaptive}, architecting communication protocols tailored to task difficulty, reducing token consumption by up to $95.33\%$ on HumanEval; and \textbf{(3) adversarially robust}, defending against agent adversarial attacks with merely $0.3\%$ accuracy drop.
☆ Language Models Encode Numbers Using Digit Representations in Base 10
Large language models (LLMs) frequently make errors when handling even simple numerical problems, such as comparing two small numbers. A natural hypothesis is that these errors stem from how LLMs represent numbers, and specifically, whether their representations of numbers capture their numeric values. We tackle this question from the observation that LLM errors on numerical tasks are often distributed across \textit{the digits} of the answer rather than normally around \textit{its numeric value}. Through a series of probing experiments and causal interventions, we show that LLMs internally represent numbers with individual circular representations per-digit in base 10. This digit-wise representation, as opposed to a value representation, sheds light on the error patterns of models on tasks involving numerical reasoning and could serve as a basis for future studies on analyzing numerical mechanisms in LLMs.
☆ MLLM can see? Dynamic Correction Decoding for Hallucination Mitigation
Multimodal Large Language Models (MLLMs) frequently exhibit hallucination phenomena, but the underlying reasons remain poorly understood. In this paper, we present an empirical analysis and find that, although MLLMs incorrectly generate the objects in the final output, they are actually able to recognize visual objects in the preceding layers. We speculate that this may be due to the strong knowledge priors of the language model suppressing the visual information, leading to hallucinations. Motivated by this, we propose a novel dynamic correction decoding method for MLLMs (DeCo), which adaptively selects the appropriate preceding layers and proportionally integrates knowledge into the final layer to adjust the output logits. Note that DeCo is model agnostic and can be seamlessly incorporated with various classic decoding strategies and applied to different MLLMs. We evaluate DeCo on widely-used benchmarks, demonstrating that it can reduce hallucination rates by a large margin compared to baselines, highlighting its potential to mitigate hallucinations. Code is available at https://github.com/zjunlp/DeCo.
comment: Ongoing work
☆ On the Training Convergence of Transformers for In-Context Classification
While transformers have demonstrated impressive capacities for in-context learning (ICL) in practice, theoretical understanding of the underlying mechanism enabling transformers to perform ICL is still in its infant stage. This work aims to theoretically study the training dynamics of transformers for in-context classification tasks. We demonstrate that, for in-context classification of Gaussian mixtures under certain assumptions, a single-layer transformer trained via gradient descent converges to a globally optimal model at a linear rate. We further quantify the impact of the training and testing prompt lengths on the ICL inference error of the trained transformer. We show that when the lengths of training and testing prompts are sufficiently large, the prediction of the trained transformer approaches the Bayes-optimal classifier. Experimental results corroborate the theoretical findings.
☆ Encoding architecture algebra
Despite the wide variety of input types in machine learning, this diversity is often not fully reflected in their representations or model architectures, leading to inefficiencies throughout a model's lifecycle. This paper introduces an algebraic approach to constructing input-encoding architectures that properly account for the data's structure, providing a step toward achieving more typeful machine learning.
comment: 25 pages, 6 figures. Keywords: typeful, algebraic data types, tensors, structured data
☆ Layer-wise Importance Matters: Less Memory for Better Performance in Parameter-efficient Fine-tuning of Large Language Models EMNLP 2024
Parameter-Efficient Fine-Tuning (PEFT) methods have gained significant popularity for adapting pre-trained Large Language Models (LLMs) to downstream tasks, primarily due to their potential to significantly reduce memory and computational overheads. However, a common limitation in most PEFT approaches is their application of a uniform architectural design across all layers. This uniformity involves identical trainable modules and ignores the varying importance of each layer, leading to sub-optimal fine-tuning results. To overcome the above limitation and obtain better performance, we develop a novel approach, Importance-aware Sparse Tuning (IST), to fully utilize the inherent sparsity and select the most important subset of full layers with effective layer-wise importance scoring. The proposed IST is a versatile and plug-and-play technique compatible with various PEFT methods that operate on a per-layer basis. By leveraging the estimated importance scores, IST dynamically updates these selected layers in PEFT modules, leading to reduced memory demands. We further provide theoretical proof of convergence and empirical evidence of superior performance to demonstrate the advantages of IST over uniform updating strategies. Extensive experiments on a range of LLMs, PEFTs, and downstream tasks substantiate the effectiveness of our proposed method, showcasing IST's capacity to enhance existing layer-based PEFT methods. Our code is available at https://github.com/Kaiseem/IST.
comment: EMNLP 2024
☆ Can Search-Based Testing with Pareto Optimization Effectively Cover Failure-Revealing Test Inputs?
Search-based software testing (SBST) is a widely adopted technique for testing complex systems with large input spaces, such as Deep Learning-enabled (DL-enabled) systems. Many SBST techniques focus on Pareto-based optimization, where multiple objectives are optimized in parallel to reveal failures. However, it is important to ensure that identified failures are spread throughout the entire failure-inducing area of a search domain and not clustered in a sub-region. This ensures that identified failures are semantically diverse and reveal a wide range of underlying causes. In this paper, we present a theoretical argument explaining why testing based on Pareto optimization is inadequate for covering failure-inducing areas within a search domain. We support our argument with empirical results obtained by applying two widely used types of Pareto-based optimization techniques, namely NSGA-II (an evolutionary algorithm) and MOPSO (a swarm-based algorithm), to two DL-enabled systems: an industrial Automated Valet Parking (AVP) system and a system for classifying handwritten digits. We measure the coverage of failure-revealing test inputs in the input space using a metric that we refer to as the Coverage Inverted Distance quality indicator. Our results show that NSGA-II and MOPSO are not more effective than a na\"ive random search baseline in covering test inputs that reveal failures. The replication package for this study is available in a GitHub repository.
comment: Accepted for publication by Empirical Software Engineering Journal (EMSE) (in October 2024)
☆ Analyzing (In)Abilities of SAEs via Formal Languages
Autoencoders have been used for finding interpretable and disentangled features underlying neural network representations in both image and text domains. While the efficacy and pitfalls of such methods are well-studied in vision, there is a lack of corresponding results, both qualitative and quantitative, for the text domain. We aim to address this gap by training sparse autoencoders (SAEs) on a synthetic testbed of formal languages. Specifically, we train SAEs on the hidden representations of models trained on formal languages (Dyck-2, Expr, and English PCFG) under a wide variety of hyperparameter settings, finding interpretable latents often emerge in the features learned by our SAEs. However, similar to vision, we find performance turns out to be highly sensitive to inductive biases of the training pipeline. Moreover, we show latents correlating to certain features of the input do not always induce a causal impact on model's computation. We thus argue that causality has to become a central target in SAE training: learning of causal features should be incentivized from the ground-up. Motivated by this, we propose and perform preliminary investigations for an approach that promotes learning of causally relevant features in our formal language setting.
comment: Under review
☆ ECGN: A Cluster-Aware Approach to Graph Neural Networks for Imbalanced Classification
Classifying nodes in a graph is a common problem. The ideal classifier must adapt to any imbalances in the class distribution. It must also use information in the clustering structure of real-world graphs. Existing Graph Neural Networks (GNNs) have not addressed both problems together. We propose the Enhanced Cluster-aware Graph Network (ECGN), a novel method that addresses these issues by integrating cluster-specific training with synthetic node generation. Unlike traditional GNNs that apply the same node update process for all nodes, ECGN learns different aggregations for different clusters. We also use the clusters to generate new minority-class nodes in a way that helps clarify the inter-class decision boundary. By combining cluster-aware embeddings with a global integration step, ECGN enhances the quality of the resulting node embeddings. Our method works with any underlying GNN and any cluster generation technique. Experimental results show that ECGN consistently outperforms its closest competitors by up to 11% on some widely studied benchmark datasets.
comment: 17 pages, 3 figures
☆ LoSAM: Local Search in Additive Noise Models with Unmeasured Confounders, a Top-Down Global Discovery Approach
We address the challenge of causal discovery in structural equation models with additive noise without imposing additional assumptions on the underlying data-generating process. We introduce local search in additive noise model (LoSAM), which generalizes an existing nonlinear method that leverages local causal substructures to the general additive noise setting, allowing for both linear and nonlinear causal mechanisms. We show that LoSAM achieves polynomial runtime, and improves runtime and efficiency by exploiting new substructures to minimize the conditioning set at each step. Further, we introduce a variant of LoSAM, LoSAM-UC, that is robust to unmeasured confounding among roots, a property that is often not satisfied by functional-causal-model-based methods. We numerically demonstrate the utility of LoSAM, showing that it outperforms existing benchmarks.
☆ Latent Action Pretraining from Videos
We introduce Latent Action Pretraining for general Action models (LAPA), an unsupervised method for pretraining Vision-Language-Action (VLA) models without ground-truth robot action labels. Existing Vision-Language-Action models require action labels typically collected by human teleoperators during pretraining, which significantly limits possible data sources and scale. In this work, we propose a method to learn from internet-scale videos that do not have robot action labels. We first train an action quantization model leveraging VQ-VAE-based objective to learn discrete latent actions between image frames, then pretrain a latent VLA model to predict these latent actions from observations and task descriptions, and finally finetune the VLA on small-scale robot manipulation data to map from latent to robot actions. Experimental results demonstrate that our method significantly outperforms existing techniques that train robot manipulation policies from large-scale videos. Furthermore, it outperforms the state-of-the-art VLA model trained with robotic action labels on real-world manipulation tasks that require language conditioning, generalization to unseen objects, and semantic generalization to unseen instructions. Training only on human manipulation videos also shows positive transfer, opening up the potential for leveraging web-scale data for robotics foundation model.
comment: Website: https://latentactionpretraining.github.io
☆ DySpec: Faster Speculative Decoding with Dynamic Token Tree Structure
While speculative decoding has recently appeared as a promising direction for accelerating the inference of large language models (LLMs), the speedup and scalability are strongly bounded by the token acceptance rate. Prevalent methods usually organize predicted tokens as independent chains or fixed token trees, which fails to generalize to diverse query distributions. In this paper, we propose DySpec, a faster speculative decoding algorithm with a novel dynamic token tree structure. We begin by bridging the draft distribution and acceptance rate from intuitive and empirical clues, and successfully show that the two variables are strongly correlated. Based on this, we employ a greedy strategy to dynamically expand the token tree at run time. Theoretically, we show that our method can achieve optimal results under mild assumptions. Empirically, DySpec yields a higher acceptance rate and speedup than fixed trees. DySpec can drastically improve the throughput and reduce the latency of token generation across various data distribution and model sizes, which significantly outperforms strong competitors, including Specinfer and Sequoia. Under low temperature setting, DySpec can improve the throughput up to 9.1$\times$ and reduce the latency up to 9.4$\times$ on Llama2-70B. Under high temperature setting, DySpec can also improve the throughput up to 6.21$\times$, despite the increasing difficulty of speculating more than one token per step for draft model.
comment: 8 pages, 4 figures
☆ Light-Weight Fault Tolerant Attention for Large Language Model Training
Large Language Models (LLMs) have demonstrated remarkable performance in various natural language processing tasks. However, the training of these models is computationally intensive and susceptible to faults, particularly in the attention mechanism, which is a critical component of transformer-based LLMs. In this paper, we investigate the impact of faults on LLM training, focusing on INF, NaN, and near-INF values in the computation results with systematic fault injection experiments. We observe the propagation patterns of these errors, which can trigger non-trainable states in the model and disrupt training, forcing the procedure to load from checkpoints.To mitigate the impact of these faults, we propose ATTNChecker, the first Algorithm-Based Fault Tolerance (ABFT) technique tailored for the attention mechanism in LLMs. ATTNChecker is designed based on fault propagation patterns of LLM and incorporates performance optimization to adapt to both system reliability and model vulnerability while providing lightweight protection for fast LLM training. Evaluations on four LLMs show that ATTNChecker on average incurs on average 7% overhead on training while detecting and correcting all extreme errors. Compared with the state-of-the-art checkpoint/restore approach, ATTNChecker reduces recovery overhead by up to 49x.
☆ Zero-shot Model-based Reinforcement Learning using Large Language Models
The emerging zero-shot capabilities of Large Language Models (LLMs) have led to their applications in areas extending well beyond natural language processing tasks. In reinforcement learning, while LLMs have been extensively used in text-based environments, their integration with continuous state spaces remains understudied. In this paper, we investigate how pre-trained LLMs can be leveraged to predict in context the dynamics of continuous Markov decision processes. We identify handling multivariate data and incorporating the control signal as key challenges that limit the potential of LLMs' deployment in this setup and propose Disentangled In-Context Learning (DICL) to address them. We present proof-of-concept applications in two reinforcement learning settings: model-based policy evaluation and data-augmented off-policy reinforcement learning, supported by theoretical analysis of the proposed methods. Our experiments further demonstrate that our approach produces well-calibrated uncertainty estimates. We release the code at https://github.com/abenechehab/dicl.
☆ On the potential of Optimal Transport in Geospatial Data Science
Prediction problems in geographic information science and transportation are frequently motivated by the possibility to enhance operational efficiency. Examples range from predicting car sharing demand for optimizing relocation to forecasting traffic congestion for navigation purposes. However, conventional accuracy metrics do not account for the spatial distribution of predictions errors, despite its relevance for operations. We put forward Optimal Transport (OT) as a spatial evaluation metric and loss function. The proposed OT metric assesses the utility of spatial prediction models in terms of the relocation costs caused by prediction errors. In experiments on real and synthetic data, we demonstrate that 1) the spatial distribution of the prediction errors is relevant in many applications and can be translated to real-world costs, 2) in contrast to other metrics, OT reflects these spatial costs, and 3) OT metrics improve comparability across spatial and temporal scales. Finally, we advocate for leveraging OT as a loss function in neural networks to improve the spatial correctness of predictions. This approach not only aligns evaluation in GeoAI with operational considerations, but also signifies a step forward in refining predictions within geospatial applications. To facilitate the adoption of OT in GIS, we provide code and tutorials at https://github.com/mie-lab/geospatialOT.
☆ BlendRL: A Framework for Merging Symbolic and Neural Policy Learning
Humans can leverage both symbolic reasoning and intuitive reactions. In contrast, reinforcement learning policies are typically encoded in either opaque systems like neural networks or symbolic systems that rely on predefined symbols and rules. This disjointed approach severely limits the agents' capabilities, as they often lack either the flexible low-level reaction characteristic of neural agents or the interpretable reasoning of symbolic agents. To overcome this challenge, we introduce BlendRL, a neuro-symbolic RL framework that harmoniously integrates both paradigms within RL agents that use mixtures of both logic and neural policies. We empirically demonstrate that BlendRL agents outperform both neural and symbolic baselines in standard Atari environments, and showcase their robustness to environmental changes. Additionally, we analyze the interaction between neural and symbolic policies, illustrating how their hybrid use helps agents overcome each other's limitations.
comment: Preprint
☆ State-space models can learn in-context by gradient descent
Deep state-space models (Deep SSMs) have shown capabilities for in-context learning on autoregressive tasks, similar to transformers. However, the architectural requirements and mechanisms enabling this in recurrent networks remain unclear. This study demonstrates that state-space model architectures can perform gradient-based learning and use it for in-context learning. We prove that a single structured state-space model layer, augmented with local self-attention, can reproduce the outputs of an implicit linear model with least squares loss after one step of gradient descent. Our key insight is that the diagonal linear recurrent layer can act as a gradient accumulator, which can be `applied' to the parameters of the implicit regression model. We validate our construction by training randomly initialized augmented SSMs on simple linear regression tasks. The empirically optimized parameters match the theoretical ones, obtained analytically from the implicit model construction. Extensions to multi-step linear and non-linear regression yield consistent results. The constructed SSM encompasses features of modern deep state-space models, with the potential for scalable training and effectiveness even in general tasks. The theoretical construction elucidates the role of local self-attention and multiplicative interactions in recurrent architectures as the key ingredients for enabling the expressive power typical of foundation models.
comment: 16 pages, 5 figures
☆ Are UFOs Driving Innovation? The Illusion of Causality in Large Language Models
Illusions of causality occur when people develop the belief that there is a causal connection between two variables with no supporting evidence. This cognitive bias has been proposed to underlie many societal problems including social prejudice, stereotype formation, misinformation and superstitious thinking. In this research we investigate whether large language models develop the illusion of causality in real-world settings. We evaluated and compared news headlines generated by GPT-4o-Mini, Claude-3.5-Sonnet, and Gemini-1.5-Pro to determine whether the models incorrectly framed correlations as causal relationships. In order to also measure sycophantic behavior, which occurs when a model aligns with a user's beliefs in order to look favorable even if it is not objectively correct, we additionally incorporated the bias into the prompts, observing if this manipulation increases the likelihood of the models exhibiting the illusion of causality. We found that Claude-3.5-Sonnet is the model that presents the lowest degree of causal illusion aligned with experiments on Correlation-to-Causation Exaggeration in human-written press releases. On the other hand, our findings suggest that while mimicry sycophancy increases the likelihood of causal illusions in these models, especially in GPT-4o-Mini, Claude-3.5-Sonnet remains the most robust against this cognitive bias.
☆ Understanding Likelihood Over-optimisation in Direct Alignment Algorithms
Direct Alignment Algorithms (DAAs), such as Direct Preference Optimisation (DPO) and Identity Preference Optimisation (IPO), have emerged as alternatives to online Reinforcement Learning from Human Feedback (RLHF) algorithms such as Proximal Policy Optimisation (PPO) for aligning language models to human preferences, without the need for explicit reward modelling. These methods generally aim to increase the likelihood of generating better (preferred) completions while discouraging worse (non-preferred) ones, while staying close to the original model's behaviour. In this work, we explore the relationship between completion likelihood and model performance in state-of-the-art DAAs, and identify a critical issue of likelihood over-optimisation. Contrary to expectations, we find that higher likelihood of better completions and larger margins between better and worse completion likelihoods do not necessarily lead to better performance, and may even degrade it. Our analysis reveals that while higher likelihood correlates with better memorisation of factual knowledge patterns, a slightly lower completion likelihood tends to improve output diversity, thus leading to better generalisation to unseen scenarios. Moreover, we identify two key indicators that signal when over-optimised output diversity begins to harm performance: Decreasing Entropy over Top-k Tokens and Diminishing Top-k Probability Mass. Our experimental results validate that these indicators are reliable signs of declining performance under different regularisations, helping prevent over-optimisation and improve alignment with human preferences.
comment: Preprint Version
☆ LLM-Mixer: Multiscale Mixing in LLMs for Time Series Forecasting
Time series forecasting remains a challenging task, particularly in the context of complex multiscale temporal patterns. This study presents LLM-Mixer, a framework that improves forecasting accuracy through the combination of multiscale time-series decomposition with pre-trained LLMs (Large Language Models). LLM-Mixer captures both short-term fluctuations and long-term trends by decomposing the data into multiple temporal resolutions and processing them with a frozen LLM, guided by a textual prompt specifically designed for time-series data. Extensive experiments conducted on multivariate and univariate datasets demonstrate that LLM-Mixer achieves competitive performance, outperforming recent state-of-the-art models across various forecasting horizons. This work highlights the potential of combining multiscale analysis and LLMs for effective and scalable time-series forecasting.
comment: Time series forecasting using LLMs
☆ Safety Filtering While Training: Improving the Performance and Sample Efficiency of Reinforcement Learning Agents
Reinforcement learning (RL) controllers are flexible and performant but rarely guarantee safety. Safety filters impart hard safety guarantees to RL controllers while maintaining flexibility. However, safety filters can cause undesired behaviours due to the separation between the controller and the safety filter, often degrading performance and robustness. In this paper, we propose several modifications to incorporating the safety filter in training RL controllers rather than solely applying it during evaluation. The modifications allow the RL controller to learn to account for the safety filter, improving performance. Additionally, our modifications significantly improve sample efficiency and eliminate training-time constraint violations. We verified the proposed modifications in simulated and real experiments with a Crazyflie 2.0 drone. In experiments, we show that the proposed training approaches require significantly fewer environment interactions and improve performance by up to 20% compared to standard RL training.
comment: 8 pages, 9 figures. Code is publicly available at https://github.com/Federico-PizarroBejarano/safe-control-gym/tree/training_rl_paper
☆ RS-MOCO: A deep learning-based topology-preserving image registration method for cardiac T1 mapping
Cardiac T1 mapping can evaluate various clinical symptoms of myocardial tissue. However, there is currently a lack of effective, robust, and efficient methods for motion correction in cardiac T1 mapping. In this paper, we propose a deep learning-based and topology-preserving image registration framework for motion correction in cardiac T1 mapping. Notably, our proposed implicit consistency constraint dubbed BLOC, to some extent preserves the image topology in registration by bidirectional consistency constraint and local anti-folding constraint. To address the contrast variation issue, we introduce a weighted image similarity metric for multimodal registration of cardiac T1-weighted images. Besides, a semi-supervised myocardium segmentation network and a dual-domain attention module are integrated into the framework to further improve the performance of the registration. Numerous comparative experiments, as well as ablation studies, demonstrated the effectiveness and high robustness of our method. The results also indicate that the proposed weighted image similarity metric, specifically crafted for our network, contributes a lot to the enhancement of the motion correction efficacy, while the bidirectional consistency constraint combined with the local anti-folding constraint ensures a more desirable topology-preserving registration mapping.
☆ Efficient, Accurate and Stable Gradients for Neural ODEs
Neural ODEs are a recently developed model class that combine the strong model priors of differential equations with the high-capacity function approximation of neural networks. One advantage of Neural ODEs is the potential for memory-efficient training via the continuous adjoint method. However, memory-efficient training comes at the cost of approximate gradients. Therefore, in practice, gradients are often obtained by simply backpropagating through the internal operations of the forward ODE solve - incurring high memory cost. Interestingly, it is possible to construct algebraically reversible ODE solvers that allow for both exact gradients and the memory-efficiency of the continuous adjoint method. Unfortunately, current reversible solvers are low-order and suffer from poor numerical stability. The use of these methods in practice is therefore limited. In this work, we present a class of algebraically reversible solvers that are both high-order and numerically stable. Moreover, any explicit numerical scheme can be made reversible by our method. This construction naturally extends to numerical schemes for Neural CDEs and SDEs.
comment: Preprint
☆ Improve Value Estimation of Q Function and Reshape Reward with Monte Carlo Tree Search
Reinforcement learning has achieved remarkable success in perfect information games such as Go and Atari, enabling agents to compete at the highest levels against human players. However, research in reinforcement learning for imperfect information games has been relatively limited due to the more complex game structures and randomness. Traditional methods face challenges in training and improving performance in imperfect information games due to issues like inaccurate Q value estimation and reward sparsity. In this paper, we focus on Uno, an imperfect information game, and aim to address these problems by reducing Q value overestimation and reshaping reward function. We propose a novel algorithm that utilizes Monte Carlo Tree Search to improve the value estimation in Q function. Even though we choose Double Deep Q Learning as the foundational framework in this paper, our method can be generalized and used in any algorithm which needs Q value estimation, such as the Actor-Critic. Additionally, we employ Monte Carlo Tree Search to reshape the reward structure in the game environment. We compared our algorithm with several traditional methods applied to games such as Double Deep Q Learning, Deep Monte Carlo and Neural Fictitious Self Play, and the experiments demonstrate that our algorithm consistently outperforms these approaches, especially as the number of players in Uno increases, indicating a higher level of difficulty.
☆ Fast Local Neural Regression for Low-Cost, Path Traced Lambertian Global Illumination
Despite recent advances in hardware acceleration of ray tracing, real-time ray budgets remain stubbornly limited at a handful of samples per pixel (spp) on commodity hardware, placing the onus on denoising algorithms to achieve high visual quality for path traced global illumination. Neural network-based solutions give excellent result quality at the cost of increased execution time relative to hand-engineered methods, making them less suitable for deployment on resource-constrained systems. We therefore propose incorporating a neural network into a computationally-efficient local linear model-based denoiser, and demonstrate faithful single-frame reconstruction of global illumination for Lambertian scenes at very low sample counts (1spp) and for low computational cost. Other contributions include improving the quality and performance of local linear model-based denoising through a simplified mathematical treatment, and demonstration of the surprising usefulness of ambient occlusion as a guide channel. We also show how our technique is straightforwardly extensible to joint denoising and upsampling of path traced renders with reference to low-cost, rasterized guide channels.
comment: 11 pages, 10 figures, 1 table
☆ Federated Learning framework for LoRaWAN-enabled IIoT communication: A case study
The development of intelligent Industrial Internet of Things (IIoT) systems promises to revolutionize operational and maintenance practices, driving improvements in operational efficiency. Anomaly detection within IIoT architectures plays a crucial role in preventive maintenance and spotting irregularities in industrial components. However, due to limited message and processing capacity, traditional Machine Learning (ML) faces challenges in deploying anomaly detection models in resource-constrained environments like LoRaWAN. On the other hand, Federated Learning (FL) solves this problem by enabling distributed model training, addressing privacy concerns, and minimizing data transmission. This study explores using FL for anomaly detection in industrial and civil construction machinery architectures that use IIoT prototypes with LoRaWAN communication. The process leverages an optimized autoencoder neural network structure and compares federated models with centralized ones. Despite uneven data distribution among machine clients, FL demonstrates effectiveness, with a mean F1 score (of 94.77), accuracy (of 92.30), TNR (of 90.65), and TPR (92.93), comparable to centralized models, considering airtime of trainning messages of 52.8 min. Local model evaluations on each machine highlight adaptability. At the same time, the performed analysis identifies message requirements, minimum training hours, and optimal round/epoch configurations for FL in LoRaWAN, guiding future implementations in constrained industrial environments.
☆ Black-box Uncertainty Quantification Method for LLM-as-a-Judge
LLM-as-a-Judge is a widely used method for evaluating the performance of Large Language Models (LLMs) across various tasks. We address the challenge of quantifying the uncertainty of LLM-as-a-Judge evaluations. While uncertainty quantification has been well-studied in other domains, applying it effectively to LLMs poses unique challenges due to their complex decision-making capabilities and computational demands. In this paper, we introduce a novel method for quantifying uncertainty designed to enhance the trustworthiness of LLM-as-a-Judge evaluations. The method quantifies uncertainty by analyzing the relationships between generated assessments and possible ratings. By cross-evaluating these relationships and constructing a confusion matrix based on token probabilities, the method derives labels of high or low uncertainty. We evaluate our method across multiple benchmarks, demonstrating a strong correlation between the accuracy of LLM evaluations and the derived uncertainty scores. Our findings suggest that this method can significantly improve the reliability and consistency of LLM-as-a-Judge evaluations.
☆ PaSTe: Improving the Efficiency of Visual Anomaly Detection at the Edge
Visual Anomaly Detection (VAD) has gained significant research attention for its ability to identify anomalous images and pinpoint the specific areas responsible for the anomaly. A key advantage of VAD is its unsupervised nature, which eliminates the need for costly and time-consuming labeled data collection. However, despite its potential for real-world applications, the literature has given limited focus to resource-efficient VAD, particularly for deployment on edge devices. This work addresses this gap by leveraging lightweight neural networks to reduce memory and computation requirements, enabling VAD deployment on resource-constrained edge devices. We benchmark the major VAD algorithms within this framework and demonstrate the feasibility of edge-based VAD using the well-known MVTec dataset. Furthermore, we introduce a novel algorithm, Partially Shared Teacher-student (PaSTe), designed to address the high resource demands of the existing Student Teacher Feature Pyramid Matching (STFPM) approach. Our results show that PaSTe decreases the inference time by 25%, while reducing the training time by 33% and peak RAM usage during training by 76%. These improvements make the VAD process significantly more efficient, laying a solid foundation for real-world deployment on edge devices.
comment: 13 pages, 6 figures
☆ On-the-fly Modulation for Balanced Multimodal Learning
Multimodal learning is expected to boost model performance by integrating information from different modalities. However, its potential is not fully exploited because the widely-used joint training strategy, which has a uniform objective for all modalities, leads to imbalanced and under-optimized uni-modal representations. Specifically, we point out that there often exists modality with more discriminative information, e.g., vision of playing football and sound of blowing wind. They could dominate the joint training process, resulting in other modalities being significantly under-optimized. To alleviate this problem, we first analyze the under-optimized phenomenon from both the feed-forward and the back-propagation stages during optimization. Then, On-the-fly Prediction Modulation (OPM) and On-the-fly Gradient Modulation (OGM) strategies are proposed to modulate the optimization of each modality, by monitoring the discriminative discrepancy between modalities during training. Concretely, OPM weakens the influence of the dominant modality by dropping its feature with dynamical probability in the feed-forward stage, while OGM mitigates its gradient in the back-propagation stage. In experiments, our methods demonstrate considerable improvement across a variety of multimodal tasks. These simple yet effective strategies not only enhance performance in vanilla and task-oriented multimodal models, but also in more complex multimodal tasks, showcasing their effectiveness and flexibility. The source code is available at \url{https://github.com/GeWu-Lab/BML_TPAMI2024}.
comment: Accepted by T-PAMI 2024
☆ Machine Learning via rough mereology
Rough sets (RS)proved a thriving realm with successes inn many fields of ML and AI. In this note, we expand RS to RM - rough mereology which provides a measurable degree of uncertainty to those areas.
comment: 18 pages, 1 figure
☆ Why Go Full? Elevating Federated Learning Through Partial Network Updates NeurIPS 2024
Federated learning is a distributed machine learning paradigm designed to protect user data privacy, which has been successfully implemented across various scenarios. In traditional federated learning, the entire parameter set of local models is updated and averaged in each training round. Although this full network update method maximizes knowledge acquisition and sharing for each model layer, it prevents the layers of the global model from cooperating effectively to complete the tasks of each client, a challenge we refer to as layer mismatch. This mismatch problem recurs after every parameter averaging, consequently slowing down model convergence and degrading overall performance. To address the layer mismatch issue, we introduce the FedPart method, which restricts model updates to either a single layer or a few layers during each communication round. Furthermore, to maintain the efficiency of knowledge acquisition and sharing, we develop several strategies to select trainable layers in each round, including sequential updating and multi-round cycle training. Through both theoretical analysis and experiments, our findings demonstrate that the FedPart method significantly surpasses conventional full network update strategies in terms of convergence speed and accuracy, while also reducing communication and computational overheads.
comment: 27 pages, 8 figures, Accepted to NeurIPS 2024
☆ Efficiera Residual Networks: Hardware-Friendly Fully Binary Weight with 2-bit Activation Model Achieves Practical ImageNet Accuracy
The edge-device environment imposes severe resource limitations, encompassing computation costs, hardware resource usage, and energy consumption for deploying deep neural network models. Ultra-low-bit quantization and hardware accelerators have been explored as promising approaches to address these challenges. Ultra-low-bit quantization significantly reduces the model size and the computational cost. Despite progress so far, many competitive ultra-low-bit models still partially rely on float or non-ultra-low-bit quantized computation such as the input and output layer. We introduce Efficiera Residual Networks (ERNs), a model optimized for low-resource edge devices. ERNs achieve full ultra-low-bit quantization, with all weights, including the initial and output layers, being binary, and activations set at 2 bits. We introduce the shared constant scaling factor technique to enable integer-valued computation in residual connections, allowing our model to operate without float values until the final convolution layer. Demonstrating competitiveness, ERNs achieve an ImageNet top-1 accuracy of 72.5pt with a ResNet50-compatible architecture and 63.6pt with a model size less than 1MB. Moreover, ERNs exhibit impressive inference times, reaching 300FPS with the smallest model and 60FPS with the largest model on a cost-efficient FPGA device.
comment: 11pages, 2 figures, the model implementation is available at https://github.com/LeapMind/ERN
☆ LoKO: Low-Rank Kalman Optimizer for Online Fine-Tuning of Large Models
Training large models with millions or even billions of parameters from scratch incurs substantial computational costs. Parameter Efficient Fine-Tuning (PEFT) methods, particularly Low-Rank Adaptation (LoRA), address this challenge by adapting only a reduced number of parameters to specific tasks with gradient-based optimizers. In this paper, we cast PEFT as an optimal filtering/state estimation problem and present Low-Rank Kalman Optimizer (LoKO) to estimate the optimal trainable parameters in an online manner. We leverage the low-rank decomposition in LoRA to significantly reduce matrix sizes in Kalman iterations and further capitalize on a diagonal approximation of the covariance matrix to effectively decrease computational complexity from quadratic to linear in the number of trainable parameters. Moreover, we discovered that the initialization of the covariance matrix within the Kalman algorithm and the accurate estimation of the observation noise covariance are the keys in this formulation, and we propose robust approaches that work well across a vast range of well-established computer vision and language models. Our results show that LoKO converges with fewer iterations and yields better performance models compared to commonly used optimizers with LoRA in both image classifications and language tasks. Our study opens up the possibility of leveraging the Kalman filter as an effective optimizer for the online fine-tuning of large models.
☆ A model learning framework for inferring the dynamics of transmission rate depending on exogenous variables for epidemic forecasts
In this work, we aim to formalize a novel scientific machine learning framework to reconstruct the hidden dynamics of the transmission rate, whose inaccurate extrapolation can significantly impair the quality of the epidemic forecasts, by incorporating the influence of exogenous variables (such as environmental conditions and strain-specific characteristics). We propose an hybrid model that blends a data-driven layer with a physics-based one. The data-driven layer is based on a neural ordinary differential equation that learns the dynamics of the transmission rate, conditioned on the meteorological data and wave-specific latent parameters. The physics-based layer, instead, consists of a standard SEIR compartmental model, wherein the transmission rate represents an input. The learning strategy follows an end-to-end approach: the loss function quantifies the mismatch between the actual numbers of infections and its numerical prediction obtained from the SEIR model incorporating as an input the transmission rate predicted by the neural ordinary differential equation. We validate this original approach using both a synthetic test case and a realistic test case based on meteorological data (temperature and humidity) and influenza data from Italy between 2010 and 2020. In both scenarios, we achieve low generalization error on the test set and observe strong alignment between the reconstructed model and established findings on the influence of meteorological factors on epidemic spread. Finally, we implement a data assimilation strategy to adapt the neural equation to the specific characteristics of an epidemic wave under investigation, and we conduct sensitivity tests on the network hyperparameters.
☆ Data Quality Control in Federated Instruction-tuning of Large Language Models
By leveraging massively distributed data, federated learning (FL) enables collaborative instruction tuning of large language models (LLMs) in a privacy-preserving way. While FL effectively expands the data quantity, the issue of data quality remains under-explored in the current literature on FL for LLMs. To address this gap, we propose a new framework of federated instruction tuning of LLMs with data quality control (FedDQC), which measures data quality to facilitate the subsequent filtering and hierarchical training processes. Our approach introduces an efficient metric to assess each client's instruction-response alignment (IRA), identifying potentially noisy data through single-shot inference. Low-IRA samples are potentially noisy and filtered to mitigate their negative impacts. To further utilize this IRA value, we propose a quality-aware hierarchical training paradigm, where LLM is progressively fine-tuned from high-IRA to low-IRA data, mirroring the easy-to-hard learning process. We conduct extensive experiments on 4 synthetic and a real-world dataset, and compare our method with baselines adapted from centralized setting. Results show that our method consistently and significantly improves the performance of LLMs trained on mix-quality data in FL.
☆ Transfer Learning with Foundational Models for Time Series Forecasting using Low-Rank Adaptations
High computational power and the availability of large datasets have supported the development of Foundational Models. They are a new emerging technique widely used in Generative Artificial Intelligence, characterized by their scalability and their use in Transfer Learning. The enormous and heterogeneous amounts of data used in their initial training phase, known as pre-training, give them a higher generalization capacity than any other specific model, constituting a solid base that can be adapted or adjusted to a wide range of tasks, increasing their applicability. This study proposes LLIAM, the Llama Lora-Integrated Autorregresive Model. Low-Rank Adaptations are used to enhance the knowledge of the model with diverse time series datasets, known as the fine-tuning phase. To illustrate the capabilities of our proposal, two sets of experiments have been carried out that obtained favorable and promising results with lower training times than other Deep Learning approaches. With this work, we also encourage the use of available resources (such as these pre-trained models) to avoid unnecessary and costly training, narrowing the gap between the goals of traditional Artificial Intelligence and those specified by the definition of Green Artificial Intelligence.
☆ It Takes Two to Tango: Directly Optimizing for Constrained Synthesizability in Generative Molecular Design
Constrained synthesizability is an unaddressed challenge in generative molecular design. In particular, designing molecules satisfying multi-parameter optimization objectives, while simultaneously being synthesizable and enforcing the presence of specific commercial building blocks in the synthesis. This is practically important for molecule re-purposing, sustainability, and efficiency. In this work, we propose a novel reward function called TANimoto Group Overlap (TANGO), which uses chemistry principles to transform a sparse reward function into a dense and learnable reward function -- crucial for reinforcement learning. TANGO can augment general-purpose molecular generative models to directly optimize for constrained synthesizability while simultaneously optimizing for other properties relevant to drug discovery using reinforcement learning. Our framework is general and addresses starting-material, intermediate, and divergent synthesis constraints. Contrary to most existing works in the field, we show that incentivizing a general-purpose (without any inductive biases) model is a productive approach to navigating challenging optimization scenarios. We demonstrate this by showing that the trained models explicitly learn a desirable distribution. Our framework is the first generative approach to tackle constrained synthesizability.
☆ Leveraging LLM Embeddings for Cross Dataset Label Alignment and Zero Shot Music Emotion Prediction
In this work, we present a novel method for music emotion recognition that leverages Large Language Model (LLM) embeddings for label alignment across multiple datasets and zero-shot prediction on novel categories. First, we compute LLM embeddings for emotion labels and apply non-parametric clustering to group similar labels, across multiple datasets containing disjoint labels. We use these cluster centers to map music features (MERT) to the LLM embedding space. To further enhance the model, we introduce an alignment regularization that enables dissociation of MERT embeddings from different clusters. This further enhances the model's ability to better adaptation to unseen datasets. We demonstrate the effectiveness of our approach by performing zero-shot inference on a new dataset, showcasing its ability to generalize to unseen labels without additional training.
☆ Network Representation Learning for Biophysical Neural Network Analysis
The analysis of biophysical neural networks (BNNs) has been a longstanding focus in computational neuroscience. A central yet unresolved challenge in BNN analysis lies in deciphering the correlations between neuronal and synaptic dynamics, their connectivity patterns, and learning process. To address this, we introduce a novel BNN analysis framework grounded in network representation learning (NRL), which leverages attention scores to uncover intricate correlations between network components and their features. Our framework integrates a new computational graph (CG)-based BNN representation, a bio-inspired graph attention network (BGAN) that enables multiscale correlation analysis across BNN representations, and an extensive BNN dataset. The CG-based representation captures key computational features, information flow, and structural relationships underlying neuronal and synaptic dynamics, while BGAN reflects the compositional structure of neurons, including dendrites, somas, and axons, as well as bidirectional information flows between BNN components. The dataset comprises publicly available models from ModelDB, reconstructed using the Python and standardized in NeuroML format, and is augmented with data derived from canonical neuron and synapse models. To our knowledge, this study is the first to apply an NRL-based approach to the full spectrum of BNNs and their analysis.
comment: 14 pages, Work-In-Progress
☆ Offline Model-Based Optimization by Learning to Rank
Offline model-based optimization (MBO) aims to identify a design that maximizes a black-box function using only a fixed, pre-collected dataset of designs and their corresponding scores. A common approach in offline MBO is to train a regression-based surrogate model by minimizing mean squared error (MSE) and then find the best design within this surrogate model by different optimizers (e.g., gradient ascent). However, a critical challenge is the risk of out-of-distribution errors, i.e., the surrogate model may typically overestimate the scores and mislead the optimizers into suboptimal regions. Prior works have attempted to address this issue in various ways, such as using regularization techniques and ensemble learning to enhance the robustness of the model, but it still remains. In this paper, we argue that regression models trained with MSE are not well-aligned with the primary goal of offline MBO, which is to select promising designs rather than to predict their scores precisely. Notably, if a surrogate model can maintain the order of candidate designs based on their relative score relationships, it can produce the best designs even without precise predictions. To validate it, we conduct experiments to compare the relationship between the quality of the final designs and MSE, finding that the correlation is really very weak. In contrast, a metric that measures order-maintaining quality shows a significantly stronger correlation. Based on this observation, we propose learning a ranking-based model that leverages learning to rank techniques to prioritize promising designs based on their relative scores. We show that the generalization error on ranking loss can be well bounded. Empirical results across diverse tasks demonstrate the superior performance of our proposed ranking-based models than twenty existing methods.
☆ On Rank-Dependent Generalisation Error Bounds for Transformers
In this paper, we introduce various covering number bounds for linear function classes, each subject to different constraints on input and matrix norms. These bounds are contingent on the rank of each class of matrices. We then apply these bounds to derive generalization errors for single layer transformers. Our results improve upon several existing generalization bounds in the literature and are independent of input sequence length, highlighting the advantages of employing low-rank matrices in transformer design. More specifically, our achieved generalisation error bound decays as $O(1/\sqrt{n})$ where $n$ is the sample length, which improves existing results in research literature of the order $O((\log n)/(\sqrt{n}))$. It also decays as $O(\log r_w)$ where $r_w$ is the rank of the combination of query and and key matrices.
comment: 20 pages
☆ BSM: Small but Powerful Biological Sequence Model for Genes and Proteins
Modeling biological sequences such as DNA, RNA, and proteins is crucial for understanding complex processes like gene regulation and protein synthesis. However, most current models either focus on a single type or treat multiple types of data separately, limiting their ability to capture cross-modal relationships. We propose that by learning the relationships between these modalities, the model can enhance its understanding of each type. To address this, we introduce BSM, a small but powerful mixed-modal biological sequence foundation model, trained on three types of data: RefSeq, Gene Related Sequences, and interleaved biological sequences from the web. These datasets capture the genetic flow, gene-protein relationships, and the natural co-occurrence of diverse biological data, respectively. By training on mixed-modal data, BSM significantly enhances learning efficiency and cross-modal representation, outperforming models trained solely on unimodal data. With only 110M parameters, BSM achieves performance comparable to much larger models across both single-modal and mixed-modal tasks, and uniquely demonstrates in-context learning capability for mixed-modal tasks, which is absent in existing models. Further scaling to 270M parameters demonstrates even greater performance gains, highlighting the potential of BSM as a significant advancement in multimodal biological sequence modeling.
☆ Towards Fair Graph Representation Learning in Social Networks
With the widespread use of Graph Neural Networks (GNNs) for representation learning from network data, the fairness of GNN models has raised great attention lately. Fair GNNs aim to ensure that node representations can be accurately classified, but not easily associated with a specific group. Existing advanced approaches essentially enhance the generalisation of node representation in combination with data augmentation strategy, and do not directly impose constraints on the fairness of GNNs. In this work, we identify that a fundamental reason for the unfairness of GNNs in social network learning is the phenomenon of social homophily, i.e., users in the same group are more inclined to congregate. The message-passing mechanism of GNNs can cause users in the same group to have similar representations due to social homophily, leading model predictions to establish spurious correlations with sensitive attributes. Inspired by this reason, we propose a method called Equity-Aware GNN (EAGNN) towards fair graph representation learning. Specifically, to ensure that model predictions are independent of sensitive attributes while maintaining prediction performance, we introduce constraints for fair representation learning based on three principles: sufficiency, independence, and separation. We theoretically demonstrate that our EAGNN method can effectively achieve group fairness. Extensive experiments on three datasets with varying levels of social homophily illustrate that our EAGNN method achieves the state-of-the-art performance across two fairness metrics and offers competitive effectiveness.
☆ Advancing Training Efficiency of Deep Spiking Neural Networks through Rate-based Backpropagation NeurIPS 2024
Recent insights have revealed that rate-coding is a primary form of information representation captured by surrogate-gradient-based Backpropagation Through Time (BPTT) in training deep Spiking Neural Networks (SNNs). Motivated by these findings, we propose rate-based backpropagation, a training strategy specifically designed to exploit rate-based representations to reduce the complexity of BPTT. Our method minimizes reliance on detailed temporal derivatives by focusing on averaged dynamics, streamlining the computational graph to reduce memory and computational demands of SNNs training. We substantiate the rationality of the gradient approximation between BPTT and the proposed method through both theoretical analysis and empirical observations. Comprehensive experiments on CIFAR-10, CIFAR-100, ImageNet, and CIFAR10-DVS validate that our method achieves comparable performance to BPTT counterparts, and surpasses state-of-the-art efficient training techniques. By leveraging the inherent benefits of rate-coding, this work sets the stage for more scalable and efficient SNNs training within resource-constrained environments. Our code is available at https://github.com/Tab-ct/rate-based-backpropagation.
comment: Accepted by NeurIPS 2024
☆ Poisson-Dirac Neural Networks for Modeling Coupled Dynamical Systems across Domains
Deep learning has achieved great success in modeling dynamical systems, providing data-driven simulators to predict complex phenomena, even without known governing equations. However, existing models have two major limitations: their narrow focus on mechanical systems and their tendency to treat systems as monolithic. These limitations reduce their applicability to dynamical systems in other domains, such as electrical and hydraulic systems, and to coupled systems. To address these limitations, we propose Poisson-Dirac Neural Networks (PoDiNNs), a novel framework based on the Dirac structure that unifies the port-Hamiltonian and Poisson formulations from geometric mechanics. This framework enables a unified representation of various dynamical systems across multiple domains as well as their interactions and degeneracies arising from couplings. Our experiments demonstrate that PoDiNNs offer improved accuracy and interpretability in modeling unknown coupled dynamical systems from data.
☆ How Transformers Implement Induction Heads: Approximation and Optimization Analysis
Transformers have demonstrated exceptional in-context learning capabilities, yet the theoretical understanding of the underlying mechanisms remain limited. A recent work (Elhage et al., 2021) identified a "rich" in-context mechanism known as induction head, contrasting with "lazy" $n$-gram models that overlook long-range dependencies. In this work, we provide both approximation and optimization analyses of how transformers implement induction heads. In the approximation analysis, we formalize both standard and generalized induction head mechanisms, and examine how transformers can efficiently implement them, with an emphasis on the distinct role of each transformer submodule. For the optimization analysis, we study the training dynamics on a synthetic mixed target, composed of a 4-gram and an in-context 2-gram component. This setting enables us to precisely characterize the entire training process and uncover an {\em abrupt transition} from lazy (4-gram) to rich (induction head) mechanisms as training progresses.
comment: 39 pages
☆ Can sparse autoencoders make sense of latent representations?
Sparse autoencoders (SAEs) have lately been used to uncover interpretable latent features in large language models. Here, we explore their potential for decomposing latent representations in complex and high-dimensional biological data, where the underlying variables are often unknown. On simulated data we show that generative hidden variables can be captured in learned representations in the form of superpositions. The degree to which they are learned depends on the completeness of the representations. Superpositions, however, are not identifiable if these generative variables are unknown. SAEs can to some extent recover these variables, yielding interpretable features. Applied to single-cell multi-omics data, we show that an SAE can uncover key biological processes such as carbon dioxide transport and ion homeostasis, which are crucial for red blood cell differentiation and immune function. Our findings highlight how SAEs can be used in advancing interpretability in biological and other scientific domains.
☆ CoActionGraphRec: Sequential Multi-Interest Recommendations Using Co-Action Graphs
There are unique challenges to developing item recommender systems for e-commerce platforms like eBay due to sparse data and diverse user interests. While rich user-item interactions are important, eBay's data sparsity exceeds other e-commerce sites by an order of magnitude. To address this challenge, we propose CoActionGraphRec (CAGR), a text based two-tower deep learning model (Item Tower and User Tower) utilizing co-action graph layers. In order to enhance user and item representations, a graph-based solution tailored to eBay's environment is utilized. For the Item Tower, we represent each item using its co-action items to capture collaborative signals in a co-action graph that is fully leveraged by the graph neural network component. For the User Tower, we build a fully connected graph of each user's behavior sequence, with edges encoding pairwise relationships. Furthermore, an explicit interaction module learns representations capturing behavior interactions. Extensive offline and online A/B test experiments demonstrate the effectiveness of our proposed approach and results show improved performance over state-of-the-art methods on key metrics.
☆ Advanced Persistent Threats (APT) Attribution Using Deep Reinforcement Learning
This paper investigates the application of Deep Reinforcement Learning (DRL) for attributing malware to specific Advanced Persistent Threat (APT) groups through detailed behavioural analysis. By analysing over 3500 malware samples from 12 distinct APT groups, the study utilises sophisticated tools like Cuckoo Sandbox to extract behavioural data, providing a deep insight into the operational patterns of malware. The research demonstrates that the DRL model significantly outperforms traditional machine learning approaches such as SGD, SVC, KNN, MLP, and Decision Tree Classifiers, achieving an impressive test accuracy of 89.27 %. It highlights the model capability to adeptly manage complex, variable, and elusive malware attributes. Furthermore, the paper discusses the considerable computational resources and extensive data dependencies required for deploying these advanced AI models in cybersecurity frameworks. Future research is directed towards enhancing the efficiency of DRL models, expanding the diversity of the datasets, addressing ethical concerns, and leveraging Large Language Models (LLMs) to refine reward mechanisms and optimise the DRL framework. By showcasing the transformative potential of DRL in malware attribution, this research advocates for a responsible and balanced approach to AI integration, with the goal of advancing cybersecurity through more adaptable, accurate, and robust systems.
comment: 21 Pages
☆ Nonlinear Gaussian process tomography with imposed non-negativity constraints on physical quantities for plasma diagnostics
We propose a novel tomographic method, nonlinear Gaussian process tomography (nonlinear GPT) that employs the Laplace approximation to ensure the non-negative physical quantity, such as the emissivity of plasma optical diagnostics. This new method implements a logarithmic Gaussian process (log-GP) to model plasma distribution more naturally, thereby expanding the limitations of standard GPT, which are restricted to linear problems and may yield non-physical negative values. The effectiveness of the proposed log-GP tomography is demonstrated through a case study using the Ring Trap 1 (RT-1) device, where log-GPT outperforms existing methods, standard GPT, and the Minimum Fisher Information (MFI) methods in terms of reconstruction accuracy. The result highlights the effectiveness of nonlinear GPT for imposing physical constraints in applications to an inverse problem.
comment: 19 pages, 7 figures
☆ Conditional Density Estimation with Histogram Trees
Conditional density estimation (CDE) goes beyond regression by modeling the full conditional distribution, providing a richer understanding of the data than just the conditional mean in regression. This makes CDE particularly useful in critical application domains. However, interpretable CDE methods are understudied. Current methods typically employ kernel-based approaches, using kernel functions directly for kernel density estimation or as basis functions in linear models. In contrast, despite their conceptual simplicity and visualization suitability, tree-based methods -- which are arguably more comprehensible -- have been largely overlooked for CDE tasks. Thus, we propose the Conditional Density Tree (CDTree), a fully non-parametric model consisting of a decision tree in which each leaf is formed by a histogram model. Specifically, we formalize the problem of learning a CDTree using the minimum description length (MDL) principle, which eliminates the need for tuning the hyperparameter for regularization. Next, we propose an iterative algorithm that, although greedily, searches the optimal histogram for every possible node split. Our experiments demonstrate that, in comparison to existing interpretable CDE methods, CDTrees are both more accurate (as measured by the log-loss) and more robust against irrelevant features. Further, our approach leads to smaller tree sizes than existing tree-based models, which benefits interpretability.
comment: Accepted to Neurips 2024
☆ Meta-DT: Offline Meta-RL as Conditional Sequence Modeling with World Model Disentanglement NeurIPS 2024
A longstanding goal of artificial general intelligence is highly capable generalists that can learn from diverse experiences and generalize to unseen tasks. The language and vision communities have seen remarkable progress toward this trend by scaling up transformer-based models trained on massive datasets, while reinforcement learning (RL) agents still suffer from poor generalization capacity under such paradigms. To tackle this challenge, we propose Meta Decision Transformer (Meta-DT), which leverages the sequential modeling ability of the transformer architecture and robust task representation learning via world model disentanglement to achieve efficient generalization in offline meta-RL. We pretrain a context-aware world model to learn a compact task representation, and inject it as a contextual condition to the causal transformer to guide task-oriented sequence generation. Then, we subtly utilize history trajectories generated by the meta-policy as a self-guided prompt to exploit the architectural inductive bias. We select the trajectory segment that yields the largest prediction error on the pretrained world model to construct the prompt, aiming to encode task-specific information complementary to the world model maximally. Notably, the proposed framework eliminates the requirement of any expert demonstration or domain knowledge at test time. Experimental results on MuJoCo and Meta-World benchmarks across various dataset types show that Meta-DT exhibits superior few and zero-shot generalization capacity compared to strong baselines while being more practical with fewer prerequisites. Our code is available at https://github.com/NJU-RL/Meta-DT.
comment: NeurIPS 2024
☆ On Championing Foundation Models: From Explainability to Interpretability
Understanding the inner mechanisms of black-box foundation models (FMs) is essential yet challenging in artificial intelligence and its applications. Over the last decade, the long-running focus has been on their explainability, leading to the development of post-hoc explainable methods to rationalize the specific decisions already made by black-box FMs. However, these explainable methods have certain limitations in terms of faithfulness, detail capture and resource requirement. Consequently, in response to these issues, a new class of interpretable methods should be considered to unveil the underlying mechanisms in an accurate, comprehensive, heuristic and resource-light way. This survey aims to review interpretable methods that comply with the aforementioned principles and have been successfully applied to FMs. These methods are deeply rooted in machine learning theory, covering the analysis of generalization performance, expressive capability, and dynamic behavior. They provide a thorough interpretation of the entire workflow of FMs, ranging from the inference capability and training dynamics to their ethical implications. Ultimately, drawing upon these interpretations, this review identifies the next frontier research directions for FMs.
comment: 45 pages, 14 figures
☆ Are High-Degree Representations Really Unnecessary in Equivariant Graph Neural Networks?
Equivariant Graph Neural Networks (GNNs) that incorporate E(3) symmetry have achieved significant success in various scientific applications. As one of the most successful models, EGNN leverages a simple scalarization technique to perform equivariant message passing over only Cartesian vectors (i.e., 1st-degree steerable vectors), enjoying greater efficiency and efficacy compared to equivariant GNNs using higher-degree steerable vectors. This success suggests that higher-degree representations might be unnecessary. In this paper, we disprove this hypothesis by exploring the expressivity of equivariant GNNs on symmetric structures, including $k$-fold rotations and regular polyhedra. We theoretically demonstrate that equivariant GNNs will always degenerate to a zero function if the degree of the output representations is fixed to 1 or other specific values. Based on this theoretical insight, we propose HEGNN, a high-degree version of EGNN to increase the expressivity by incorporating high-degree steerable vectors while maintaining EGNN's efficiency through the scalarization trick. Our extensive experiments demonstrate that HEGNN not only aligns with our theoretical analyses on toy datasets consisting of symmetric structures, but also shows substantial improvements on more complicated datasets such as $N$-body and MD17. Our theoretical findings and empirical results potentially open up new possibilities for the research of equivariant GNNs.
☆ Hessian-Informed Flow Matching
Modeling complex systems that evolve toward equilibrium distributions is important in various physical applications, including molecular dynamics and robotic control. These systems often follow the stochastic gradient descent of an underlying energy function, converging to stationary distributions around energy minima. The local covariance of these distributions is shaped by the energy landscape's curvature, often resulting in anisotropic characteristics. While flow-based generative models have gained traction in generating samples from equilibrium distributions in such applications, they predominately employ isotropic conditional probability paths, limiting their ability to capture such covariance structures. In this paper, we introduce Hessian-Informed Flow Matching (HI-FM), a novel approach that integrates the Hessian of an energy function into conditional flows within the flow matching framework. This integration allows HI-FM to account for local curvature and anisotropic covariance structures. Our approach leverages the linearization theorem from dynamical systems and incorporates additional considerations such as time transformations and equivariance. Empirical evaluations on the MNIST and Lennard-Jones particles datasets demonstrate that HI-FM improves the likelihood of test samples.
comment: In submission
☆ KLay: Accelerating Neurosymbolic AI
A popular approach to neurosymbolic AI involves mapping logic formulas to arithmetic circuits (computation graphs consisting of sums and products) and passing the outputs of a neural network through these circuits. This approach enforces symbolic constraints onto a neural network in a principled and end-to-end differentiable way. Unfortunately, arithmetic circuits are challenging to run on modern AI accelerators as they exhibit a high degree of irregular sparsity. To address this limitation, we introduce knowledge layers (KLay), a new data structure to represent arithmetic circuits that can be efficiently parallelized on GPUs. Moreover, we contribute two algorithms used in the translation of traditional circuit representations to KLay and a further algorithm that exploits parallelization opportunities during circuit evaluations. We empirically show that KLay achieves speedups of multiple orders of magnitude over the state of the art, thereby paving the way towards scaling neurosymbolic AI to larger real-world applications.
☆ Enhancing Unimodal Latent Representations in Multimodal VAEs through Iterative Amortized Inference
Multimodal variational autoencoders (VAEs) aim to capture shared latent representations by integrating information from different data modalities. A significant challenge is accurately inferring representations from any subset of modalities without training an impractical number (2^M) of inference networks for all possible modality combinations. Mixture-based models simplify this by requiring only as many inference models as there are modalities, aggregating unimodal inferences. However, they suffer from information loss when modalities are missing. Alignment-based VAEs address this by aligning unimodal inference models with a multimodal model through minimizing the Kullback-Leibler (KL) divergence but face issues due to amortization gaps, which compromise inference accuracy. To tackle these problems, we introduce multimodal iterative amortized inference, an iterative refinement mechanism within the multimodal VAE framework. This method overcomes information loss from missing modalities and minimizes the amortization gap by iteratively refining the multimodal inference using all available modalities. By aligning unimodal inference to this refined multimodal posterior, we achieve unimodal inferences that effectively incorporate multimodal information while requiring only unimodal inputs during inference. Experiments on benchmark datasets show that our approach improves inference performance, evidenced by higher linear classification accuracy and competitive cosine similarity, and enhances cross-modal generation, indicated by lower FID scores. This demonstrates that our method enhances inferred representations from unimodal inputs.
comment: 22 pages, 12 figures
☆ RSSI-Assisted CSI-Based Passenger Counting with Multiple Wi-Fi Receivers
Passenger counting is crucial for public transport vehicle scheduling and traffic capacity evaluation. However, most existing methods are either costly or with low counting accuracy, leading to the recent use of Wi-Fi signals for this purpose. In this paper, we develop an efficient edge computing-based passenger counting system consists of multiple Wi-Fi receivers and an edge server. It leverages channel state information (CSI) and received signal strength indicator (RSSI) to facilitate the collaboration among multiple receivers. Specifically, we design a novel CSI feature fusion module called Adaptive RSSI-weighted CSI Feature Concatenation, which integrates locally extracted CSI and RSSI features from multiple receivers for information fusion at the edge server. Performance of our proposed system is evaluated using a real-world dataset collected from a double-decker bus in Hong Kong, with up to 20 passengers. The experimental results reveal that our system achieves an average accuracy and F1-score of over 94%, surpassing other cooperative sensing baselines by at least 2.27% in accuracy and 2.34% in F1-score.
comment: 6 pages, 9 figures, this article was submitted to IEEE for possible publication
☆ Convergence to the Truth
This article reviews and develops an epistemological tradition in philosophy of science, called convergentism, which holds that inference methods should be assessed in terms of their abilities to converge to the truth. This tradition is compared with three competing ones: (1) explanationism, which holds that theory choice should be guided by a theory's overall balance of explanatory virtues, such as simplicity and fit with data; (2) instrumentalism, according to which scientific inference should be driven by the goal of obtaining useful models, rather than true theories; (3) Bayesianism, which features a shift of focus from all-or-nothing beliefs to degrees of belief.
☆ FOOGD: Federated Collaboration for Both Out-of-distribution Generalization and Detection NeurIPS 2024
Federated learning (FL) is a promising machine learning paradigm that collaborates with client models to capture global knowledge. However, deploying FL models in real-world scenarios remains unreliable due to the coexistence of in-distribution data and unexpected out-of-distribution (OOD) data, such as covariate-shift and semantic-shift data. Current FL researches typically address either covariate-shift data through OOD generalization or semantic-shift data via OOD detection, overlooking the simultaneous occurrence of various OOD shifts. In this work, we propose FOOGD, a method that estimates the probability density of each client and obtains reliable global distribution as guidance for the subsequent FL process. Firstly, SM3D in FOOGD estimates score model for arbitrary distributions without prior constraints, and detects semantic-shift data powerfully. Then SAG in FOOGD provides invariant yet diverse knowledge for both local covariate-shift generalization and client performance generalization. In empirical validations, FOOGD significantly enjoys three main advantages: (1) reliably estimating non-normalized decentralized distributions, (2) detecting semantic shift data via score values, and (3) generalizing to covariate-shift data by regularizing feature extractor. The prejoct is open in https://github.com/XeniaLLL/FOOGD-main.git.
comment: NeurIPS 2024
☆ Investigating Data Hierarchies in Multifidelity Machine Learning for Excitation Energies
Recent progress in machine learning (ML) has made high-accuracy quantum chemistry (QC) calculations more accessible. Of particular interest are multifidelity machine learning (MFML) methods where training data from differing accuracies or fidelities are used. These methods usually employ a fixed scaling factor, $\gamma$, to relate the number of training samples across different fidelities, which reflects the cost and assumed sparsity of the data. This study investigates the impact of modifying $\gamma$ on model efficiency and accuracy for the prediction of vertical excitation energies using the QeMFi benchmark dataset. Further, this work introduces QC compute time informed scaling factors, denoted as $\theta$, that vary based on QC compute times at different fidelities. A novel error metric, error contours of MFML, is proposed to provide a comprehensive view of model error contributions from each fidelity. The results indicate that high model accuracy can be achieved with just 2 training samples at the target fidelity when a larger number of samples from lower fidelities are used. This is further illustrated through a novel concept, the $\Gamma$-curve, which compares model error against the time-cost of generating training samples, demonstrating that multifidelity models can achieve high accuracy while minimizing training data costs.
☆ Benchmarking Data Efficiency in $Δ$-ML and Multifidelity Models for Quantum Chemistry
The development of machine learning (ML) methods has made quantum chemistry (QC) calculations more accessible by reducing the compute cost incurred in conventional QC methods. This has since been translated into the overhead cost of generating training data. Increased work in reducing the cost of generating training data resulted in the development of $\Delta$-ML and multifidelity machine learning methods which use data at more than one QC level of accuracy, or fidelity. This work compares the data costs associated with $\Delta$-ML, multifidelity machine learning (MFML), and optimized MFML (o-MFML) in contrast with a newly introduced Multifidelity$\Delta$-Machine Learning (MF$\Delta$ML) method for the prediction of ground state energies over the multifidelity benchmark dataset QeMFi. This assessment is made on the basis of training data generation cost associated with each model and is compared with the single fidelity kernel ridge regression (KRR) case. The results indicate that the use of multifidelity methods surpasses the standard $\Delta$-ML approaches in cases of a large number of predictions. For cases, where $\Delta$-ML method might be favored, such as small test set regimes, the MF$\Delta$-ML method is shown to be more efficient than conventional $\Delta$-ML.
comment: Supplementary information (sections S1,S2, and figure S1) included
☆ Experimental Design Using Interlacing Polynomials
We present a unified deterministic approach for experimental design problems using the method of interlacing polynomials. Our framework recovers the best-known approximation guarantees for the well-studied D/A/E-design problems with simple analysis. Furthermore, we obtain improved non-trivial approximation guarantee for E-design in the challenging small budget regime. Additionally, our approach provides an optimal approximation guarantee for a generalized ratio objective that generalizes both D-design and A-design.
comment: 16 pages
☆ Point-Calibrated Spectral Neural Operators
Two typical neural models have been extensively studied for operator learning, learning in spatial space via attention mechanism or learning in spectral space via spectral analysis technique such as Fourier Transform. Spatial learning enables point-level flexibility but lacks global continuity constraint, while spectral learning enforces spectral continuity prior but lacks point-wise adaptivity. This work innovatively combines the continuity prior and the point-level flexibility, with the introduced Point-Calibrated Spectral Transform. It achieves this by calibrating the preset spectral eigenfunctions with the predicted point-wise frequency preference via neural gate mechanism. Beyond this, we introduce Point-Calibrated Spectral Neural Operators, which learn operator mappings by approximating functions with the point-level adaptive spectral basis, thereby not only preserving the benefits of spectral prior but also boasting the superior adaptability comparable to the attention mechanism. Comprehensive experiments demonstrate its consistent performance enhancement in extensive PDE solving scenarios.
Survey and Evaluation of Converging Architecture in LLMs based on Footsteps of Operations
The advent of the Attention mechanism and Transformer architecture enables contextually natural text generation and compresses the burden of processing entire source information into singular vectors. Based on these two main ideas, model sizes gradually increases to accommodate more precise and comprehensive information, leading to the current state-of-the-art LLMs being very large, with parameters around 70 billion. As the model sizes are growing, the demand for substantial storage and computational capacity increases. This leads to the development of high-bandwidth memory and accelerators, as well as a variety of model architectures designed to meet these requirements. We note that LLM architectures have increasingly converged. This paper analyzes how these converged architectures perform in terms of layer configurations, operational mechanisms, and model sizes, considering various hyperparameter settings. In this paper, we conduct a concise survey of the history of LLMs by tracing the evolution of their operational improvements. Furthermore, we summarize the performance trends of LLMs under various hyperparameter settings using the RTX 6000, which features the state-of-the-art Ada Lovelace architecture. We conclude that even the same model can exhibit different behaviors depending on the hyperparameters or whether it is deployed in server or edge environments.
comment: 13 pages and 16 figures
☆ WPFed: Web-based Personalized Federation for Decentralized Systems
Decentralized learning has become crucial for collaborative model training in environments where data privacy and trust are paramount. In web-based applications, clients are liberated from traditional fixed network topologies, enabling the establishment of arbitrary peer-to-peer (P2P) connections. While this flexibility is highly promising, it introduces a fundamental challenge: the optimal selection of neighbors to ensure effective collaboration. To address this, we introduce WPFed, a fully decentralized, web-based learning framework designed to enable globally optimal neighbor selection. WPFed employs a dynamic communication graph and a weighted neighbor selection mechanism. By assessing inter-client similarity through Locality-Sensitive Hashing (LSH) and evaluating model quality based on peer rankings, WPFed enables clients to identify personalized optimal neighbors on a global scale while preserving data privacy. To enhance security and deter malicious behavior, WPFed integrates verification mechanisms for both LSH codes and performance rankings, leveraging blockchain-driven announcements to ensure transparency and verifiability. Through extensive experiments on multiple real-world datasets, we demonstrate that WPFed significantly improves learning outcomes and system robustness compared to traditional federated learning methods. Our findings highlight WPFed's potential to facilitate effective and secure decentralized collaborative learning across diverse and interconnected web environments.
☆ DODT: Enhanced Online Decision Transformer Learning through Dreamer's Actor-Critic Trajectory Forecasting
Advancements in reinforcement learning have led to the development of sophisticated models capable of learning complex decision-making tasks. However, efficiently integrating world models with decision transformers remains a challenge. In this paper, we introduce a novel approach that combines the Dreamer algorithm's ability to generate anticipatory trajectories with the adaptive learning strengths of the Online Decision Transformer. Our methodology enables parallel training where Dreamer-produced trajectories enhance the contextual decision-making of the transformer, creating a bidirectional enhancement loop. We empirically demonstrate the efficacy of our approach on a suite of challenging benchmarks, achieving notable improvements in sample efficiency and reward maximization over existing methods. Our results indicate that the proposed integrated framework not only accelerates learning but also showcases robustness in diverse and dynamic scenarios, marking a significant step forward in model-based reinforcement learning.
☆ Reducing Labeling Costs in Sentiment Analysis via Semi-Supervised Learning
Labeling datasets is a noteworthy challenge in machine learning, both in terms of cost and time. This research, however, leverages an efficient answer. By exploring label propagation in semi-supervised learning, we can significantly reduce the number of labels required compared to traditional methods. We employ a transductive label propagation method based on the manifold assumption for text classification. Our approach utilizes a graph-based method to generate pseudo-labels for unlabeled data for the text classification task, which are then used to train deep neural networks. By extending labels based on cosine proximity within a nearest neighbor graph from network embeddings, we combine unlabeled data into supervised learning, thereby reducing labeling costs. Based on previous successes in other domains, this study builds and evaluates this approach's effectiveness in sentiment analysis, presenting insights into semi-supervised learning.
comment: 12 pages, 7 figures, accepted at the 2024 8th International Conference on Natural Language Processing and Information Retrieval (NLPIR 2024), Okayama, Japan, 2024
☆ Toward a Well-Calibrated Discrimination via Survival Outcome-Aware Contrastive Learning NeurIPS 2024
Previous deep learning approaches for survival analysis have primarily relied on ranking losses to improve discrimination performance, which often comes at the expense of calibration performance. To address such an issue, we propose a novel contrastive learning approach specifically designed to enhance discrimination \textit{without} sacrificing calibration. Our method employs weighted sampling within a contrastive learning framework, assigning lower penalties to samples with similar survival outcomes. This aligns well with the assumption that patients with similar event times share similar clinical statuses. Consequently, when augmented with the commonly used negative log-likelihood loss, our approach significantly improves discrimination performance without directly manipulating the model outputs, thereby achieving better calibration. Experiments on multiple real-world clinical datasets demonstrate that our method outperforms state-of-the-art deep survival models in both discrimination and calibration. Through comprehensive ablation studies, we further validate the effectiveness of our approach through quantitative and qualitative analyses.
comment: Accepted at NeurIPS 2024
☆ DIAR: Diffusion-model-guided Implicit Q-learning with Adaptive Revaluation
We propose a novel offline reinforcement learning (offline RL) approach, introducing the Diffusion-model-guided Implicit Q-learning with Adaptive Revaluation (DIAR) framework. We address two key challenges in offline RL: out-of-distribution samples and long-horizon problems. We leverage diffusion models to learn state-action sequence distributions and incorporate value functions for more balanced and adaptive decision-making. DIAR introduces an Adaptive Revaluation mechanism that dynamically adjusts decision lengths by comparing current and future state values, enabling flexible long-term decision-making. Furthermore, we address Q-value overestimation by combining Q-network learning with a value function guided by a diffusion model. The diffusion model generates diverse latent trajectories, enhancing policy robustness and generalization. As demonstrated in tasks like Maze2D, AntMaze, and Kitchen, DIAR consistently outperforms state-of-the-art algorithms in long-horizon, sparse-reward environments.
comment: Preprint, under review. Comments welcome
☆ SHAKTI: A 2.5 Billion Parameter Small Language Model Optimized for Edge AI and Low-Resource Environments
We introduce Shakti, a 2.5 billion parameter language model specifically optimized for resource-constrained environments such as edge devices, including smartphones, wearables, and IoT systems. Shakti combines high-performance NLP with optimized efficiency and precision, making it ideal for real-time AI applications where computational resources and memory are limited. With support for vernacular languages and domain-specific tasks, Shakti excels in industries such as healthcare, finance, and customer service. Benchmark evaluations demonstrate that Shakti performs competitively against larger models while maintaining low latency and on-device efficiency, positioning it as a leading solution for edge AI.
comment: Paper in pdf format is 11 pages and contains 4 tables
☆ Evolutionary Retrofitting
AfterLearnER (After Learning Evolutionary Retrofitting) consists in applying non-differentiable optimization, including evolutionary methods, to refine fully-trained machine learning models by optimizing a set of carefully chosen parameters or hyperparameters of the model, with respect to some actual, exact, and hence possibly non-differentiable error signal, performed on a subset of the standard validation set. The efficiency of AfterLearnER is demonstrated by tackling non-differentiable signals such as threshold-based criteria in depth sensing, the word error rate in speech re-synthesis, image quality in 3D generative adversarial networks (GANs), image generation via Latent Diffusion Models (LDM), the number of kills per life at Doom, computational accuracy or BLEU in code translation, and human appreciations in image synthesis. In some cases, this retrofitting is performed dynamically at inference time by taking into account user inputs. The advantages of AfterLearnER are its versatility (no gradient is needed), the possibility to use non-differentiable feedback including human evaluations, the limited overfitting, supported by a theoretical study and its anytime behavior. Last but not least, AfterLearnER requires only a minimal amount of feedback, i.e., a few dozens to a few hundreds of scalars, rather than the tens of thousands needed in most related published works. Compared to fine-tuning (typically using the same loss, and gradient-based optimization on a smaller but still big dataset at a fine grain), AfterLearnER uses a minimum amount of data on the real objective function without requiring differentiability.
☆ Sequential LLM Framework for Fashion Recommendation
The fashion industry is one of the leading domains in the global e-commerce sector, prompting major online retailers to employ recommendation systems for product suggestions and customer convenience. While recommendation systems have been widely studied, most are designed for general e-commerce problems and struggle with the unique challenges of the fashion domain. To address these issues, we propose a sequential fashion recommendation framework that leverages a pre-trained large language model (LLM) enhanced with recommendation-specific prompts. Our framework employs parameter-efficient fine-tuning with extensive fashion data and introduces a novel mix-up-based retrieval technique for translating text into relevant product suggestions. Extensive experiments show our proposed framework significantly enhances fashion recommendation performance.
☆ Diffusion-Based Offline RL for Improved Decision-Making in Augmented ARC Task
Effective long-term strategies enable AI systems to navigate complex environments by making sequential decisions over extended horizons. Similarly, reinforcement learning (RL) agents optimize decisions across sequences to maximize rewards, even without immediate feedback. To verify that Latent Diffusion-Constrained Q-learning (LDCQ), a prominent diffusion-based offline RL method, demonstrates strong reasoning abilities in multi-step decision-making, we aimed to evaluate its performance on the Abstraction and Reasoning Corpus (ARC). However, applying offline RL methodologies to enhance strategic reasoning in AI for solving tasks in ARC is challenging due to the lack of sufficient experience data in the ARC training set. To address this limitation, we introduce an augmented offline RL dataset for ARC, called Synthesized Offline Learning Data for Abstraction and Reasoning (SOLAR), along with the SOLAR-Generator, which generates diverse trajectory data based on predefined rules. SOLAR enables the application of offline RL methods by offering sufficient experience data. We synthesized SOLAR for a simple task and used it to train an agent with the LDCQ method. Our experiments demonstrate the effectiveness of the offline RL approach on a simple ARC task, showing the agent's ability to make multi-step sequential decisions and correctly identify answer states. These results highlight the potential of the offline RL approach to enhance AI's strategic reasoning capabilities.
comment: Preprint, Under review. Comments welcome
☆ KA-GNN: Kolmogorov-Arnold Graph Neural Networks for Molecular Property Prediction
Molecular property prediction is a crucial task in the process of Artificial Intelligence-Driven Drug Discovery (AIDD). The challenge of developing models that surpass traditional non-neural network methods continues to be a vibrant area of research. This paper presents a novel graph neural network model-the Kolmogorov-Arnold Network (KAN)-based Graph Neural Network (KA-GNN), which incorporates Fourier series, specifically designed for molecular property prediction. This model maintains the high interpretability characteristic of KAN methods while being extremely efficient in computational resource usage, making it an ideal choice for deployment in resource-constrained environments. Tested and validated on seven public datasets, KA-GNN has shown significant improvements in property predictions over the existing state-of-the-art (SOTA) benchmarks.
☆ Deciphering the Chaos: Enhancing Jailbreak Attacks via Adversarial Prompt Translation
Automatic adversarial prompt generation provides remarkable success in jailbreaking safely-aligned large language models (LLMs). Existing gradient-based attacks, while demonstrating outstanding performance in jailbreaking white-box LLMs, often generate garbled adversarial prompts with chaotic appearance. These adversarial prompts are difficult to transfer to other LLMs, hindering their performance in attacking unknown victim models. In this paper, for the first time, we delve into the semantic meaning embedded in garbled adversarial prompts and propose a novel method that "translates" them into coherent and human-readable natural language adversarial prompts. In this way, we can effectively uncover the semantic information that triggers vulnerabilities of the model and unambiguously transfer it to the victim model, without overlooking the adversarial information hidden in the garbled text, to enhance jailbreak attacks. It also offers a new approach to discovering effective designs for jailbreak prompts, advancing the understanding of jailbreak attacks. Experimental results demonstrate that our method significantly improves the success rate of jailbreak attacks against various safety-aligned LLMs and outperforms state-of-the-arts by large margins. With at most 10 queries, our method achieves an average attack success rate of 81.8% in attacking 7 commercial closed-source LLMs, including GPT and Claude-3 series, on HarmBench. Our method also achieves over 90% attack success rates against Llama-2-Chat models on AdvBench, despite their outstanding resistance to jailbreak attacks. Code at: https://github.com/qizhangli/Adversarial-Prompt-Translator.
☆ Communication-Control Codesign for Large-Scale Wireless Networked Control Systems
Wireless Networked Control Systems (WNCSs) are essential to Industry 4.0, enabling flexible control in applications, such as drone swarms and autonomous robots. The interdependence between communication and control requires integrated design, but traditional methods treat them separately, leading to inefficiencies. Current codesign approaches often rely on simplified models, focusing on single-loop or independent multi-loop systems. However, large-scale WNCSs face unique challenges, including coupled control loops, time-correlated wireless channels, trade-offs between sensing and control transmissions, and significant computational complexity. To address these challenges, we propose a practical WNCS model that captures correlated dynamics among multiple control loops with spatially distributed sensors and actuators sharing limited wireless resources over multi-state Markov block-fading channels. We formulate the codesign problem as a sequential decision-making task that jointly optimizes scheduling and control inputs across estimation, control, and communication domains. To solve this problem, we develop a Deep Reinforcement Learning (DRL) algorithm that efficiently handles the hybrid action space, captures communication-control correlations, and ensures robust training despite sparse cross-domain variables and floating control inputs. Extensive simulations show that the proposed DRL approach outperforms benchmarks and solves the large-scale WNCS codesign problem, providing a scalable solution for industrial automation.
comment: This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible
☆ Towards Differentiable Multilevel Optimization: A Gradient-Based Approach
Multilevel optimization has gained renewed interest in machine learning due to its promise in applications such as hyperparameter tuning and continual learning. However, existing methods struggle with the inherent difficulty of efficiently handling the nested structure. This paper introduces a novel gradient-based approach for multilevel optimization that overcomes these limitations by leveraging a hierarchically structured decomposition of the full gradient and employing advanced propagation techniques. Extending to n-level scenarios, our method significantly reduces computational complexity while improving both solution accuracy and convergence speed. We demonstrate the effectiveness of our approach through numerical experiments, comparing it with existing methods across several benchmarks. The results show a notable improvement in solution accuracy. To the best of our knowledge, this is one of the first algorithms to provide a general version of implicit differentiation with both theoretical guarantees and superior empirical performance.
comment: 18 pages
☆ QSpec: Speculative Decoding with Complementary Quantization Schemes
Quantization has been substantially adopted to accelerate inference and reduce memory consumption of large language models (LLMs). While activation-weight joint quantization speeds up the inference process through low-precision kernels, we demonstrate that it suffers severe performance degradation on multi-step reasoning tasks, rendering it ineffective. We propose a novel quantization paradigm called QSPEC, which seamlessly integrates two complementary quantization schemes for speculative decoding. Leveraging nearly cost-free execution switching, QSPEC drafts tokens with low-precision, fast activation-weight quantization, and verifies them with high-precision weight-only quantization, effectively combining the strengths of both quantization schemes. Compared to high-precision quantization methods, QSPEC empirically boosts token generation throughput by up to 1.80x without any quality compromise, distinguishing it from other low-precision quantization approaches. This enhancement is also consistent across various serving tasks, model sizes, quantization methods, and batch sizes. Unlike existing speculative decoding techniques, our approach reuses weights and the KV cache, avoiding additional memory overhead. Furthermore, QSPEC offers a plug-and-play advantage without requiring any training. We believe that QSPEC demonstrates unique strengths for future deployment of high-fidelity quantization schemes, particularly in memory-constrained scenarios (e.g., edge devices).
☆ Data Selection for Task-Specific Model Finetuning
Finetuning foundation models for specific tasks is an emerging paradigm in modern machine learning. The efficacy of task-specific finetuning largely depends on the selection of appropriate training data. We present a framework to select data for task-specific model finetuning, guided by a small but representative set of examples from the target task. To do so, we formulate data selection for task-specific finetuning as an optimization problem with a distribution alignment loss based on optimal transport to capture the discrepancy between the selected data and the target distribution. In addition, we add a regularizer to encourage the diversity of the selected data and incorporate kernel density estimation into the regularizer to reduce the negative effects of near-duplicates among the candidate data. We connect our optimization problem to nearest neighbor search and design efficient algorithms to compute the optimal solution based on approximate nearest neighbor search techniques. We evaluate our method on data selection for both continued pretraining and instruction tuning of language models. We show that instruction tuning using data selected by our method with a 1% selection ratio often outperforms using the full dataset and beats the baseline selection methods by 1.5 points in F1 score on average.
comment: 31 pages, 1 figure
☆ Have the VLMs Lost Confidence? A Study of Sycophancy in VLMs
In the study of LLMs, sycophancy represents a prevalent hallucination that poses significant challenges to these models. Specifically, LLMs often fail to adhere to original correct responses, instead blindly agreeing with users' opinions, even when those opinions are incorrect or malicious. However, research on sycophancy in visual language models (VLMs) has been scarce. In this work, we extend the exploration of sycophancy from LLMs to VLMs, introducing the MM-SY benchmark to evaluate this phenomenon. We present evaluation results from multiple representative models, addressing the gap in sycophancy research for VLMs. To mitigate sycophancy, we propose a synthetic dataset for training and employ methods based on prompts, supervised fine-tuning, and DPO. Our experiments demonstrate that these methods effectively alleviate sycophancy in VLMs. Additionally, we probe VLMs to assess the semantic impact of sycophancy and analyze the attention distribution of visual tokens. Our findings indicate that the ability to prevent sycophancy is predominantly observed in higher layers of the model. The lack of attention to image knowledge in these higher layers may contribute to sycophancy, and enhancing image attention at high layers proves beneficial in mitigating this issue.
☆ Sorted Weight Sectioning for Energy-Efficient Unstructured Sparse DNNs on Compute-in-Memory Crossbars
We introduce $\textit{sorted weight sectioning}$ (SWS): a weight allocation algorithm that places sorted deep neural network (DNN) weight sections on bit-sliced compute-in-memory (CIM) crossbars to reduce analog-to-digital converter (ADC) energy consumption. Data conversions are the most energy-intensive process in crossbar operation. SWS effectively reduces this cost leveraging (1) small weights and (2) zero weights (weight sparsity). DNN weights follow bell-shaped distributions, with most weights near zero. Using SWS, we only need low-order crossbar columns for sections with low-magnitude weights. This reduces the quantity and resolution of ADCs used, exponentially decreasing ADC energy costs without significantly degrading DNN accuracy. Unstructured sparsification further sharpens the weight distribution with small accuracy loss. However, it presents challenges in hardware tracking of zeros: we cannot switch zero rows to other layer weights in unsorted crossbars without index matching. SWS efficiently addresses unstructured sparse models using offline remapping of zeros into earlier sections, which reveals full sparsity potential and maximizes energy efficiency. Our method reduces ADC energy use by 89.5% on unstructured sparse BERT models. Overall, this paper introduces a novel algorithm to promote energy-efficient CIM crossbars for unstructured sparse DNN workloads.
comment: 5 pages, 4 figures
☆ TraM : Enhancing User Sleep Prediction with Transformer-based Multivariate Time Series Modeling and Machine Learning Ensembles
This paper presents a novel approach that leverages Transformer-based multivariate time series model and Machine Learning Ensembles to predict the quality of human sleep, emotional states, and stress levels. A formula to calculate the labels was developed, and the various models were applied to user data. Time Series Transformer was used for labels where time series characteristics are crucial, while Machine Learning Ensembles were employed for labels requiring comprehensive daily activity statistics. Time Series Transformer excels in capturing the characteristics of time series through pre-training, while Machine Learning Ensembles select machine learning models that meet our categorization criteria. The proposed model, TraM, scored 6.10 out of 10 in experiments, demonstrating superior performance compared to other methodologies. The code and configuration for the TraM framework are available at: https://github.com/jin-jae/ETRI-Paper-Contest.
☆ Backdoor Attack on Vertical Federated Graph Neural Network Learning
Federated Graph Neural Network (FedGNN) is a privacy-preserving machine learning technology that combines federated learning (FL) and graph neural networks (GNNs). It offers a privacy-preserving solution for training GNNs using isolated graph data. Vertical Federated Graph Neural Network (VFGNN) is an important branch of FedGNN, where data features and labels are distributed among participants, and each participant has the same sample space. Due to the difficulty of accessing and modifying distributed data and labels, the vulnerability of VFGNN to backdoor attacks remains largely unexplored. In this context, we propose BVG, the first method for backdoor attacks in VFGNN. Without accessing or modifying labels, BVG uses multi-hop triggers and requires only four target class nodes for an effective backdoor attack. Experiments show that BVG achieves high attack success rates (ASR) across three datasets and three different GNN models, with minimal impact on main task accuracy (MTA). We also evaluate several defense methods, further validating the robustness and effectiveness of BVG. This finding also highlights the need for advanced defense mechanisms to counter sophisticated backdoor attacks in practical VFGNN applications.
☆ Subspace Optimization for Large Language Models with Convergence Guarantees
Subspace optimization algorithms, with GaLore (Zhao et al., 2024) as a representative method, have gained popularity for pre-training or fine-tuning large language models (LLMs) due to their memory efficiency. However, their convergence guarantees remain unclear, particularly in stochastic settings. In this paper, we unexpectedly discover that GaLore does not always converge to the optimal solution and substantiate this finding with an explicit counterexample. We then investigate the conditions under which GaLore can achieve convergence, demonstrating that it does so either in deterministic scenarios or when using a sufficiently large mini-batch size. More significantly, we introduce GoLore (Gradient random Low-rank projection), a novel variant of GaLore that provably converges in stochastic settings, even with standard batch sizes. Our convergence analysis can be readily extended to other sparse subspace optimization algorithms. Finally, we conduct numerical experiments to validate our theoretical results and empirically explore the proposed mechanisms. Codes are available at https://github.com/pkumelon/Golore.
☆ Calabi-Yau metrics through Grassmannian learning and Donaldson's algorithm
Motivated by recent progress in the problem of numerical K\"ahler metrics, we survey machine learning techniques in this area, discussing both advantages and drawbacks. We then revisit the algebraic ansatz pioneered by Donaldson. Inspired by his work, we present a novel approach to obtaining Ricci-flat approximations to K\"ahler metrics, applying machine learning within a `principled' framework. In particular, we use gradient descent on the Grassmannian manifold to identify an efficient subspace of sections for calculation of the metric. We combine this approach with both Donaldson's algorithm and learning on the $h$-matrix itself (the latter method being equivalent to gradient descent on the fibre bundle of Hermitian metrics on the tautological bundle over the Grassmannian). We implement our methods on the Dwork family of threefolds, commenting on the behaviour at different points in moduli space. In particular, we observe the emergence of nontrivial local minima as the moduli parameter is increased.
comment: 38 pages (including references), 3 figures, 1 appendix
☆ AdvBDGen: Adversarially Fortified Prompt-Specific Fuzzy Backdoor Generator Against LLM Alignment
With the growing adoption of reinforcement learning with human feedback (RLHF) for aligning large language models (LLMs), the risk of backdoor installation during alignment has increased, leading to unintended and harmful behaviors. Existing backdoor triggers are typically limited to fixed word patterns, making them detectable during data cleaning and easily removable post-poisoning. In this work, we explore the use of prompt-specific paraphrases as backdoor triggers, enhancing their stealth and resistance to removal during LLM alignment. We propose AdvBDGen, an adversarially fortified generative fine-tuning framework that automatically generates prompt-specific backdoors that are effective, stealthy, and transferable across models. AdvBDGen employs a generator-discriminator pair, fortified by an adversary, to ensure the installability and stealthiness of backdoors. It enables the crafting and successful installation of complex triggers using as little as 3% of the fine-tuning data. Once installed, these backdoors can jailbreak LLMs during inference, demonstrate improved stability against perturbations compared to traditional constant triggers, and are more challenging to remove. These findings underscore an urgent need for the research community to develop more robust defenses against adversarial backdoor threats in LLM alignment.
comment: Published at the Neurips Safe Generative AI Workshop 2024
☆ Advancing the Understanding of Fixed Point Iterations in Deep Neural Networks: A Detailed Analytical Study
Recent empirical studies have identified fixed point iteration phenomena in deep neural networks, where the hidden state tends to stabilize after several layers, showing minimal change in subsequent layers. This observation has spurred the development of practical methodologies, such as accelerating inference by bypassing certain layers once the hidden state stabilizes, selectively fine-tuning layers to modify the iteration process, and implementing loops of specific layers to maintain fixed point iterations. Despite these advancements, the understanding of fixed point iterations remains superficial, particularly in high-dimensional spaces, due to the inadequacy of current analytical tools. In this study, we conduct a detailed analysis of fixed point iterations in a vector-valued function modeled by neural networks. We establish a sufficient condition for the existence of multiple fixed points of looped neural networks based on varying input regions. Additionally, we expand our examination to include a robust version of fixed point iterations. To demonstrate the effectiveness and insights provided by our approach, we provide case studies that looped neural networks may exist $2^d$ number of robust fixed points under exponentiation or polynomial activation functions, where $d$ is the feature dimension. Furthermore, our preliminary empirical results support our theoretical findings. Our methodology enriches the toolkit available for analyzing fixed point iterations of deep neural networks and may enhance our comprehension of neural network mechanisms.
☆ UmambaTSF: A U-shaped Multi-Scale Long-Term Time Series Forecasting Method Using Mamba
Multivariate Time series forecasting is crucial in domains such as transportation, meteorology, and finance, especially for predicting extreme weather events. State-of-the-art methods predominantly rely on Transformer architectures, which utilize attention mechanisms to capture temporal dependencies. However, these methods are hindered by quadratic time complexity, limiting the model's scalability with respect to input sequence length. This significantly restricts their practicality in the real world. Mamba, based on state space models (SSM), provides a solution with linear time complexity, increasing the potential for efficient forecasting of sequential data. In this study, we propose UmambaTSF, a novel long-term time series forecasting framework that integrates multi-scale feature extraction capabilities of U-shaped encoder-decoder multilayer perceptrons (MLP) with Mamba's long sequence representation. To improve performance and efficiency, the Mamba blocks introduced in the framework adopt a refined residual structure and adaptable design, enabling the capture of unique temporal signals and flexible channel processing. In the experiments, UmambaTSF achieves state-of-the-art performance and excellent generality on widely used benchmark datasets while maintaining linear time complexity and low memory consumption.
☆ ILAEDA: An Imitation Learning Based Approach for Automatic Exploratory Data Analysis
Automating end-to-end Exploratory Data Analysis (AutoEDA) is a challenging open problem, often tackled through Reinforcement Learning (RL) by learning to predict a sequence of analysis operations (FILTER, GROUP, etc). Defining rewards for each operation is a challenging task and existing methods rely on various \emph{interestingness measures} to craft reward functions to capture the importance of each operation. In this work, we argue that not all of the essential features of what makes an operation important can be accurately captured mathematically using rewards. We propose an AutoEDA model trained through imitation learning from expert EDA sessions, bypassing the need for manually defined interestingness measures. Our method, based on generative adversarial imitation learning (GAIL), generalizes well across datasets, even with limited expert data. We also introduce a novel approach for generating synthetic EDA demonstrations for training. Our method outperforms the existing state-of-the-art end-to-end EDA approach on benchmarks by upto 3x, showing strong performance and generalization, while naturally capturing diverse interestingness measures in generated EDA sessions.
comment: Accepted at AIMLSystems '24
☆ Shallow diffusion networks provably learn hidden low-dimensional structure
Diffusion-based generative models provide a powerful framework for learning to sample from a complex target distribution. The remarkable empirical success of these models applied to high-dimensional signals, including images and video, stands in stark contrast to classical results highlighting the curse of dimensionality for distribution recovery. In this work, we take a step towards understanding this gap through a careful analysis of learning diffusion models over the Barron space of single layer neural networks. In particular, we show that these shallow models provably adapt to simple forms of low dimensional structure, thereby avoiding the curse of dimensionality. We combine our results with recent analyses of sampling with diffusion models to provide an end-to-end sample complexity bound for learning to sample from structured distributions. Importantly, our results do not require specialized architectures tailored to particular latent structures, and instead rely on the low-index structure of the Barron space to adapt to the underlying distribution.
☆ Reducing Source-Private Bias in Extreme Universal Domain Adaptation
Universal Domain Adaptation (UniDA) aims to transfer knowledge from a labeled source domain to an unlabeled target domain without assuming how much the label-sets of the two domains intersect. The goal of UniDA is to achieve robust performance on the target domain across different intersection levels. However, existing literature has not sufficiently explored performance under extreme intersection levels. Our experiments reveal that state-of-the-art methods struggle when the source domain has significantly more non-overlapping classes than overlapping ones, a setting we refer to as Extreme UniDA. In this paper, we demonstrate that classical partial domain alignment, which focuses on aligning only overlapping-class data between domains, is limited in mitigating the bias of feature extractors toward source-private classes in extreme UniDA scenarios. We argue that feature extractors trained with source supervised loss distort the intrinsic structure of the target data due to the inherent differences between source-private classes and the target data. To mitigate this bias, we propose using self-supervised learning to preserve the structure of the target data. Our approach can be easily integrated into existing frameworks. We apply the proposed approach to two distinct training paradigms-adversarial-based and optimal-transport-based-and show consistent improvements across various intersection levels, with significant gains in extreme UniDA settings.
☆ Bypassing the Exponential Dependency: Looped Transformers Efficiently Learn In-context by Multi-step Gradient Descent
In-context learning has been recognized as a key factor in the success of Large Language Models (LLMs). It refers to the model's ability to learn patterns on the fly from provided in-context examples in the prompt during inference. Previous studies have demonstrated that the Transformer architecture used in LLMs can implement a single-step gradient descent update by processing in-context examples in a single forward pass. Recent work has further shown that, during in-context learning, a looped Transformer can implement multi-step gradient descent updates in forward passes. However, their theoretical results require an exponential number of in-context examples, $n = \exp(\Omega(T))$, where $T$ is the number of loops or passes, to achieve a reasonably low error. In this paper, we study linear looped Transformers in-context learning on linear vector generation tasks. We show that linear looped Transformers can implement multi-step gradient descent efficiently for in-context learning. Our results demonstrate that as long as the input data has a constant condition number, e.g., $n = O(d)$, the linear looped Transformers can achieve a small error by multi-step gradient descent during in-context learning. Furthermore, our preliminary experiments validate our theoretical analysis. Our findings reveal that the Transformer architecture possesses a stronger in-context learning capability than previously understood, offering new insights into the mechanisms behind LLMs and potentially guiding the better design of efficient inference algorithms for LLMs.
☆ FedCCRL: Federated Domain Generalization with Cross-Client Representation Learning
Domain Generalization (DG) aims to train models that can effectively generalize to unseen domains. However, in the context of Federated Learning (FL), where clients collaboratively train a model without directly sharing their data, most existing DG algorithms are not directly applicable to the FL setting due to privacy constraints, as well as the limited data quantity and domain diversity at each client. To tackle these challenges, we propose FedCCRL, a novel federated domain generalization method that significantly improves the model's ability to generalize to unseen domains without compromising privacy or incurring excessive computational and communication costs. Specifically, we adapt MixStyle to the federated setting to transfer domain-specific features while AugMix is employed to perturb domain-invariant features. Furthermore, we leverage supervised contrastive loss for representation alignment and utilize Jensen-Shannon divergence to ensure consistent predictions between original and augmented samples. Extensive experimental results demonstrate that FedCCRL achieves the state-of-the-art performances on the PACS, OfficeHome and miniDomainNet datasets across varying numbers of clients. Code is available at https://github.com/SanphouWang/FedCCRL.
☆ Unveiling Options with Neural Decomposition ICLR 2024
In reinforcement learning, agents often learn policies for specific tasks without the ability to generalize this knowledge to related tasks. This paper introduces an algorithm that attempts to address this limitation by decomposing neural networks encoding policies for Markov Decision Processes into reusable sub-policies, which are used to synthesize temporally extended actions, or options. We consider neural networks with piecewise linear activation functions, so that they can be mapped to an equivalent tree that is similar to oblique decision trees. Since each node in such a tree serves as a function of the input of the tree, each sub-tree is a sub-policy of the main policy. We turn each of these sub-policies into options by wrapping it with while-loops of varied number of iterations. Given the large number of options, we propose a selection mechanism based on minimizing the Levin loss for a uniform policy on these options. Empirical results in two grid-world domains where exploration can be difficult confirm that our method can identify useful options, thereby accelerating the learning process on similar but different tasks.
comment: Published as a conference paper at ICLR 2024
☆ Beyond Linear Approximations: A Novel Pruning Approach for Attention Matrix
Large Language Models (LLMs) have shown immense potential in enhancing various aspects of our daily lives, from conversational AI to search and AI assistants. However, their growing capabilities come at the cost of extremely large model sizes, making deployment on edge devices challenging due to memory and computational constraints. This paper introduces a novel approach to LLM weight pruning that directly optimizes for approximating the attention matrix, a core component of transformer architectures. Unlike existing methods that focus on linear approximations, our approach accounts for the non-linear nature of the Softmax attention mechanism. We provide theoretical guarantees for the convergence of our Gradient Descent-based optimization method to a near-optimal pruning mask solution. Our preliminary empirical results demonstrate the effectiveness of this approach in maintaining model performance while significantly reducing computational costs. This work establishes a new theoretical foundation for pruning algorithm design in LLMs, potentially paving the way for more efficient LLM inference on resource-constrained devices.
☆ Disentangled Unsupervised Skill Discovery for Efficient Hierarchical Reinforcement Learning NeurIPS2024
A hallmark of intelligent agents is the ability to learn reusable skills purely from unsupervised interaction with the environment. However, existing unsupervised skill discovery methods often learn entangled skills where one skill variable simultaneously influences many entities in the environment, making downstream skill chaining extremely challenging. We propose Disentangled Unsupervised Skill Discovery (DUSDi), a method for learning disentangled skills that can be efficiently reused to solve downstream tasks. DUSDi decomposes skills into disentangled components, where each skill component only affects one factor of the state space. Importantly, these skill components can be concurrently composed to generate low-level actions, and efficiently chained to tackle downstream tasks through hierarchical Reinforcement Learning. DUSDi defines a novel mutual-information-based objective to enforce disentanglement between the influences of different skill components, and utilizes value factorization to optimize this objective efficiently. Evaluated in a set of challenging environments, DUSDi successfully learns disentangled skills, and significantly outperforms previous skill discovery methods when it comes to applying the learned skills to solve downstream tasks. Code and skills visualization at jiahenghu.github.io/DUSDi-site/.
comment: NeurIPS2024
☆ Learning Agents With Prioritization and Parameter Noise in Continuous State and Action Space
Among the many variants of RL, an important class of problems is where the state and action spaces are continuous -- autonomous robots, autonomous vehicles, optimal control are all examples of such problems that can lend themselves naturally to reinforcement based algorithms, and have continuous state and action spaces. In this paper, we introduce a prioritized form of a combination of state-of-the-art approaches such as Deep Q-learning (DQN) and Deep Deterministic Policy Gradient (DDPG) to outperform the earlier results for continuous state and action space problems. Our experiments also involve the use of parameter noise during training resulting in more robust deep RL models outperforming the earlier results significantly. We believe these results are a valuable addition for continuous state and action space problems.
comment: 10 pages, 3 figures. Published in Advances in Neural Networks - ISNN 2019
☆ A Unified Framework for Forward and Inverse Problems in Subsurface Imaging using Latent Space Translations
In subsurface imaging, learning the mapping from velocity maps to seismic waveforms (forward problem) and waveforms to velocity (inverse problem) is important for several applications. While traditional techniques for solving forward and inverse problems are computationally prohibitive, there is a growing interest in leveraging recent advances in deep learning to learn the mapping between velocity maps and seismic waveform images directly from data. Despite the variety of architectures explored in previous works, several open questions still remain unanswered such as the effect of latent space sizes, the importance of manifold learning, the complexity of translation models, and the value of jointly solving forward and inverse problems. We propose a unified framework to systematically characterize prior research in this area termed the Generalized Forward-Inverse (GFI) framework, building on the assumption of manifolds and latent space translations. We show that GFI encompasses previous works in deep learning for subsurface imaging, which can be viewed as specific instantiations of GFI. We also propose two new model architectures within the framework of GFI: Latent U-Net and Invertible X-Net, leveraging the power of U-Nets for domain translation and the ability of IU-Nets to simultaneously learn forward and inverse translations, respectively. We show that our proposed models achieve state-of-the-art (SOTA) performance for forward and inverse problems on a wide range of synthetic datasets, and also investigate their zero-shot effectiveness on two real-world-like datasets.
☆ Automatically Generating Visual Hallucination Test Cases for Multimodal Large Language Models
Visual hallucination (VH) occurs when a multimodal large language model (MLLM) generates responses with incorrect visual details for prompts. Existing methods for generating VH test cases primarily rely on human annotations, typically in the form of triples: (image, question, answer). In this paper, we introduce VHExpansion, the first automated method for expanding VH test cases for MLLMs. Given an initial VH test case, VHExpansion automatically expands it by perturbing the question and answer through negation as well as modifying the image using both common and adversarial perturbations. Additionally, we propose a new evaluation metric, symmetric accuracy, which measures the proportion of correctly answered VH test-case pairs. Each pair consists of a test case and its negated counterpart. Our theoretical analysis shows that symmetric accuracy is an unbiased evaluation metric that remains unaffected by the imbalance of VH testing cases with varying answers when an MLLM is randomly guessing the answers, whereas traditional accuracy is prone to such imbalance. We apply VHExpansion to expand three VH datasets annotated manually and use these expanded datasets to benchmark seven MLLMs. Our evaluation shows that VHExpansion effectively identifies more VH test cases. Moreover, symmetric accuracy, being unbiased, leads to different conclusions about the vulnerability of MLLMs to VH compared to traditional accuracy metric. Finally, we show that fine-tuning MLLMs on the expanded VH dataset generated by VHExpansion mitigates VH more effectively than fine-tuning on the original, manually annotated dataset. Our code is available at: https://github.com/lycheeefish/VHExpansion.
♻ ☆ LoRA-Pro: Are Low-Rank Adapters Properly Optimized?
Low-rank adaptation, also known as LoRA, has emerged as a prominent method for parameter-efficient fine-tuning of foundation models. Despite its computational efficiency, LoRA still yields inferior performance compared to full fine-tuning. In this paper, we first uncover a fundamental connection between the optimization processes of LoRA and full fine-tuning: using LoRA for optimization is mathematically equivalent to full fine-tuning using a low-rank gradient for parameter updates. And this low-rank gradient can be expressed in terms of the gradients of the two low-rank matrices in LoRA. Leveraging this insight, we introduce LoRA-Pro, a method that enhances LoRA's performance by strategically adjusting the gradients of these low-rank matrices. This adjustment allows the low-rank gradient to more accurately approximate the full fine-tuning gradient, thereby narrowing the performance gap between LoRA and full fine-tuning. Furthermore, we theoretically derive the optimal solutions for adjusting the gradients of the low-rank matrices, applying them during fine-tuning in LoRA-Pro. We conduct extensive experiments across natural language understanding, dialogue generation, mathematical reasoning, code generation, and image classification tasks, demonstrating that LoRA-Pro substantially improves LoRA's performance, effectively narrowing the gap with full fine-tuning. Code is publicly available at \url{https://github.com/mrflogs/LoRA-Pro}.
♻ ☆ TemporalBench: Benchmarking Fine-grained Temporal Understanding for Multimodal Video Models
Understanding fine-grained temporal dynamics is crucial for multimodal video comprehension and generation. Due to the lack of fine-grained temporal annotations, existing video benchmarks mostly resemble static image benchmarks and are incompetent at evaluating models for temporal understanding. In this paper, we introduce TemporalBench, a new benchmark dedicated to evaluating fine-grained temporal understanding in videos. TemporalBench consists of ~10K video question-answer pairs, derived from ~2K high-quality human annotations detailing the temporal dynamics in video clips. As a result, our benchmark provides a unique testbed for evaluating various temporal understanding and reasoning abilities such as action frequency, motion magnitude, event order, etc. Moreover, it enables evaluations on various tasks like both video question answering and captioning, both short and long video understanding, as well as different models such as multimodal video embedding models and text generation models. Results show that state-of-the-art models like GPT-4o achieve only 38.5% question answering accuracy on TemporalBench, demonstrating a significant gap (~30%) between humans and AI in temporal understanding. Furthermore, we notice a critical pitfall for multi-choice QA where LLMs can detect the subtle changes in negative captions and find a centralized description as a cue for its prediction, where we propose Multiple Binary Accuracy (MBA) to correct such bias. We hope that TemporalBench can foster research on improving models' temporal reasoning capabilities. Both dataset and evaluation code will be made available.
comment: Project Page: https://temporalbench.github.io/
♻ ☆ A Novel Gaussian Min-Max Theorem and its Applications
A celebrated result by Gordon allows one to compare the min-max behavior of two Gaussian processes if certain inequality conditions are met. The consequences of this result include the Gaussian min-max (GMT) and convex Gaussian min-max (CGMT) theorems which have had far-reaching implications in high-dimensional statistics, machine learning, non-smooth optimization, and signal processing. Both theorems rely on a pair of Gaussian processes, first identified by Slepian, that satisfy Gordon's comparison inequalities. In this paper, we identify such a new pair. The resulting theorems extend the classical GMT and CGMT Theorems from the case where the underlying Gaussian matrix in the primary process has iid rows to where it has independent but non-identically-distributed ones. The new CGMT is applied to the problems of multi-source Gaussian regression, as well as to binary classification of general Gaussian mixture models.
comment: Added more references to related works
♻ ☆ LoRD: Adapting Differentiable Driving Policies to Distribution Shifts
Distribution shifts between operational domains can severely affect the performance of learned models in self-driving vehicles (SDVs). While this is a well-established problem, prior work has mostly explored naive solutions such as fine-tuning, focusing on the motion prediction task. In this work, we explore novel adaptation strategies for differentiable autonomy stacks consisting of prediction, planning, and control, perform evaluation in closed-loop, and investigate the often-overlooked issue of catastrophic forgetting. Specifically, we introduce two simple yet effective techniques: a low-rank residual decoder (LoRD) and multi-task fine-tuning. Through experiments across three models conducted on two real-world autonomous driving datasets (nuPlan, exiD), we demonstrate the effectiveness of our methods and highlight a significant performance gap between open-loop and closed-loop evaluation in prior approaches. Our approach improves forgetting by up to 23.33% and the closed-loop OOD driving score by 8.83% in comparison to standard fine-tuning.
comment: Under Review
♻ ☆ Curriculum effects and compositionality emerge with in-context learning in neural networks
Human learning embodies a striking duality: sometimes, we appear capable of following logical, compositional rules and benefit from structured curricula (e.g., in formal education), while other times, we rely on an incremental approach or trial-and-error, learning better from curricula that are unstructured or randomly interleaved. Influential psychological theories explain this seemingly disparate behavioral evidence by positing two qualitatively different learning systems -- one for rapid, rule-based inferences and another for slow, incremental adaptation. It remains unclear how to reconcile such theories with neural networks, which learn via incremental weight updates and are thus a natural model for the latter type of learning, but are not obviously compatible with the former. However, recent evidence suggests that both metalearning neural networks and large language models are capable of "in-context learning" (ICL) -- the ability to flexibly grasp the structure of a new task from a few examples given at inference time. Here, we show that networks capable of ICL can reproduce human-like learning and compositional behavior on rule-governed tasks, while at the same time replicating human behavioral phenomena in tasks lacking rule-like structure via their usual in-weight learning (IWL). Our work shows how emergent ICL can equip neural networks with fundamentally different learning properties than those traditionally attributed to them, and that these can coexist with the properties of their native IWL, thus offering a novel perspective on dual-process theories and human cognitive flexibility.
comment: 27 pages (including appendix), 10 figures, 7 tables. Previous version accepted as a talk + full paper at CogSci 2024
♻ ☆ Predicting from Strings: Language Model Embeddings for Bayesian Optimization
Bayesian Optimization is ubiquitous in the field of experimental design and blackbox optimization for improving search efficiency, but has been traditionally restricted to regression models which are only applicable to fixed search spaces and tabular input features. We propose Embed-then-Regress, a paradigm for applying in-context regression over string inputs, through the use of string embedding capabilities of pretrained language models. By expressing all inputs as strings, we are able to perform general-purpose regression for Bayesian Optimization over various domains including synthetic, combinatorial, and hyperparameter optimization, obtaining comparable results to state-of-the-art Gaussian Process-based algorithms. Code can be found at https://github.com/google-research/optformer/tree/main/optformer/embed_then_regress.
♻ ☆ Active Label Refinement for Robust Training of Imbalanced Medical Image Classification Tasks in the Presence of High Label Noise MICCAI 2024
The robustness of supervised deep learning-based medical image classification is significantly undermined by label noise. Although several methods have been proposed to enhance classification performance in the presence of noisy labels, they face some challenges: 1) a struggle with class-imbalanced datasets, leading to the frequent overlooking of minority classes as noisy samples; 2) a singular focus on maximizing performance using noisy datasets, without incorporating experts-in-the-loop for actively cleaning the noisy labels. To mitigate these challenges, we propose a two-phase approach that combines Learning with Noisy Labels (LNL) and active learning. This approach not only improves the robustness of medical image classification in the presence of noisy labels, but also iteratively improves the quality of the dataset by relabeling the important incorrect labels, under a limited annotation budget. Furthermore, we introduce a novel Variance of Gradients approach in LNL phase, which complements the loss-based sample selection by also sampling under-represented samples. Using two imbalanced noisy medical classification datasets, we demonstrate that that our proposed technique is superior to its predecessors at handling class imbalance by not misidentifying clean samples from minority classes as mostly noisy samples.
comment: Accepted at MICCAI 2024
♻ ☆ AGaLiTe: Approximate Gated Linear Transformers for Online Reinforcement Learning
In this paper we investigate transformer architectures designed for partially observable online reinforcement learning. The self-attention mechanism in the transformer architecture is capable of capturing long-range dependencies and it is the main reason behind its effectiveness in processing sequential data. Nevertheless, despite their success, transformers have two significant drawbacks that still limit their applicability in online reinforcement learning: (1) in order to remember all past information, the self-attention mechanism requires access to the whole history to be provided as context. (2) The inference cost in transformers is expensive. In this paper, we introduce recurrent alternatives to the transformer self-attention mechanism that offer context-independent inference cost, leverage long-range dependencies effectively, and performs well in online reinforcement learning task. We quantify the impact of the different components of our architecture in a diagnostic environment and assess performance gains in 2D and 3D pixel-based partially-observable environments (e.g. T-Maze, Mystery Path, Craftax, and Memory Maze). Compared with a state-of-the-art architecture, GTrXL, inference in our approach is at least 40% cheaper while reducing memory use more than 50%. Our approach either performs similarly or better than GTrXL, improving more than 37% upon GTrXL performance in harder tasks.
comment: Published in Transactions on Machine Learning Research
♻ ☆ Teuken-7B-Base & Teuken-7B-Instruct: Towards European LLMs
We present two multilingual LLMs designed to embrace Europe's linguistic diversity by supporting all 24 official languages of the European Union. Trained on a dataset comprising around 60% non-English data and utilizing a custom multilingual tokenizer, our models address the limitations of existing LLMs that predominantly focus on English or a few high-resource languages. We detail the models' development principles, i.e., data composition, tokenizer optimization, and training methodologies. The models demonstrate competitive performance across multilingual benchmarks, as evidenced by their performance on European versions of ARC, HellaSwag, MMLU, and TruthfulQA.
♻ ☆ Divide And Conquer: Learning Chaotic Dynamical Systems With Multistep Penalty Neural Ordinary Differential Equations
Forecasting high-dimensional dynamical systems is a fundamental challenge in various fields, such as geosciences and engineering. Neural Ordinary Differential Equations (NODEs), which combine the power of neural networks and numerical solvers, have emerged as a promising algorithm for forecasting complex nonlinear dynamical systems. However, classical techniques used for NODE training are ineffective for learning chaotic dynamical systems. In this work, we propose a novel NODE-training approach that allows for robust learning of chaotic dynamical systems. Our method addresses the challenges of non-convexity and exploding gradients associated with underlying chaotic dynamics. Training data trajectories from such systems are split into multiple, non-overlapping time windows. In addition to the deviation from the training data, the optimization loss term further penalizes the discontinuities of the predicted trajectory between the time windows. The window size is selected based on the fastest Lyapunov time scale of the system. Multi-step penalty(MP) method is first demonstrated on Lorenz equation, to illustrate how it improves the loss landscape and thereby accelerates the optimization convergence. MP method can optimize chaotic systems in a manner similar to least-squares shadowing with significantly lower computational costs. Our proposed algorithm, denoted the Multistep Penalty NODE, is applied to chaotic systems such as the Kuramoto-Sivashinsky equation, the two-dimensional Kolmogorov flow, and ERA5 reanalysis data for the atmosphere. It is observed that MP-NODE provide viable performance for such chaotic systems, not only for short-term trajectory predictions but also for invariant statistics that are hallmarks of the chaotic nature of these dynamics.
comment: 25 pages, 17 Figures, submitted to Computer Methods in Applied Mechanics and Engineering
♻ ☆ Socialized Learning: A Survey of the Paradigm Shift for Edge Intelligence in Networked Systems
Amidst the robust impetus from artificial intelligence (AI) and big data, edge intelligence (EI) has emerged as a nascent computing paradigm, synthesizing AI with edge computing (EC) to become an exemplary solution for unleashing the full potential of AI services. Nonetheless, challenges in communication costs, resource allocation, privacy, and security continue to constrain its proficiency in supporting services with diverse requirements. In response to these issues, this paper introduces socialized learning (SL) as a promising solution, further propelling the advancement of EI. SL is a learning paradigm predicated on social principles and behaviors, aimed at amplifying the collaborative capacity and collective intelligence of agents within the EI system. SL not only enhances the system's adaptability but also optimizes communication, and networking processes, essential for distributed intelligence across diverse devices and platforms. Therefore, a combination of SL and EI may greatly facilitate the development of collaborative intelligence in the future network. This paper presents the findings of a literature review on the integration of EI and SL, summarizing the latest achievements in existing research on EI and SL. Subsequently, we delve comprehensively into the limitations of EI and how it could benefit from SL. Special emphasis is placed on the communication challenges and networking strategies and other aspects within these systems, underlining the role of optimized network solutions in improving system efficiency. Based on these discussions, we elaborate in detail on three integrated components: socialized architecture, socialized training, and socialized inference, analyzing their strengths and weaknesses. Finally, we identify some possible future applications of combining SL and EI, discuss open problems and suggest some future research.
comment: This paper has been accepted by IEEE Communications Surveys and Tutorials
♻ ☆ Prompt a Robot to Walk with Large Language Models
Large language models (LLMs) pre-trained on vast internet-scale data have showcased remarkable capabilities across diverse domains. Recently, there has been escalating interest in deploying LLMs for robotics, aiming to harness the power of foundation models in real-world settings. However, this approach faces significant challenges, particularly in grounding these models in the physical world and in generating dynamic robot motions. To address these issues, we introduce a novel paradigm in which we use few-shot prompts collected from the physical environment, enabling the LLM to autoregressively generate low-level control commands for robots without task-specific fine-tuning. Experiments across various robots and environments validate that our method can effectively prompt a robot to walk. We thus illustrate how LLMs can proficiently function as low-level feedback controllers for dynamic motion control even in high-dimensional robotic systems. The project website and source code can be found at: https://prompt2walk.github.io/ .
comment: Conference on Decision and Control (CDC), 2024
♻ ☆ Self-Data Distillation for Recovering Quality in Pruned Large Language Models NeurIPS 2024
Large language models have driven significant progress in natural language processing, but their deployment requires substantial compute and memory resources. As models scale, compression techniques become essential for balancing model quality with computational efficiency. Structured pruning, which removes less critical components of the model, is a promising strategy for reducing complexity. However, one-shot pruning often results in significant quality degradation, particularly in tasks requiring multi-step reasoning. To recover lost quality, supervised fine-tuning (SFT) is commonly applied, but it can lead to catastrophic forgetting by shifting the model's learned data distribution. Therefore, addressing the degradation from both pruning and SFT is essential to preserve the original model's quality. In this work, we propose self-data distilled fine-tuning to address these challenges. Our approach leverages the original, unpruned model to generate a distilled dataset that preserves semantic richness and mitigates catastrophic forgetting by maintaining alignment with the base model's knowledge. Empirically, we demonstrate that self-data distillation consistently outperforms standard SFT, improving average accuracy by up to 8% on the HuggingFace OpenLLM Leaderboard v1. Specifically, when pruning 6 decoder blocks on Llama3.1-8B Instruct (i.e., 32 to 26 layers, reducing the model size from 8.03B to 6.72B parameters), our method retains 91.2% of the original model's accuracy compared to 81.7% with SFT, while reducing real-world FLOPs by 16.30%. Furthermore, our approach scales effectively across datasets, with the quality improving as the dataset size increases.
comment: Accepted at the NeurIPS 2024 Machine Learning and Compression Workshop
♻ ☆ Augmentation-aware Self-supervised Learning with Conditioned Projector NeurIPS 2023
Self-supervised learning (SSL) is a powerful technique for learning from unlabeled data. By learning to remain invariant to applied data augmentations, methods such as SimCLR and MoCo can reach quality on par with supervised approaches. However, this invariance may be detrimental for solving downstream tasks that depend on traits affected by augmentations used during pretraining, such as color. In this paper, we propose to foster sensitivity to such characteristics in the representation space by modifying the projector network, a common component of self-supervised architectures. Specifically, we supplement the projector with information about augmentations applied to images. For the projector to take advantage of this auxiliary conditioning when solving the SSL task, the feature extractor learns to preserve the augmentation information in its representations. Our approach, coined Conditional Augmentation-aware Self-supervised Learning (CASSLE), is directly applicable to typical joint-embedding SSL methods regardless of their objective functions. Moreover, it does not require major changes in the network architecture or prior knowledge of downstream tasks. In addition to an analysis of sensitivity towards different data augmentations, we conduct a series of experiments, which show that CASSLE improves over various SSL methods, reaching state-of-the-art performance in multiple downstream tasks.
comment: A short version of this paper appeared at the NeurIPS 2023 Workshop: Self-Supervised Learning - Theory and Practice (https://sslneurips23.github.io). The full paper was published (OA) in Knowledge-Based Systems (https://www.sciencedirect.com/science/article/pii/S0950705124012061)
♻ ☆ The Fragility of Fairness: Causal Sensitivity Analysis for Fair Machine Learning
Fairness metrics are a core tool in the fair machine learning literature (FairML), used to determine that ML models are, in some sense, "fair". Real-world data, however, are typically plagued by various measurement biases and other violated assumptions, which can render fairness assessments meaningless. We adapt tools from causal sensitivity analysis to the FairML context, providing a general framework which (1) accommodates effectively any combination of fairness metric and bias that can be posed in the "oblivious setting"; (2) allows researchers to investigate combinations of biases, resulting in non-linear sensitivity; and (3) enables flexible encoding of domain-specific constraints and assumptions. Employing this framework, we analyze the sensitivity of the most common parity metrics under 3 varieties of classifier across 14 canonical fairness datasets. Our analysis reveals the striking fragility of fairness assessments to even minor dataset biases. We show that causal sensitivity analysis provides a powerful and necessary toolkit for gauging the informativeness of parity metric evaluations. Our repository is available here: https://github.com/Jakefawkes/fragile_fair.
comment: Published at Neurips 2024 in the Dataset and Benchmarks Track
♻ ☆ LLM-Based Robust Product Classification in Commerce and Compliance EMNLP 2024
Product classification is a crucial task in international trade, as compliance regulations are verified and taxes and duties are applied based on product categories. Manual classification of products is time-consuming and error-prone, and the sheer volume of products imported and exported renders the manual process infeasible. Consequently, e-commerce platforms and enterprises involved in international trade have turned to automatic product classification using machine learning. However, current approaches do not consider the real-world challenges associated with product classification, such as very abbreviated and incomplete product descriptions. In addition, recent advancements in generative Large Language Models (LLMs) and their reasoning capabilities are mainly untapped in product classification and e-commerce. In this research, we explore the real-life challenges of industrial classification and we propose data perturbations that allow for realistic data simulation. Furthermore, we employ LLM-based product classification to improve the robustness of the prediction in presence of incomplete data. Our research shows that LLMs with in-context learning outperform the supervised approaches in the clean-data scenario. Additionally, we illustrate that LLMs are significantly more robust than the supervised approaches when data attacks are present.
comment: Camera-ready version for Customizable NLP Workshop at EMNLP 2024. 11 pages
♻ ☆ NeuroSEM: A hybrid framework for simulating multiphysics problems by coupling PINNs and spectral elements
Multiphysics problems that are characterized by complex interactions among fluid dynamics, heat transfer, structural mechanics, and electromagnetics, are inherently challenging due to their coupled nature. While experimental data on certain state variables may be available, integrating these data with numerical solvers remains a significant challenge. Physics-informed neural networks (PINNs) have shown promising results in various engineering disciplines, particularly in handling noisy data and solving inverse problems in partial differential equations (PDEs). However, their effectiveness in forecasting nonlinear phenomena in multiphysics regimes, particularly involving turbulence, is yet to be fully established. This study introduces NeuroSEM, a hybrid framework integrating PINNs with the high-fidelity Spectral Element Method (SEM) solver, Nektar++. NeuroSEM leverages the strengths of both PINNs and SEM, providing robust solutions for multiphysics problems. PINNs are trained to assimilate data and model physical phenomena in specific subdomains, which are then integrated into the Nektar++ solver. We demonstrate the efficiency and accuracy of NeuroSEM for thermal convection in cavity flow and flow past a cylinder. We applied NeuroSEM to the Rayleigh-B\'enard convection system, including cases with missing thermal boundary conditions and noisy datasets, and to real particle image velocimetry (PIV) data to capture flow patterns characterized by horseshoe vortical structures. The framework's plug-and-play nature facilitates its extension to other multiphysics or multiscale problems. Furthermore, NeuroSEM is optimized for efficient execution on emerging integrated GPU-CPU architectures. This hybrid approach enhances the accuracy and efficiency of simulations, making it a powerful tool for tackling complex engineering challenges in various scientific domains.
♻ ☆ Dash: Accelerating Distributed Private Convolutional Neural Network Inference with Arithmetic Garbled Circuits
The adoption of machine learning solutions is rapidly increasing across all parts of society. As the models grow larger, both training and inference of machine learning models is increasingly outsourced, e.g. to cloud service providers. This means that potentially sensitive data is processed on untrusted platforms, which bears inherent data security and privacy risks. In this work, we investigate how to protect distributed machine learning systems, focusing on deep convolutional neural networks. The most common and best-performing mixed MPC approaches are based on HE, secret sharing, and garbled circuits. They commonly suffer from large performance overheads, big accuracy losses, and communication overheads that grow linearly in the depth of the neural network. To improve on these problems, we present Dash, a fast and distributed private convolutional neural network inference scheme secure against malicious attackers. Building on arithmetic garbling gadgets [BMR16] and fancy-garbling [BCM+19], Dash is based purely on arithmetic garbled circuits. We introduce LabelTensors that allow us to leverage the massive parallelity of modern GPUs. Combined with state-of-the-art garbling optimizations, Dash outperforms previous garbling approaches up to a factor of about 100. Furthermore, we introduce an efficient scaling operation over the residues of the Chinese remainder theorem representation to arithmetic garbled circuits, which allows us to garble larger networks and achieve much higher accuracy than previous approaches. Finally, Dash requires only a single communication round per inference step, regardless of the depth of the neural network, and a very small constant online communication volume.
♻ ☆ Machine Learning for K-adaptability in Two-stage Robust Optimization
Two-stage robust optimization problems constitute one of the hardest optimization problem classes. One of the solution approaches to this class of problems is K-adaptability. This approach simultaneously seeks the best partitioning of the uncertainty set of scenarios into K subsets, and optimizes decisions corresponding to each of these subsets. In general case, it is solved using the K-adaptability branch-and-bound algorithm, which requires exploration of exponentially-growing solution trees. To accelerate finding high-quality solutions in such trees, we propose a machine learning-based node selection strategy. In particular, we construct a feature engineering scheme based on general two-stage robust optimization insights that allows us to train our machine learning tool on a database of resolved B&B trees, and to apply it as-is to problems of different sizes and/or types. We experimentally show that using our learned node selection strategy outperforms a vanilla, random node selection strategy when tested on problems of the same type as the training problems, also in case the K-value or the problem size differs from the training ones.
♻ ☆ Mitigate Position Bias in Large Language Models via Scaling a Single Dimension
Large Language Models (LLMs) are increasingly applied in various real-world scenarios due to their excellent generalization capabilities and robust generative abilities. However, they exhibit position bias, also known as "lost in the middle", a phenomenon that is especially pronounced in long-context scenarios, which indicates the placement of the key information in different positions of a prompt can significantly affect accuracy. This paper first explores the micro-level manifestations of position bias, concluding that attention weights are a micro-level expression of position bias. It further identifies that, in addition to position embeddings, causal attention mask also contributes to position bias by creating position-specific hidden states. Based on these insights, we propose a method to mitigate position bias by scaling this positional hidden states. Experiments on the NaturalQuestions Multi-document QA, KV retrieval, LongBench and timeline reorder tasks, using various models including RoPE models, context windowextended models, and Alibi models, demonstrate the effectiveness and generalizability of our approach. Our method can improve performance by up to 15.2% by modifying just one dimension of hidden states. Our code is available at https://aka.ms/PositionalHidden.
♻ ☆ Learning Truncated Causal History Model for Video Restoration NeurIPS 2024
One key challenge to video restoration is to model the transition dynamics of video frames governed by motion. In this work, we propose TURTLE to learn the truncated causal history model for efficient and high-performing video restoration. Unlike traditional methods that process a range of contextual frames in parallel, TURTLE enhances efficiency by storing and summarizing a truncated history of the input frame latent representation into an evolving historical state. This is achieved through a sophisticated similarity-based retrieval mechanism that implicitly accounts for inter-frame motion and alignment. The causal design in TURTLE enables recurrence in inference through state-memorized historical features while allowing parallel training by sampling truncated video clips. We report new state-of-the-art results on a multitude of video restoration benchmark tasks, including video desnowing, nighttime video deraining, video raindrops and rain streak removal, video super-resolution, real-world and synthetic video deblurring, and blind video denoising while reducing the computational cost compared to existing best contextual methods on all these tasks.
comment: Accepted to NeurIPS 2024. 24 pages
♻ ☆ Phantom: General Trigger Attacks on Retrieval Augmented Language Generation
Retrieval Augmented Generation (RAG) expands the capabilities of modern large language models (LLMs), by anchoring, adapting, and personalizing their responses to the most relevant knowledge sources. It is particularly useful in chatbot applications, allowing developers to customize LLM output without expensive retraining. Despite their significant utility in various applications, RAG systems present new security risks. In this work, we propose new attack vectors that allow an adversary to inject a single malicious document into a RAG system's knowledge base, and mount a backdoor poisoning attack. We design Phantom, a general two-stage optimization framework against RAG systems, that crafts a malicious poisoned document leading to an integrity violation in the model's output. First, the document is constructed to be retrieved only when a specific trigger sequence of tokens appears in the victim's queries. Second, the document is further optimized with crafted adversarial text that induces various adversarial objectives on the LLM output, including refusal to answer, reputation damage, privacy violations, and harmful behaviors. We demonstrate our attacks on multiple LLM architectures, including Gemma, Vicuna, and Llama, and show that they transfer to GPT-3.5 Turbo and GPT-4. Finally, we successfully conducted a Phantom attack on NVIDIA's black-box production RAG system, "Chat with RTX".
♻ ☆ Federated Continual Learning Goes Online: Uncertainty-Aware Memory Management for Vision Tasks and Beyond
Given the ability to model more realistic and dynamic problems, Federated Continual Learning (FCL) has been increasingly investigated recently. A well-known problem encountered in this setting is the so-called catastrophic forgetting, for which the learning model is inclined to focus on more recent tasks while forgetting the previously learned knowledge. The majority of the current approaches in FCL propose generative-based solutions to solve said problem. However, this setting requires multiple training epochs over the data, implying an offline setting where datasets are stored locally and remain unchanged over time. Furthermore, the proposed solutions are tailored for vision tasks solely. To overcome these limitations, we propose a new approach to deal with different modalities in the online scenario where new data arrive in streams of mini-batches that can only be processed once. To solve catastrophic forgetting, we propose an uncertainty-aware memory-based approach. Specifically, we suggest using an estimator based on the Bregman Information (BI) to compute the model's variance at the sample level. Through measures of predictive uncertainty, we retrieve samples with specific characteristics, and - by retraining the model on such samples - we demonstrate the potential of this approach to reduce the forgetting effect in realistic settings while maintaining data confidentiality and competitive communication efficiency compared to state-of-the-art approaches.
♻ ☆ Data Interpreter: An LLM Agent For Data Science
Large Language Model (LLM)-based agents have shown effectiveness across many applications. However, their use in data science scenarios requiring solving long-term interconnected tasks, dynamic data adjustments and domain expertise remains challenging. Previous approaches primarily focus on individual tasks, making it difficult to assess the complete data science workflow. Moreover, they struggle to handle real-time changes in intermediate data and fail to adapt dynamically to evolving task dependencies inherent to data science problems. In this paper, we present Data Interpreter, an LLM-based agent designed to automatically solve various data science problems end-to-end. Our Data Interpreter incorporates two key modules: 1) Hierarchical Graph Modeling, which breaks down complex problems into manageable subproblems, enabling dynamic node generation and graph optimization; and 2) Programmable Node Generation, a technique that refines and verifies each subproblem to iteratively improve code generation results and robustness. Extensive experiments consistently demonstrate the superiority of Data Interpreter. On InfiAgent-DABench, it achieves a 25% performance boost, raising accuracy from 75.9% to 94.9%. For machine learning and open-ended tasks, it improves performance from 88% to 95%, and from 60% to 97%, respectively. Moreover, on the MATH dataset, Data Interpreter achieves remarkable performance with a 26% improvement compared to state-of-the-art baselines. The code is available at https://github.com/geekan/MetaGPT.
♻ ☆ An Experimental Design for Anytime-Valid Causal Inference on Multi-Armed Bandits
Experimentation is crucial for managers to rigorously quantify the value of a change and determine if it leads to a statistically significant improvement over the status quo. As companies increasingly mandate that all changes undergo experimentation before widespread release, two challenges arise: (1) minimizing the proportion of customers assigned to the inferior treatment and (2) increasing experimentation velocity by enabling data-dependent stopping. This paper addresses both challenges by introducing the Mixture Adaptive Design (MAD), a new experimental design for multi-armed bandit (MAB) algorithms that enables anytime-valid inference on the Average Treatment Effect (ATE) for \emph{any} MAB algorithm. Intuitively, MAD "mixes" any bandit algorithm with a Bernoulli design, where at each time step, the probability of assigning a unit via the Bernoulli design is determined by a user-specified deterministic sequence that can converge to zero. This sequence lets managers directly control the trade-off between regret minimization and inferential precision. Under mild conditions on the rate the sequence converges to zero, we provide a confidence sequence that is asymptotically anytime-valid and guaranteed to shrink around the true ATE. Hence, when the true ATE converges to a non-zero value, the MAD confidence sequence is guaranteed to exclude zero in finite time. Therefore, the MAD enables managers to stop experiments early while ensuring valid inference, enhancing both the efficiency and reliability of adaptive experiments. Empirically, we demonstrate that the MAD achieves finite-sample anytime-validity while accurately and precisely estimating the ATE, all without incurring significant losses in reward compared to standard bandit designs.
♻ ☆ AuToMATo: An Out-Of-The-Box Persistence-Based Clustering Algorithm
We present AuToMATo, a novel clustering algorithm based on persistent homology. While AuToMATo is not parameter-free per se, we provide default choices for its parameters that make it into an out-of-the-box clustering algorithm that performs well across the board. AuToMATo combines the existing ToMATo clustering algorithm with a bootstrapping procedure in order to separate significant peaks of an estimated density function from non-significant ones. We perform a thorough comparison of AuToMATo (with its parameters fixed to their defaults) against many other state-of-the-art clustering algorithms. We find not only that AuToMATo compares favorably against parameter-free clustering algorithms, but in many instances also significantly outperforms even the best selection of parameters for other algorithms. AuToMATo is motivated by applications in topological data analysis, in particular the Mapper algorithm, where it is desirable to work with a clustering algorithm that does not need tuning of its parameters. Indeed, we provide evidence that AuToMATo performs well when used with Mapper. Finally, we provide an open-source implementation of AuToMATo in Python that is fully compatible with the standard scikit-learn architecture.
♻ ☆ The $μ\mathcal{G}$ Language for Programming Graph Neural Networks
Graph neural networks form a class of deep learning architectures specifically designed to work with graph-structured data. As such, they share the inherent limitations and problems of deep learning, especially regarding the issues of explainability and trustworthiness. We propose $\mu\mathcal{G}$, an original domain-specific language for the specification of graph neural networks that aims to overcome these issues. The language's syntax is introduced, and its meaning is rigorously defined by a denotational semantics. An equivalent characterization in the form of an operational semantics is also provided and, together with a type system, is used to prove the type soundness of $\mu\mathcal{G}$. We show how $\mu\mathcal{G}$ programs can be represented in a more user-friendly graphical visualization, and provide examples of its generality by showing how it can be used to define some of the most popular graph neural network models, or to develop any custom graph processing application.
♻ ☆ Shelf-Supervised Cross-Modal Pre-Training for 3D Object Detection
State-of-the-art 3D object detectors are often trained on massive labeled datasets. However, annotating 3D bounding boxes remains prohibitively expensive and time-consuming, particularly for LiDAR. Instead, recent works demonstrate that self-supervised pre-training with unlabeled data can improve detection accuracy with limited labels. Contemporary methods adapt best-practices for self-supervised learning from the image domain to point clouds (such as contrastive learning). However, publicly available 3D datasets are considerably smaller and less diverse than those used for image-based self-supervised learning, limiting their effectiveness. We do note, however, that such 3D data is naturally collected in a multimodal fashion, often paired with images. Rather than pre-training with only self-supervised objectives, we argue that it is better to bootstrap point cloud representations using image-based foundation models trained on internet-scale data. Specifically, we propose a shelf-supervised approach (e.g. supervised with off-the-shelf image foundation models) for generating zero-shot 3D bounding boxes from paired RGB and LiDAR data. Pre-training 3D detectors with such pseudo-labels yields significantly better semi-supervised detection accuracy than prior self-supervised pretext tasks. Importantly, we show that image-based shelf-supervision is helpful for training LiDAR-only, RGB-only and multi-modal (RGB + LiDAR) detectors. We demonstrate the effectiveness of our approach on nuScenes and WOD, significantly improving over prior work in limited data settings. Our code is available at https://github.com/meharkhurana03/cm3d
comment: The first two authors contributed equally. This work has been accepted to the Conference on Robot Learning (CoRL) 2024
♻ ☆ Equivariant Diffusion Policy
Recent work has shown diffusion models are an effective approach to learning the multimodal distributions arising from demonstration data in behavior cloning. However, a drawback of this approach is the need to learn a denoising function, which is significantly more complex than learning an explicit policy. In this work, we propose Equivariant Diffusion Policy, a novel diffusion policy learning method that leverages domain symmetries to obtain better sample efficiency and generalization in the denoising function. We theoretically analyze the $\mathrm{SO}(2)$ symmetry of full 6-DoF control and characterize when a diffusion model is $\mathrm{SO}(2)$-equivariant. We furthermore evaluate the method empirically on a set of 12 simulation tasks in MimicGen, and show that it obtains a success rate that is, on average, 21.9% higher than the baseline Diffusion Policy. We also evaluate the method on a real-world system to show that effective policies can be learned with relatively few training samples, whereas the baseline Diffusion Policy cannot.
comment: Conference on Robot Learning 2024, Oral Presentation
♻ ☆ GSE: Group-wise Sparse and Explainable Adversarial Attacks
Sparse adversarial attacks fool deep neural networks (DNNs) through minimal pixel perturbations, often regularized by the $\ell_0$ norm. Recent efforts have replaced this norm with a structural sparsity regularizer, such as the nuclear group norm, to craft group-wise sparse adversarial attacks. The resulting perturbations are thus explainable and hold significant practical relevance, shedding light on an even greater vulnerability of DNNs. However, crafting such attacks poses an optimization challenge, as it involves computing norms for groups of pixels within a non-convex objective. We address this by presenting a two-phase algorithm that generates group-wise sparse attacks within semantically meaningful areas of an image. Initially, we optimize a quasinorm adversarial loss using the $1/2-$quasinorm proximal operator tailored for non-convex programming. Subsequently, the algorithm transitions to a projected Nesterov's accelerated gradient descent with $2-$norm regularization applied to perturbation magnitudes. Rigorous evaluations on CIFAR-10 and ImageNet datasets demonstrate a remarkable increase in group-wise sparsity, e.g., $50.9\%$ on CIFAR-10 and $38.4\%$ on ImageNet (average case, targeted attack). This performance improvement is accompanied by significantly faster computation times, improved explainability, and a $100\%$ attack success rate.
♻ ☆ Bridging The Gap between Low-rank and Orthogonal Adaptation via Householder Reflection Adaptation
While following different technical routes, both low-rank and orthogonal adaptation techniques can efficiently adapt large-scale pre-training models in specific tasks or domains based on a small piece of trainable parameters. In this study, we bridge the gap between these two techniques, proposing a simple but effective adaptation method based on Householder reflections. Given a pre-trained model, our method fine-tunes its layers by multiplying each frozen weight matrix with an orthogonal matrix constructed by a chain of learnable Householder reflections (HRs). This HR-based orthogonal fine-tuning is equivalent to an adaptive low-rank adaptation. Moreover, we show that the orthogonality of the reflection planes corresponding to the HRs impacts the model capacity and regularity. The analysis motivates us to regularize the orthogonality of the HRs, leading to different implementations of the proposed Householder reflection adaptation (HRA) method. Compared with state-of-the-art methods, HRA achieves superior performance with fewer learnable parameters when adapting large language models and conditional image generators. The code of the experiments is available at \url{https://github.com/DaShenZi721/HRA}, and the method has been merged into the \href{https://github.com/huggingface/peft}{PEFT} package.
♻ ☆ Inducing anxiety in large language models can induce bias
Large language models (LLMs) are transforming research on machine learning while galvanizing public debates. Understanding not only when these models work well and succeed but also why they fail and misbehave is of great societal relevance. We propose to turn the lens of psychiatry, a framework used to describe and modify maladaptive behavior, to the outputs produced by these models. We focus on twelve established LLMs and subject them to a questionnaire commonly used in psychiatry. Our results show that six of the latest LLMs respond robustly to the anxiety questionnaire, producing comparable anxiety scores to humans. Moreover, the LLMs' responses can be predictably changed by using anxiety-inducing prompts. Anxiety-induction not only influences LLMs' scores on an anxiety questionnaire but also influences their behavior in a previously-established benchmark measuring biases such as racism and ageism. Importantly, greater anxiety-inducing text leads to stronger increases in biases, suggesting that how anxiously a prompt is communicated to large language models has a strong influence on their behavior in applied settings. These results demonstrate the usefulness of methods taken from psychiatry for studying the capable algorithms to which we increasingly delegate authority and autonomy.
♻ ☆ Optimal Time Complexity Algorithms for Computing General Random Walk Graph Kernels on Sparse Graphs
We present the first linear time complexity randomized algorithms for unbiased approximation of the celebrated family of general random walk kernels (RWKs) for sparse graphs. This includes both labelled and unlabelled instances. The previous fastest methods for general RWKs were of cubic time complexity and not applicable to labelled graphs. Our method samples dependent random walks to compute novel graph embeddings in $\mathbb{R}^d$ whose dot product is equal to the true RWK in expectation. It does so without instantiating the direct product graph in memory, meaning we can scale to massive datasets that cannot be stored on a single machine. We derive exponential concentration bounds to prove that our estimator is sharp, and show that the ability to approximate general RWKs (rather than just special cases) unlocks efficient implicit graph kernel learning. Our method is up to $\mathbf{27\times}$ faster than its counterparts for efficient computation on large graphs and scales to graphs $\mathbf{128 \times}$ bigger than largest examples amenable to brute-force computation.
♻ ☆ Linear Transformer Topological Masking with Graph Random Features
When training transformers on graph-structured data, incorporating information about the underlying topology is crucial for good performance. Topological masking, a type of relative position encoding, achieves this by upweighting or downweighting attention depending on the relationship between the query and keys in a graph. In this paper, we propose to parameterise topological masks as a learnable function of a weighted adjacency matrix -- a novel, flexible approach which incorporates a strong structural inductive bias. By approximating this mask with graph random features (for which we prove the first known concentration bounds), we show how this can be made fully compatible with linear attention, preserving $\mathcal{O}(N)$ time and space complexity with respect to the number of input tokens. The fastest previous alternative was $\mathcal{O}(N \log N)$ and only suitable for specific graphs. Our efficient masking algorithms provide strong performance gains for tasks on image and point cloud data, including with $>30$k nodes.
♻ ☆ Discovering Knowledge-Critical Subnetworks in Pretrained Language Models EMNLP 2024
Pretrained language models (LMs) encode implicit representations of knowledge in their parameters. However, localizing these representations and disentangling them from each other remains an open problem. In this work, we investigate whether pretrained language models contain various knowledge-critical subnetworks: particular sparse computational subgraphs that can, if removed, precisely suppress specific knowledge the model has memorized. We propose a multi-objective differentiable masking scheme that can be applied to both weights and neurons to discover such subnetworks and show that we can use them to precisely remove specific knowledge from models while minimizing adverse effects on the behavior of the original model. We demonstrate our method on multiple GPT2 variants, uncovering highly sparse subnetworks (98%+ sparsity) that are critical for expressing specific collections of relational knowledge. When these subnetworks are removed, the remaining network maintains most of its initial abilities but struggles to represent the suppressed knowledge.
comment: EMNLP 2024
♻ ☆ Sufficient and Necessary Explanations (and What Lies in Between)
As complex machine learning models continue to find applications in high-stakes decision-making scenarios, it is crucial that we can explain and understand their predictions. Post-hoc explanation methods provide useful insights by identifying important features in an input $\mathbf{x}$ with respect to the model output $f(\mathbf{x})$. In this work, we formalize and study two precise notions of feature importance for general machine learning models: sufficiency and necessity. We demonstrate how these two types of explanations, albeit intuitive and simple, can fall short in providing a complete picture of which features a model finds important. To this end, we propose a unified notion of importance that circumvents these limitations by exploring a continuum along a necessity-sufficiency axis. Our unified notion, we show, has strong ties to other popular definitions of feature importance, like those based on conditional independence and game-theoretic quantities like Shapley values. Crucially, we demonstrate how a unified perspective allows us to detect important features that could be missed by either of the previous approaches alone.
♻ ☆ Agent Planning with World Knowledge Model NeurIPS 2024
Recent endeavors towards directly using large language models (LLMs) as agent models to execute interactive planning tasks have shown commendable results. Despite their achievements, however, they still struggle with brainless trial-and-error in global planning and generating hallucinatory actions in local planning due to their poor understanding of the ``real'' physical world. Imitating humans' mental world knowledge model which provides global prior knowledge before the task and maintains local dynamic knowledge during the task, in this paper, we introduce parametric World Knowledge Model (WKM) to facilitate agent planning. Concretely, we steer the agent model to self-synthesize knowledge from both expert and sampled trajectories. Then we develop WKM, providing prior task knowledge to guide the global planning and dynamic state knowledge to assist the local planning. Experimental results on three complex real-world simulated datasets with three state-of-the-art open-source LLMs, Mistral-7B, Gemma-7B, and Llama-3-8B, demonstrate that our method can achieve superior performance compared to various strong baselines. Besides, we analyze to illustrate that our WKM can effectively alleviate the blind trial-and-error and hallucinatory action issues, providing strong support for the agent's understanding of the world. Other interesting findings include: 1) our instance-level task knowledge can generalize better to unseen tasks, 2) weak WKM can guide strong agent model planning, and 3) unified WKM training has promising potential for further development. The code is available at https://github.com/zjunlp/WKM.
comment: NeurIPS 2024
♻ ☆ Statistical Taylor Expansion
Statistical Taylor expansion replaces the input precise variables in a conventional Taylor expansion with random variables each with known distribution, to calculate the result mean and deviation. It is based on the uncorrelated uncertainty assumption: Each input variable is measured independently with fine enough statistical precision, so that their uncertainties are independent of each other. Statistical Taylor expansion reviews that the intermediate analytic expressions can no longer be regarded as independent of each other, and the result of analytic expression should be path independent. This conclusion differs fundamentally from the conventional common approach in applied mathematics to find the best execution path for a result. This paper also presents an implementation of statistical Taylor expansion called variance arithmetic, and the tests on variance arithmetic.
comment: 65 pages, 53 figures
♻ ☆ Uncovering mesa-optimization algorithms in Transformers
Some autoregressive models exhibit in-context learning capabilities: being able to learn as an input sequence is processed, without undergoing any parameter changes, and without being explicitly trained to do so. The origins of this phenomenon are still poorly understood. Here we analyze a series of Transformer models trained to perform synthetic sequence prediction tasks, and discover that standard next-token prediction error minimization gives rise to a subsidiary learning algorithm that adjusts the model as new inputs are revealed. We show that this process corresponds to gradient-based optimization of a principled objective function, which leads to strong generalization performance on unseen sequences. Our findings explain in-context learning as a product of autoregressive loss minimization and inform the design of new optimization-based Transformer layers.
♻ ☆ Exploring and Addressing Reward Confusion in Offline Preference Learning NeurIPS2024
Spurious correlations in a reward model's training data can prevent Reinforcement Learning from Human Feedback (RLHF) from identifying the desired goal and induce unwanted behaviors. This paper shows that offline RLHF is susceptible to reward confusion, especially in the presence of spurious correlations in offline data. We create a benchmark to study this problem and propose a method that can significantly reduce reward confusion by leveraging transitivity of preferences while building a global preference chain with active learning.
comment: NeurIPS2024 Workshop on Bayesian Decision-making and Uncertainty
♻ ☆ Generalized Simplicial Attention Neural Networks
Graph machine learning methods excel at leveraging pairwise relations present in the data. However, graphs are unable to fully capture the multi-way interactions inherent in many complex systems. An effective way to incorporate them is to model the data on higher-order combinatorial topological spaces, such as Simplicial Complexes (SCs) or Cell Complexes. For this reason, we introduce Generalized Simplicial Attention Neural Networks (GSANs), novel neural network architectures designed to process data living on simplicial complexes using masked self-attentional layers. Hinging on topological signal processing principles, we devise a series of principled self-attention mechanisms able to process data associated with simplices of various order, such as nodes, edges, triangles, and beyond. These schemes learn how to combine data associated with neighbor simplices of consecutive order in a task-oriented fashion, leveraging on the simplicial Dirac operator and its Dirac decomposition. We also prove that GSAN satisfies two fundamental properties: permutation equivariance and simplicial-awareness. Finally, we illustrate how our approach compares favorably with other simplicial and graph models when applied to several (inductive and transductive) tasks such as trajectory prediction, missing data imputation, graph classification, and simplex prediction.
comment: arXiv admin note: text overlap with arXiv:2203.07485 (preliminary version of this paper). 16 pages, accepted for publication at IEEE TSIPN
♻ ☆ A Probabilistic Model Behind Self-Supervised Learning
In self-supervised learning (SSL), representations are learned via an auxiliary task without annotated labels. A common task is to classify augmentations or different modalities of the data, which share semantic content (e.g. an object in an image) but differ in style (e.g. the object's location). Many approaches to self-supervised learning have been proposed, e.g. SimCLR, CLIP, and DINO, which have recently gained much attention for their representations achieving downstream performance comparable to supervised learning. However, a theoretical understanding of self-supervised methods eludes. Addressing this, we present a generative latent variable model for self-supervised learning and show that several families of discriminative SSL, including contrastive methods, induce a comparable distribution over representations, providing a unifying theoretical framework for these methods. The proposed model also justifies connections drawn to mutual information and the use of a ''projection head''. Learning representations by fitting the model generatively (termed SimVAE) improves performance over discriminative and other VAE-based methods on simple image benchmarks and significantly narrows the gap between generative and discriminative representation learning in more complex settings. Importantly, as our analysis predicts, SimVAE outperforms self-supervised learning where style information is required, taking an important step toward understanding self-supervised methods and achieving task-agnostic representations.
♻ ☆ Vanilla Gradient Descent for Oblique Decision Trees ECAI
Decision Trees (DTs) constitute one of the major highly non-linear AI models, valued, e.g., for their efficiency on tabular data. Learning accurate DTs is, however, complicated, especially for oblique DTs, and does take a significant training time. Further, DTs suffer from overfitting, e.g., they proverbially "do not generalize" in regression tasks. Recently, some works proposed ways to make (oblique) DTs differentiable. This enables highly efficient gradient-descent algorithms to be used to learn DTs. It also enables generalizing capabilities by learning regressors at the leaves simultaneously with the decisions in the tree. Prior approaches to making DTs differentiable rely either on probabilistic approximations at the tree's internal nodes (soft DTs) or on approximations in gradient computation at the internal node (quantized gradient descent). In this work, we propose DTSemNet, a novel semantically equivalent and invertible encoding for (hard, oblique) DTs as Neural Networks (NNs), that uses standard vanilla gradient descent. Experiments across various classification and regression benchmarks show that oblique DTs learned using DTSemNet are more accurate than oblique DTs of similar size learned using state-of-the-art techniques. Further, DT training time is significantly reduced. We also experimentally demonstrate that DTSemNet can learn DT policies as efficiently as NN policies in the Reinforcement Learning (RL) setup with physical inputs (dimensions $\leq32$). The code is available at https://github.com/CPS-research-group/dtsemnet.
comment: Published in European Conference on Artificial Intelligence (ECAI), 2024. Full version (includes supplementary material)
♻ ☆ DiaSynth: Synthetic Dialogue Generation Framework for Low Resource Dialogue Applications
The scarcity of domain-specific dialogue datasets limits the development of dialogue systems across applications. Existing research is constrained by general or niche datasets that lack sufficient scale for training dialogue systems. To address this gap, we introduce DiaSynth - a synthetic dialogue generation framework capable of generating high-quality, contextually rich dialogues across a wide range of domains. Unlike existing frameworks, DiaSynth uses Large Language Models (LLMs) and Chain of Thought (CoT) reasoning to generate dynamic, domain-specific dialogues with simulated personas and diverse conversational features. We perform our experiments by generating synthetic data using different LLMs and few-shot examples from DialogSum and SAMSum. The pretrained language models fine-tuned on the synthetic data outperform the base models by 16.47% on dialogue summarization, while the comparison between models fine-tuned on in-domain data and synthetic data shows that the synthetic data is able to capture 90.48% of the performance distribution of the in-domain data on dialogue summarization. The quality of the data generated also increases as we increase the size of LLM from 3B to 8B. These results validate DiaSynth's potential as a robust alternative to traditional data collection methods. We open source the code and data generated for future research.
comment: 13 pages, 1 figure
♻ ☆ The Empirical Impact of Neural Parameter Symmetries, or Lack Thereof NeurIPS 2024
Many algorithms and observed phenomena in deep learning appear to be affected by parameter symmetries -- transformations of neural network parameters that do not change the underlying neural network function. These include linear mode connectivity, model merging, Bayesian neural network inference, metanetworks, and several other characteristics of optimization or loss-landscapes. However, theoretical analysis of the relationship between parameter space symmetries and these phenomena is difficult. In this work, we empirically investigate the impact of neural parameter symmetries by introducing new neural network architectures that have reduced parameter space symmetries. We develop two methods, with some provable guarantees, of modifying standard neural networks to reduce parameter space symmetries. With these new methods, we conduct a comprehensive experimental study consisting of multiple tasks aimed at assessing the effect of removing parameter symmetries. Our experiments reveal several interesting observations on the empirical impact of parameter symmetries; for instance, we observe linear mode connectivity between our networks without alignment of weight spaces, and we find that our networks allow for faster and more effective Bayesian neural network training. Our code is available at https://github.com/cptq/asymmetric-networks
comment: NeurIPS 2024. v2: added / updated some citations. v3 added link to code, and some additional ablations
♻ ☆ Enhancing learning in spiking neural networks through neuronal heterogeneity and neuromodulatory signaling
Recent progress in artificial intelligence (AI) has been driven by insights from neuroscience, particularly with the development of artificial neural networks (ANNs). This has significantly enhanced the replication of complex cognitive tasks such as vision and natural language processing. Despite these advances, ANNs struggle with continual learning, adaptable knowledge transfer, robustness, and resource efficiency - capabilities that biological systems handle seamlessly. Specifically, ANNs often overlook the functional and morphological diversity of the brain, hindering their computational capabilities. Furthermore, incorporating cell-type specific neuromodulatory effects into ANNs with neuronal heterogeneity could enable learning at two spatial scales: spiking behavior at the neuronal level, and synaptic plasticity at the circuit level, thereby potentially enhancing their learning abilities. In this article, we summarize recent bio-inspired models, learning rules and architectures and propose a biologically-informed framework for enhancing ANNs. Our proposed dual-framework approach highlights the potential of spiking neural networks (SNNs) for emulating diverse spiking behaviors and dendritic compartments to simulate morphological and functional diversity of neuronal computations. Finally, we outline how the proposed approach integrates brain-inspired compartmental models and task-driven SNNs, balances bioinspiration and complexity, and provides scalable solutions for pressing AI challenges, such as continual learning, adaptability, robustness, and resource-efficiency.
comment: 30 pages, 4 figures, 3 boxes
♻ ☆ Random Function Descent
Classical worst-case optimization theory neither explains the success of optimization in machine learning, nor does it help with step size selection. In this paper we demonstrate the viability and advantages of replacing the classical 'convex function' framework with a 'random function' framework. With complexity $\mathcal{O}(n^3d^3)$, where $n$ is the number of steps and $d$ the number of dimensions, Bayesian optimization with gradients has not been viable in large dimension so far. By bridging the gap between Bayesian optimization (i.e. random function optimization theory) and classical optimization we establish viability. Specifically, we use a 'stochastic Taylor approximation' to rediscover gradient descent, which is scalable in high dimension due to $\mathcal{O}(nd)$ complexity. This rediscovery yields a specific step size schedule we call Random Function Descent (RFD). The advantage of this random function framework is that RFD is scale invariant and that it provides a theoretical foundation for common step size heuristics such as gradient clipping and gradual learning rate warmup.
♻ ☆ Adaptive Hybrid Model Pruning in Federated Learning through Loss Exploration
The rapid proliferation of smart devices coupled with the advent of 6G networks has profoundly reshaped the domain of collaborative machine learning. Alongside growing privacy-security concerns in sensitive fields, these developments have positioned federated learning (FL) as a pivotal technology for decentralized model training. Despite its vast potential, specially in the age of complex foundation models, FL encounters challenges such as elevated communication costs, computational constraints, and the complexities of non-IID data distributions. We introduce AutoFLIP, an innovative approach that utilizes a federated loss exploration phase to drive adaptive hybrid pruning, operating in a structured and unstructured way. This innovative mechanism automatically identifies and prunes model substructure by distilling knowledge on model gradients behavior across different non-IID client losses topology, thereby optimizing computational efficiency and enhancing model performance on resource constrained scenarios. Extensive experiments on various datasets and FL tasks reveal that AutoFLIP not only efficiently accelerates global convergence, but also achieves superior accuracy and robustness compared to traditional methods. On average, AutoFLIP reduces computational overhead by 48.8% and communication costs by 35.5%, while improving global accuracy. By significantly reducing these overheads, AutoFLIP offer the way for efficient FL deployment in real-world applications for a scalable and broad applicability.
♻ ☆ Conjugate Bayesian Two-step Change Point Detection for Hawkes Process NeurIPS 2024
The Bayesian two-step change point detection method is popular for the Hawkes process due to its simplicity and intuitiveness. However, the non-conjugacy between the point process likelihood and the prior requires most existing Bayesian two-step change point detection methods to rely on non-conjugate inference methods. These methods lack analytical expressions, leading to low computational efficiency and impeding timely change point detection. To address this issue, this work employs data augmentation to propose a conjugate Bayesian two-step change point detection method for the Hawkes process, which proves to be more accurate and efficient. Extensive experiments on both synthetic and real data demonstrate the superior effectiveness and efficiency of our method compared to baseline methods. Additionally, we conduct ablation studies to explore the robustness of our method concerning various hyperparameters. Our code is publicly available at https://github.com/Aurora2050/CoBay-CPD.
comment: 10 pages, accepted by NeurIPS 2024
♻ ☆ Introducing MeMo: A Multimodal Dataset for Memory Modelling in Multiparty Conversations
Conversational memory is the process by which humans encode, retain and retrieve verbal, non-verbal and contextual information from a conversation. Since human memory is selective, differing recollections of the same events can lead to misunderstandings and misalignments within a group. Yet, conversational facilitation systems, aimed at advancing the quality of group interactions, usually focus on tracking users' states within an individual session, ignoring what remains in each participant's memory after the interaction. Understanding conversational memory can be used as a source of information on the long-term development of social connections within a group. This paper introduces the MeMo corpus, the first conversational dataset annotated with participants' memory retention reports, aimed at facilitating computational modelling of human conversational memory. The MeMo corpus includes 31 hours of small-group discussions on Covid-19, repeated 3 times over the term of 2 weeks. It integrates validated behavioural and perceptual measures, audio, video, and multimodal annotations, offering a valuable resource for studying and modelling conversational memory and group dynamics. By introducing the MeMo corpus, analysing its validity, and demonstrating its usefulness for future research, this paper aims to pave the way for future research in conversational memory modelling for intelligent system development.
♻ ☆ Explaining Learned Reward Functions with Counterfactual Trajectories
Learning rewards from human behaviour or feedback is a promising approach to aligning AI systems with human values but fails to consistently extract correct reward functions. Interpretability tools could enable users to understand and evaluate possible flaws in learned reward functions. We propose Counterfactual Trajectory Explanations (CTEs) to interpret reward functions in reinforcement learning by contrasting an original with a counterfactual partial trajectory and the rewards they each receive. We derive six quality criteria for CTEs and propose a novel Monte-Carlo-based algorithm for generating CTEs that optimises these quality criteria. Finally, we measure how informative the generated explanations are to a proxy-human model by training it on CTEs. CTEs are demonstrably informative for the proxy-human model, increasing the similarity between its predictions and the reward function on unseen trajectories. Further, it learns to accurately judge differences in rewards between trajectories and generalises to out-of-distribution examples. Although CTEs do not lead to a perfect understanding of the reward, our method, and more generally the adaptation of XAI methods, are presented as a fruitful approach for interpreting learned reward functions.
♻ ☆ CoRA: Collaborative Information Perception by Large Language Model's Weights for Recommendation
Involving collaborative information in Large Language Models (LLMs) is a promising technique for adapting LLMs for recommendation. Existing methods achieve this by concatenating collaborative features with text tokens into a unified sequence input and then fine-tuning to align these features with LLM's input space. Although effective, in this work, we identify two limitations when adapting LLMs to recommendation tasks, which hinder the integration of general knowledge and collaborative information, resulting in sub-optimal recommendation performance. (1) Fine-tuning LLM with recommendation data can undermine its inherent world knowledge and fundamental competencies, which are crucial for interpreting and inferring recommendation text. (2) Incorporating collaborative features into textual prompts disrupts the semantics of the original prompts, preventing LLM from generating appropriate outputs. In this paper, we propose a new paradigm, CoRA (an acronym for Collaborative LoRA), with a collaborative weights generator. Rather than input space alignment, this method aligns collaborative information with LLM's parameter space, representing them as incremental weights to update LLM's output. This way, LLM perceives collaborative information without altering its general knowledge and text inference capabilities. Specifically, we employ a collaborative filtering model to extract user and item embeddings, converting them into collaborative weights with low-rank properties through the collaborative weights generator. We then merge the collaborative weights into LLM's weights, enabling LLM to perceive the collaborative signals and generate personalized recommendations without fine-tuning or extra collaborative tokens in prompts. Extensive experiments confirm that CoRA effectively integrates collaborative information into LLM, enhancing recommendation performance.
♻ ☆ Information propagation dynamics in Deep Graph Networks
Graphs are a highly expressive abstraction for modeling entities and their relations, such as molecular structures, social networks, and traffic networks. Deep Graph Networks (DGNs) have emerged as a family of deep learning models that can effectively process and learn such structured information. However, learning effective information propagation patterns within DGNs remains a critical challenge that heavily influences the model capabilities, both in the static domain and in the temporal domain (where features and/or topology evolve). Given this challenge, this thesis investigates the dynamics of information propagation within DGNs for static and dynamic graphs, focusing on their design as dynamical systems. Throughout this work, we provide theoretical and empirical evidence to demonstrate the effectiveness of our proposed architectures in propagating and preserving long-term dependencies between nodes, and in learning complex spatio-temporal patterns from irregular and sparsely sampled dynamic graphs. In summary, this thesis provides a comprehensive exploration of the intersection between graphs, deep learning, and dynamical systems, offering insights and advancements for the field of graph representation learning and paving the way for more effective and versatile graph-based learning models.
comment: PhD thesis
♻ ☆ Unraveling the Mechanics of Learning-Based Demonstration Selection for In-Context Learning
Large Language Models (LLMs) have demonstrated impressive in-context learning (ICL) capabilities from few-shot demonstration exemplars. While recent learning-based demonstration selection methods have proven beneficial to ICL by choosing more useful exemplars, their underlying mechanisms are opaque, hindering efforts to address limitations such as high training costs and poor generalization across tasks. These methods generally assume the selection process captures similarities between the exemplar and the target instance, however, it remains unknown what kinds of similarities are captured and vital to performing ICL. To dive into this question, we analyze the working mechanisms of the learning-based demonstration selection methods and empirically identify two important factors related to similarity measurement: 1) The ability to integrate different levels of task-agnostic text similarities between the input of exemplars and test cases enhances generalization power across different tasks. 2) Incorporating task-specific labels when measuring the similarities significantly improves the performance on each specific task. We validate these two findings through extensive quantitative and qualitative analyses across ten datasets and various LLMs. Based on our findings, we introduce two effective yet simplified exemplar selection methods catering to task-agnostic and task-specific demands, eliminating the costly LLM inference overhead.
comment: 17 pages, 7 figures and 9 tables
♻ ☆ Mathador-LM: A Dynamic Benchmark for Mathematical Reasoning on Large Language Models EMNLP 2024
We introduce Mathador-LM, a new benchmark for evaluating the mathematical reasoning on large language models (LLMs), combining ruleset interpretation, planning, and problem-solving. This benchmark is inspired by the Mathador game, where the objective is to reach a target number using basic arithmetic operations on a given set of base numbers, following a simple set of rules. We show that, across leading LLMs, we obtain stable average performance while generating benchmark instances \emph{dynamically}, following a target difficulty level. Thus, our benchmark alleviates concerns about test-set leakage into training data, an issue that often undermines popular benchmarks. Additionally, we conduct a comprehensive evaluation of both open and closed-source state-of-the-art LLMs on Mathador-LM. Our findings reveal that contemporary models struggle with Mathador-LM, scoring significantly lower than average 3rd graders. This stands in stark contrast to their strong performance on popular mathematical reasoning benchmarks. The implementation of Mathador-LM benchmark is available at \href{https://github.com/IST-DASLab/Mathador-LM}{github.com/IST-DASLab/Mathador-LM}.
comment: EMNLP 2024
♻ ☆ Balanced Neural ODEs: nonlinear model order reduction and Koopman operator approximations
Variational Autoencoders (VAEs) are a powerful framework for learning compact latent representations, while NeuralODEs excel in learning transient system dynamics. This work combines the strengths of both to create fast surrogate models with adjustable complexity. By leveraging the VAE's dimensionality reduction using a non-hierarchical prior, our method adaptively assigns stochastic noise, naturally complementing known NeuralODE training enhancements and enabling probabilistic time series modeling. We show that standard Latent ODEs struggle with dimensionality reduction in systems with time-varying inputs. Our approach mitigates this by continuously propagating variational parameters through time, establishing fixed information channels in latent space. This results in a flexible and robust method that can learn different system complexities, e.g. deep neural networks or linear matrices. Hereby, it enables efficient approximation of the Koopman operator without the need for predefining its dimensionality. As our method balances dimensionality reduction and reconstruction accuracy, we call it Balanced Neural ODE (B-NODE). We demonstrate the effectiveness of this method on academic test cases and apply it to a real-world example of a thermal power plant.
comment: Conference paper under review
♻ ☆ Few for Many: Tchebycheff Set Scalarization for Many-Objective Optimization
Multi-objective optimization can be found in many real-world applications where some conflicting objectives can not be optimized by a single solution. Existing optimization methods often focus on finding a set of Pareto solutions with different optimal trade-offs among the objectives. However, the required number of solutions to well approximate the whole Pareto optimal set could be exponentially large with respect to the number of objectives, which makes these methods unsuitable for handling many optimization objectives. In this work, instead of finding a dense set of Pareto solutions, we propose a novel Tchebycheff set scalarization method to find a few representative solutions (e.g., 5) to cover a large number of objectives (e.g., $>100$) in a collaborative and complementary manner. In this way, each objective can be well addressed by at least one solution in the small solution set. In addition, we further develop a smooth Tchebycheff set scalarization approach for efficient optimization with good theoretical guarantees. Experimental studies on different problems with many optimization objectives demonstrate the effectiveness of our proposed method.
♻ ☆ TADA: Temporal Adversarial Data Augmentation for Time Series Data
Domain generalization aim to train models to effectively perform on samples that are unseen and outside of the distribution. Adversarial data augmentation (ADA) is a widely used technique in domain generalization. It enhances the model robustness by including synthetic samples designed to simulate potential unseen scenarios into the training datasets, which is then used to train the model. However, in time series data, traditional ADA approaches often fail to address distribution shifts related to temporal characteristics. To address this limitation, we propose Temporal Adversarial Data Augmentation (TADA) for time series data, which incorporate time warping into ADA. Although time warping is inherently non-differentiable, ADA relies on generating samples through backpropagation. We resolve this issue by leveraging the duality between phase shifts in the frequency domain and time shifts in the time domain, thereby making the process differentiable. Our evaluations across various time series datasets demonstrate that TADA outperforms existing methods for domain generalization. In addition, using distribution visualization, we confirmed that the distribution shifts induced by TADA are clearly different from those induced by ADA, and together, they effectively simulate real-world distribution shifts.
♻ ☆ The Reachability Problem for Neural-Network Control Systems
A control system consists of a plant component and a controller which periodically computes a control input for the plant. We consider systems where the controller is implemented by a feedforward neural network with ReLU activations. The reachability problem asks, given a set of initial states, whether a set of target states can be reached. We show that this problem is undecidable even for trivial plants and fixed-depth neural networks with three inputs and outputs. We also show that the problem becomes semi-decidable when the plant as well as the input and target sets are given by automata over infinite words.
♻ ☆ Foundation Models for ECG: Leveraging Hybrid Self-Supervised Learning for Advanced Cardiac Diagnostics
Using foundation models enhanced by self-supervised learning (SSL) methods presents an innovative approach to electrocardiogram (ECG) analysis, which is crucial for cardiac health monitoring and diagnosis. This study comprehensively evaluates foundation models for ECGs, leveraging SSL methods, including generative and contrastive learning, on a vast dataset comprising approximately 1.3 million ECG samples. By integrating these methods with consideration of the unique characteristics of ECGs, we developed a Hybrid Learning (HL) for foundation models that improve the precision and reliability of cardiac diagnostics. The HL-based foundation model adeptly captures the intricate details of ECGs, enhancing diagnostic capability. The results underscore the considerable potential of SSL-enhanced foundation models in clinical settings, setting the stage for future research into their scalable applications across a broader range of medical diagnostics. This work sets a new standard in the ECG field, emphasizing the transformative influence of tailored, data-driven model training on the effectiveness and accuracy of medical diagnostics.
comment: 27 pages
♻ ☆ Dissecting Fine-Tuning Unlearning in Large Language Models EMNLP 2024
Fine-tuning-based unlearning methods prevail for preventing targeted harmful, sensitive, or copyrighted information within large language models while preserving overall capabilities. However, the true effectiveness of these methods is unclear. In this work, we delve into the limitations of fine-tuning-based unlearning through activation patching and parameter restoration experiments. Our findings reveal that these methods alter the model's knowledge retrieval process, providing further evidence that they do not genuinely erase the problematic knowledge embedded in the model parameters. Instead, the coefficients generated by the MLP components in the model's final layer are the primary contributors to these seemingly positive unlearning effects, playing a crucial role in controlling the model's behaviors. Furthermore, behavioral tests demonstrate that this unlearning mechanism inevitably impacts the global behavior of the models, affecting unrelated knowledge or capabilities. The code is released at https://github.com/yihuaihong/Dissecting-FT-Unlearning.
comment: Accepted in EMNLP 2024 Main (Short paper)
♻ ☆ Anderson Acceleration in Nonsmooth Problems: Local Convergence via Active Manifold Identification
Anderson acceleration is an effective technique for enhancing the efficiency of fixed-point iterations; however, analyzing its convergence in nonsmooth settings presents significant challenges. In this paper, we investigate a class of nonsmooth optimization algorithms characterized by the active manifold identification property. This class includes a diverse array of methods such as the proximal point method, proximal gradient method, proximal linear method, proximal coordinate descent method, Douglas-Rachford splitting (or the alternating direction method of multipliers), and the iteratively reweighted $\ell_1$ method, among others. Under the assumption that the optimization problem possesses an active manifold at a stationary point, we establish a local R-linear convergence rate for the Anderson-accelerated algorithm. Our extensive numerical experiments further highlight the robust performance of the proposed Anderson-accelerated methods.
♻ ☆ An accelerated first-order regularized momentum descent ascent algorithm for stochastic nonconvex-concave minimax problems
Stochastic nonconvex minimax problems have attracted wide attention in machine learning, signal processing and many other fields in recent years. In this paper, we propose an accelerated first-order regularized momentum descent ascent algorithm (FORMDA) for solving stochastic nonconvex-concave minimax problems. The iteration complexity of the algorithm is proved to be $\tilde{\mathcal{O}}(\varepsilon ^{-6.5})$ to obtain an $\varepsilon$-stationary point, which achieves the best-known complexity bound for single-loop algorithms to solve the stochastic nonconvex-concave minimax problems under the stationarity of the objective function.
♻ ☆ Mechanistic Permutability: Match Features Across Layers
Understanding how features evolve across layers in deep neural networks is a fundamental challenge in mechanistic interpretability, particularly due to polysemanticity and feature superposition. While Sparse Autoencoders (SAEs) have been used to extract interpretable features from individual layers, aligning these features across layers has remained an open problem. In this paper, we introduce SAE Match, a novel, data-free method for aligning SAE features across different layers of a neural network. Our approach involves matching features by minimizing the mean squared error between the folded parameters of SAEs, a technique that incorporates activation thresholds into the encoder and decoder weights to account for differences in feature scales. Through extensive experiments on the Gemma 2 language model, we demonstrate that our method effectively captures feature evolution across layers, improving feature matching quality. We also show that features persist over several layers and that our approach can approximate hidden states across layers. Our work advances the understanding of feature dynamics in neural networks and provides a new tool for mechanistic interpretability studies.
♻ ☆ GraphCLIP: Enhancing Transferability in Graph Foundation Models for Text-Attributed Graphs
Recently, research on Text-Attributed Graphs (TAGs) has gained significant attention due to the prevalence of free-text node features in real-world applications and the advancements in Large Language Models (LLMs) that bolster TAG methodologies. However, current TAG approaches face two primary challenges: (i) Heavy reliance on label information and (ii) Limited cross-domain zero/few-shot transferability. These issues constrain the scaling of both data and model size, owing to high labor costs and scaling laws, complicating the development of graph foundation models with strong transferability. In this work, we propose the GraphCLIP framework to address these challenges by learning graph foundation models with strong cross-domain zero/few-shot transferability through a self-supervised contrastive graph-summary pretraining method. Specifically, we generate and curate large-scale graph-summary pair data with the assistance of LLMs, and introduce a novel graph-summary pretraining method, combined with invariant learning, to enhance graph foundation models with strong cross-domain zero-shot transferability. For few-shot learning, we propose a novel graph prompt tuning technique aligned with our pretraining objective to mitigate catastrophic forgetting and minimize learning costs. Extensive experiments show the superiority of GraphCLIP in both zero-shot and few-shot settings, while evaluations across various downstream tasks confirm the versatility of GraphCLIP. Our code is available at: https://github.com/ZhuYun97/GraphCLIP
comment: Under Review
♻ ☆ MLKV: Multi-Layer Key-Value Heads for Memory Efficient Transformer Decoding
Auto-regressive inference of transformers benefit greatly from Key-Value (KV) caching, but can lead to major memory bottlenecks as model size, batch size, and sequence length grow at scale. We introduce Multi-Layer Key-Value (MLKV) sharing, a novel approach extending KV sharing across transformer layers to reduce memory usage beyond what was possible with Multi-Query Attention (MQA) and Grouped-Query Attention (GQA). Evaluations on various NLP benchmarks and inference metrics using uptrained Pythia-160M variants demonstrate that MLKV significantly reduces memory usage with minimal performance loss, reducing KV cache size down to a factor of 6x compared to MQA. These results highlight MLKV's potential for efficient deployment of transformer models at scale. We provide code at https://github.com/zaydzuhri/pythia-mlkv
♻ ☆ AnimateLCM: Accelerating the Animation of Personalized Diffusion Models and Adapters with Decoupled Consistency Learning SIGGRAPH
This paper introduces an effective method for computation-efficient personalized style video generation without requiring access to any personalized video data. It reduces the necessary generation time of similarly sized video diffusion models from 25 seconds to around 1 second while maintaining the same level of performance. The method's effectiveness lies in its dual-level decoupling learning approach: 1) separating the learning of video style from video generation acceleration, which allows for personalized style video generation without any personalized style video data, and 2) separating the acceleration of image generation from the acceleration of video motion generation, enhancing training efficiency and mitigating the negative effects of low-quality video data.
comment: Accepted as a Short Paper by SIGGRAPH ASIA 2024 Technical Communications. This is a short version of the original work. Project Page: https://animatelcm.github.io/
♻ ☆ MTL-LoRA: Low-Rank Adaptation for Multi-Task Learning
Parameter-efficient fine-tuning (PEFT) has been widely employed for domain adaptation, with LoRA being one of the most prominent methods due to its simplicity and effectiveness. However, in multi-task learning (MTL) scenarios, LoRA tends to obscure the distinction between tasks by projecting sparse high-dimensional features from different tasks into the same dense low-dimensional intrinsic space. This leads to task interference and suboptimal performance for LoRA and its variants. To tackle this challenge, we propose MTL-LoRA, which retains the advantages of low-rank adaptation while significantly enhancing multi-task learning capabilities. MTL-LoRA augments LoRA by incorporating additional task-adaptive parameters that differentiate task-specific information and effectively capture shared knowledge across various tasks within low-dimensional spaces. This approach enables large language models (LLMs) pre-trained on general corpus to adapt to different target task domains with a limited number of trainable parameters. Comprehensive experimental results, including evaluations on public academic benchmarks for natural language understanding, commonsense reasoning, and image-text understanding, as well as real-world industrial text Ads relevance datasets, demonstrate that MTL-LoRA outperforms LoRA and its various variants with comparable or even fewer learnable parameters in multitask learning.
comment: 12 Pages, 4 Figures
♻ ☆ The Buffer Mechanism for Multi-Step Information Reasoning in Language Models
Large language models have consistently struggled with complex reasoning tasks, such as mathematical problem-solving. Investigating the internal reasoning mechanisms of these models can help us design better model architectures and training strategies, ultimately enhancing their reasoning capability. In this study, we constructed a symbolic dataset to investigate the mechanisms by which Transformer models employ vertical thinking strategy based on their inherent structure and horizontal thinking strategy based on Chain of Thought to achieve multi-step reasoning. We introduced the concept of buffer mechanism: the model stores various information in distinct buffers and selectively extracts them through the query-key matrix. We proposed a random matrix-based algorithm to enhance the model's reasoning ability, resulting in a 75% reduction in the training time required for the GPT-2 model to achieve generalization capability on the PrOntoQA dataset. These findings provide new insights into understanding the mechanisms of large language models.
♻ ☆ Any2Graph: Deep End-To-End Supervised Graph Prediction With An Optimal Transport Loss
We propose Any2graph, a generic framework for end-to-end Supervised Graph Prediction (SGP) i.e. a deep learning model that predicts an entire graph for any kind of input. The framework is built on a novel Optimal Transport loss, the Partially-Masked Fused Gromov-Wasserstein, that exhibits all necessary properties (permutation invariance, differentiability and scalability) and is designed to handle any-sized graphs. Numerical experiments showcase the versatility of the approach that outperform existing competitors on a novel challenging synthetic dataset and a variety of real-world tasks such as map construction from satellite image (Sat2Graph) or molecule prediction from fingerprint (Fingerprint2Graph).
♻ ☆ Latent Schr{ö}dinger Bridge Diffusion Model for Generative Learning
This paper aims to conduct a comprehensive theoretical analysis of current diffusion models. We introduce a novel generative learning methodology utilizing the Schr{\"o}dinger bridge diffusion model in latent space as the framework for theoretical exploration in this domain. Our approach commences with the pre-training of an encoder-decoder architecture using data originating from a distribution that may diverge from the target distribution, thus facilitating the accommodation of a large sample size through the utilization of pre-existing large-scale models. Subsequently, we develop a diffusion model within the latent space utilizing the Schr{\"o}dinger bridge framework. Our theoretical analysis encompasses the establishment of end-to-end error analysis for learning distributions via the latent Schr{\"o}dinger bridge diffusion model. Specifically, we control the second-order Wasserstein distance between the generated distribution and the target distribution. Furthermore, our obtained convergence rates effectively mitigate the curse of dimensionality, offering robust theoretical support for prevailing diffusion models.
♻ ☆ Disentangling Representations through Multi-task Learning
Intelligent perception and interaction with the world hinges on internal representations that capture its underlying structure ("disentangled" or "abstract" representations). Disentangled representations serve as world models, isolating latent factors of variation in the world along orthogonal directions, thus facilitating feature-based generalization. We provide experimental and theoretical results guaranteeing the emergence of disentangled representations in agents that optimally solve multi-task evidence aggregation classification tasks, canonical in the cognitive neuroscience literature. The key conceptual finding is that, by producing accurate multi-task classification estimates, a system implicitly represents a set of coordinates specifying a disentangled representation of the underlying latent state of the data it receives. The theory provides conditions for the emergence of these representations in terms of noise, number of tasks, and evidence aggregation time. We experimentally validate these predictions in RNNs trained on multi-task classification, which learn disentangled representations in the form of continuous attractors, leading to zero-shot out-of-distribution (OOD) generalization in predicting latent factors. We demonstrate the robustness of our framework across autoregressive architectures, decision boundary geometries and in tasks requiring classification confidence estimation. We find that transformers are particularly suited for disentangling representations, which might explain their unique world understanding abilities. Overall, our framework puts forth parallel processing as a general principle for the formation of cognitive maps that capture the structure of the world in both biological and artificial systems, and helps explain why ANNs often arrive at human-interpretable concepts, and how they both may acquire exceptional zero-shot generalization capabilities.
comment: 39 pages, 15 figures
♻ ☆ Improved Depth Estimation of Bayesian Neural Networks NeurIPS 2024
This paper proposes improvements over earlier work by Nazareth and Blei (2022) for estimating the depth of Bayesian neural networks. Here, we propose a discrete truncated normal distribution over the network depth to independently learn its mean and variance. Posterior distributions are inferred by minimizing the variational free energy, which balances the model complexity and accuracy. Our method improves test accuracy on the spiral data set and reduces the variance in posterior depth estimates.
comment: NeurIPS 2024 Workshop on Bayesian Decision-making and Uncertainty. Available at https://openreview.net/forum?id=6TLRVdWGzI
♻ ☆ Score Neural Operator: A Generative Model for Learning and Generalizing Across Multiple Probability Distributions
Most existing generative models are limited to learning a single probability distribution from the training data and cannot generalize to novel distributions for unseen data. An architecture that can generate samples from both trained datasets and unseen probability distributions would mark a significant breakthrough. Recently, score-based generative models have gained considerable attention for their comprehensive mode coverage and high-quality image synthesis, as they effectively learn an operator that maps a probability distribution to its corresponding score function. In this work, we introduce the $\emph{Score Neural Operator}$, which learns the mapping from multiple probability distributions to their score functions within a unified framework. We employ latent space techniques to facilitate the training of score matching, which tends to over-fit in the original image pixel space, thereby enhancing sample generation quality. Our trained Score Neural Operator demonstrates the ability to predict score functions of probability measures beyond the training space and exhibits strong generalization performance in both 2-dimensional Gaussian Mixture Models and 1024-dimensional MNIST double-digit datasets. Importantly, our approach offers significant potential for few-shot learning applications, where a single image from a new distribution can be leveraged to generate multiple distinct images from that distribution.
♻ ☆ Multi-Type Preference Learning: Empowering Preference-Based Reinforcement Learning with Equal Preferences
Preference-Based reinforcement learning (PBRL) learns directly from the preferences of human teachers regarding agent behaviors without needing meticulously designed reward functions. However, existing PBRL methods often learn primarily from explicit preferences, neglecting the possibility that teachers may choose equal preferences. This neglect may hinder the understanding of the agent regarding the task perspective of the teacher, leading to the loss of important information. To address this issue, we introduce the Equal Preference Learning Task, which optimizes the neural network by promoting similar reward predictions when the behaviors of two agents are labeled as equal preferences. Building on this task, we propose a novel PBRL method, Multi-Type Preference Learning (MTPL), which allows simultaneous learning from equal preferences while leveraging existing methods for learning from explicit preferences. To validate our approach, we design experiments applying MTPL to four existing state-of-the-art baselines across ten locomotion and robotic manipulation tasks in the DeepMind Control Suite. The experimental results indicate that simultaneous learning from both equal and explicit preferences enables the PBRL method to more comprehensively understand the feedback from teachers, thereby enhancing feedback efficiency. Project page: \url{https://github.com/FeiCuiLengMMbb/paper_MTPL}
comment: 8 pages, 6 figures
♻ ☆ On the Adversarial Risk of Test Time Adaptation: An Investigation into Realistic Test-Time Data Poisoning
Test-time adaptation (TTA) updates the model weights during the inference stage using testing data to enhance generalization. However, this practice exposes TTA to adversarial risks. Existing studies have shown that when TTA is updated with crafted adversarial test samples, also known as test-time poisoned data, the performance on benign samples can deteriorate. Nonetheless, the perceived adversarial risk may be overstated if the poisoned data is generated under overly strong assumptions. In this work, we first review realistic assumptions for test-time data poisoning, including white-box versus grey-box attacks, access to benign data, attack budget, and more. We then propose an effective and realistic attack method that better produces poisoned samples without access to benign samples, and derive an effective in-distribution attack objective. We also design two TTA-aware attack objectives. Our benchmarks of existing attack methods reveal that the TTA methods are more robust than previously believed. In addition, we analyze effective defense strategies to help develop adversarially robust TTA methods.
comment: 19 pages, 4 figures, 8 tables
♻ ☆ CycleNet: Enhancing Time Series Forecasting through Modeling Periodic Patterns NeurIPS 2024
The stable periodic patterns present in time series data serve as the foundation for conducting long-horizon forecasts. In this paper, we pioneer the exploration of explicitly modeling this periodicity to enhance the performance of models in long-term time series forecasting (LTSF) tasks. Specifically, we introduce the Residual Cycle Forecasting (RCF) technique, which utilizes learnable recurrent cycles to model the inherent periodic patterns within sequences, and then performs predictions on the residual components of the modeled cycles. Combining RCF with a Linear layer or a shallow MLP forms the simple yet powerful method proposed in this paper, called CycleNet. CycleNet achieves state-of-the-art prediction accuracy in multiple domains including electricity, weather, and energy, while offering significant efficiency advantages by reducing over 90% of the required parameter quantity. Furthermore, as a novel plug-and-play technique, the RCF can also significantly improve the prediction accuracy of existing models, including PatchTST and iTransformer. The source code is available at: https://github.com/ACAT-SCUT/CycleNet.
comment: NeurIPS 2024 Spotlight
♻ ☆ Generative Semi-supervised Graph Anomaly Detection NeurIPS 2024
This work considers a practical semi-supervised graph anomaly detection (GAD) scenario, where part of the nodes in a graph are known to be normal, contrasting to the extensively explored unsupervised setting with a fully unlabeled graph. We reveal that having access to the normal nodes, even just a small percentage of normal nodes, helps enhance the detection performance of existing unsupervised GAD methods when they are adapted to the semi-supervised setting. However, their utilization of these normal nodes is limited. In this paper, we propose a novel Generative GAD approach (namely GGAD) for the semi-supervised scenario to better exploit the normal nodes. The key idea is to generate pseudo anomaly nodes, referred to as 'outlier nodes', for providing effective negative node samples in training a discriminative one-class classifier. The main challenge here lies in the lack of ground truth information about real anomaly nodes. To address this challenge, GGAD is designed to leverage two important priors about the anomaly nodes -- asymmetric local affinity and egocentric closeness -- to generate reliable outlier nodes that assimilate anomaly nodes in both graph structure and feature representations. Comprehensive experiments on six real-world GAD datasets are performed to establish a benchmark for semi-supervised GAD and show that GGAD substantially outperforms state-of-the-art unsupervised and semi-supervised GAD methods with varying numbers of training normal nodes. Code will be made available at https://github.com/mala-lab/GGAD.
comment: Accepted by NeurIPS 2024
♻ ☆ A Training-free Sub-quadratic Cost Transformer Model Serving Framework With Hierarchically Pruned Attention
In modern large language models (LLMs), increasing the context length is crucial for improving comprehension and coherence in long-context, multi-modal, and retrieval-augmented language generation. While many recent transformer models attempt to extend their context length over a million tokens, they remain impractical due to the quadratic time and space complexities. Although recent works on linear and sparse attention mechanisms can achieve this goal, their real-world applicability is often limited by the need to re-train from scratch and significantly worse performance. In response, we propose a novel approach, Hierarchically Pruned Attention (HiP), which reduces the time complexity of the attention mechanism to $O(T \log T)$ and the space complexity to $O(T)$, where $T$ is the sequence length. We notice a pattern in the attention scores of pretrained LLMs where tokens close together tend to have similar scores, which we call ``attention locality''. Based on this observation, we utilize a novel tree-search-like algorithm that estimates the top-$k$ key tokens for a given query on the fly, which is mathematically guaranteed to have better performance than random attention pruning. In addition to improving the time complexity of the attention mechanism, we further optimize GPU memory usage by implementing KV cache offloading, which stores only $O(\log T)$ tokens on the GPU while maintaining similar decoding throughput. Experiments on benchmarks show that HiP, with its training-free nature, significantly reduces both prefill and decoding latencies, as well as memory usage, while maintaining high-quality generation with minimal degradation. HiP enables pretrained LLMs to scale up to millions of tokens on commodity GPUs, potentially unlocking long-context LLM applications previously deemed infeasible.
comment: 44 pages
♻ ☆ Variational Mode-Driven Graph Convolutional Network for Spatiotemporal Traffic Forecasting
This paper focuses on spatiotemporal (ST) traffic prediction using graph neural networks. Given that ST data consists of non-stationary and complex time events, interpreting and predicting such trends is comparatively complicated. Representation of ST data in modes helps us to infer behavior and assess the impact of noise on prediction applications. We propose a framework that decomposes ST data into modes using the variational mode decomposition (VMD) method, which is then fed into the neural network for forecasting future states. This hybrid approach is known as a variational mode graph convolutional network (VMGCN). Instead of exhaustively searching for the number of modes, they are determined using the reconstruction loss from the real-time application data. We also study the significance of each mode and the impact of bandwidth constraints on different horizon predictions in traffic flow data. We evaluate the performance of our proposed network on the LargeST dataset for both short and long-term predictions. Our framework yields better results compared to state-of-the-art methods.
comment: IEEE Transactions on Intelligent Transportation Systems Submission, 2024
♻ ☆ STGformer: Efficient Spatiotemporal Graph Transformer for Traffic Forecasting
Traffic forecasting is a cornerstone of smart city management, enabling efficient resource allocation and transportation planning. Deep learning, with its ability to capture complex nonlinear patterns in spatiotemporal (ST) data, has emerged as a powerful tool for traffic forecasting. While graph neural networks (GCNs) and transformer-based models have shown promise, their computational demands often hinder their application to real-world road networks, particularly those with large-scale spatiotemporal interactions. To address these challenges, we propose a novel spatiotemporal graph transformer (STGformer) architecture. STGformer effectively balances the strengths of GCNs and Transformers, enabling efficient modeling of both global and local traffic patterns while maintaining a manageable computational footprint. Unlike traditional approaches that require multiple attention layers, STG attention block captures high-order spatiotemporal interactions in a single layer, significantly reducing computational cost. In particular, STGformer achieves a 100x speedup and a 99.8\% reduction in GPU memory usage compared to STAEformer during batch inference on a California road graph with 8,600 sensors. We evaluate STGformer on the LargeST benchmark and demonstrate its superiority over state-of-the-art Transformer-based methods such as PDFormer and STAEformer, which underline STGformer's potential to revolutionize traffic forecasting by overcoming the computational and memory limitations of existing approaches, making it a promising foundation for future spatiotemporal modeling tasks.
♻ ☆ Recommenadation aided Caching using Combinatorial Multi-armed Bandits
We study content caching with recommendations in a wireless network where the users are connected through a base station equipped with a finite-capacity cache. We assume a fixed set of contents with unknown user preferences and content popularities. The base station can cache a subset of the contents and can also recommend subsets of the contents to different users in order to encourage them to request the recommended contents. Recommendations, depending on their acceptability, can thus be used to increase cache hits. We first assume that the users' recommendation acceptabilities are known and formulate the cache hit optimization problem as a combinatorial multi-armed bandit (CMAB). We propose a UCB-based algorithm to decide which contents to cache and recommend and provide an upper bound on the regret of this algorithm. Subsequently, we consider a more general scenario where the users' recommendation acceptabilities are also unknown and propose another UCB-based algorithm that learns these as well. We numerically demonstrate the performance of our algorithms and compare these to state-of-the-art algorithms.
♻ ☆ Delay as Payoff in MAB
In this paper, we investigate a variant of the classical stochastic Multi-armed Bandit (MAB) problem, where the payoff received by an agent (either cost or reward) is both delayed, and directly corresponds to the magnitude of the delay. This setting models faithfully many real world scenarios such as the time it takes for a data packet to traverse a network given a choice of route (where delay serves as the agent's cost); or a user's time spent on a web page given a choice of content (where delay serves as the agent's reward). Our main contributions are tight upper and lower bounds for both the cost and reward settings. For the case that delays serve as costs, which we are the first to consider, we prove optimal regret that scales as $\sum_{i:\Delta_i > 0}\frac{\log T}{\Delta_i} + d^*$, where $T$ is the maximal number of steps, $\Delta_i$ are the sub-optimality gaps and $d^*$ is the minimal expected delay amongst arms. For the case that delays serves as rewards, we show optimal regret of $\sum_{i:\Delta_i > 0}\frac{\log T}{\Delta_i} + \bar{d}$, where $\bar d$ is the second maximal expected delay. These improve over the regret in the general delay-dependent payoff setting, which scales as $\sum_{i:\Delta_i > 0}\frac{\log T}{\Delta_i} + D$, where $D$ is the maximum possible delay. Our regret bounds highlight the difference between the cost and reward scenarios, showing that the improvement in the cost scenario is more significant than for the reward. Finally, we accompany our theoretical results with an empirical evaluation.
♻ ☆ Rethinking Reward Model Evaluation: Are We Barking up the Wrong Tree?
Reward Models (RMs) are crucial for aligning language models with human preferences. Currently, the evaluation of RMs depends on measuring accuracy against a validation set of manually annotated preference data. Although this method is straightforward and widely adopted, the relationship between RM accuracy and downstream policy performance remains under-explored. In this work, we conduct experiments in a synthetic setting to investigate how differences in RM measured by accuracy translate into gaps in optimized policy performance. Our findings reveal that while there is a weak positive correlation between accuracy and downstream performance, policies optimized towards RMs with similar accuracy can exhibit quite different performance. Moreover, we discover that the way of measuring accuracy significantly impacts its ability to predict the final policy performance. Through the lens of Regressional Goodhart's effect, we identify the existence of exogenous variables impacting the relationship between RM quality measured by accuracy and policy model capability. This underscores the inadequacy of relying solely on accuracy to reflect their impact on policy optimization.
♻ ☆ Edge Unlearning is Not "on Edge"! An Adaptive Exact Unlearning System on Resource-Constrained Devices
The right to be forgotten mandates that machine learning models enable the erasure of a data owner's data and information from a trained model. Removing data from the dataset alone is inadequate, as machine learning models can memorize information from the training data, increasing the potential privacy risk to users. To address this, multiple machine unlearning techniques have been developed and deployed. Among them, approximate unlearning is a popular solution, but recent studies report that its unlearning effectiveness is not fully guaranteed. Another approach, exact unlearning, tackles this issue by discarding the data and retraining the model from scratch, but at the cost of considerable computational and memory resources. However, not all devices have the capability to perform such retraining. In numerous machine learning applications, such as edge devices, Internet-of-Things (IoT), mobile devices, and satellites, resources are constrained, posing challenges for deploying existing exact unlearning methods. In this study, we propose a Constraint-aware Adaptive Exact Unlearning System at the network Edge (CAUSE), an approach to enabling exact unlearning on resource-constrained devices. Aiming to minimize the retrain overhead by storing sub-models on the resource-constrained device, CAUSE innovatively applies a Fibonacci-based replacement strategy and updates the number of shards adaptively in the user-based data partition process. To further improve the effectiveness of memory usage, CAUSE leverages the advantage of model pruning to save memory via compression with minimal accuracy sacrifice. The experimental results demonstrate that CAUSE significantly outperforms other representative systems in realizing exact unlearning on the resource-constrained device by 9.23%-80.86%, 66.21%-83.46%, and 5.26%-194.13% in terms of unlearning speed, energy consumption, and accuracy.
comment: Accepted to IEEE Symposium on Security and Privacy 2025 (Oakland 2025)
♻ ☆ Investigating Annotator Bias in Large Language Models for Hate Speech Detection NeurIPS
Data annotation, the practice of assigning descriptive labels to raw data, is pivotal in optimizing the performance of machine learning models. However, it is a resource-intensive process susceptible to biases introduced by annotators. The emergence of sophisticated Large Language Models (LLMs) presents a unique opportunity to modernize and streamline this complex procedure. While existing research extensively evaluates the efficacy of LLMs, as annotators, this paper delves into the biases present in LLMs when annotating hate speech data. Our research contributes to understanding biases in four key categories: gender, race, religion, and disability with four LLMs: GPT-3.5, GPT-4o, Llama-3.1 and Gemma-2. Specifically targeting highly vulnerable groups within these categories, we analyze annotator biases. Furthermore, we conduct a comprehensive examination of potential factors contributing to these biases by scrutinizing the annotated data. We introduce our custom hate speech detection dataset, HateBiasNet, to conduct this research. Additionally, we perform the same experiments on the ETHOS (Mollas et al. 2022) dataset also for comparative analysis. This paper serves as a crucial resource, guiding researchers and practitioners in harnessing the potential of LLMs for data annotation, thereby fostering advancements in this critical field.
comment: Accepted at NeurIPS Safe Generative AI Workshop, 2024
♻ ☆ Differential Privacy of Cross-Attention with Provable Guarantee
Cross-attention has become a fundamental module nowadays in many important artificial intelligence applications, e.g., retrieval-augmented generation (RAG), system prompt, guided stable diffusion, and many more. Ensuring cross-attention privacy is crucial and urgently needed because its key and value matrices may contain sensitive information about model providers and their users. In this work, we design a novel differential privacy (DP) data structure to address the privacy security of cross-attention with a theoretical guarantee. In detail, let $n$ be the input token length of system prompt/RAG data, $d$ be the feature dimension, $0 < \alpha \le 1$ be the relative error parameter, $R$ be the maximum value of the query and key matrices, $R_w$ be the maximum value of the value matrix, and $r,s,\epsilon_s$ be parameters of polynomial kernel methods. Then, our data structure requires $\widetilde{O}(ndr^2)$ memory consumption with $\widetilde{O}(nr^2)$ initialization time complexity and $\widetilde{O}(\alpha^{-1} r^2)$ query time complexity for a single token query. In addition, our data structure can guarantee that the process of answering user query satisfies $(\epsilon, \delta)$-DP with $\widetilde{O}(n^{-1} \epsilon^{-1} \alpha^{-1/2} R^{2s} R_w r^2)$ additive error and $n^{-1} (\alpha + \epsilon_s)$ relative error between our output and the true answer. Furthermore, our result is robust to adaptive queries in which users can intentionally attack the cross-attention system. To our knowledge, this is the first work to provide DP for cross-attention and is promising to inspire more privacy algorithm design in large generative models (LGMs).
♻ ☆ Multi-Layer Transformers Gradient Can be Approximated in Almost Linear Time
The computational complexity of the self-attention mechanism in popular transformer architectures poses significant challenges for training and inference, and becomes the bottleneck for long inputs. Is it possible to significantly reduce the quadratic time complexity of computing the gradients in multi-layer transformer models? This paper proves that a novel fast approximation method can calculate the gradients in almost linear time $n^{1+o(1)}$ where $n$ is the input sequence length, while it maintains a polynomially small approximation error $1 / \mathrm{poly}(n)$ across the entire model. Our theory holds for general loss functions and when the multi-layer transformer model contains many practical sub-modules, such as residual connection, casual mask, and multi-head attention. By improving the efficiency of gradient computation, we hope that this work will facilitate more effective training and deployment of long-context language models based on our theoretical results.
♻ ☆ Graph Continual Learning with Debiased Lossless Memory Replay ECAI 2024
Real-life graph data often expands continually, rendering the learning of graph neural networks (GNNs) on static graph data impractical. Graph continual learning (GCL) tackles this problem by continually adapting GNNs to the expanded graph of the current task while maintaining the performance over the graph of previous tasks. Memory replay-based methods, which aim to replay data of previous tasks when learning new tasks, have been explored as one principled approach to mitigate the forgetting of the knowledge learned from the previous tasks. In this paper we extend this methodology with a novel framework, called Debiased Lossless Memory replay (DeLoMe). Unlike existing methods that sample nodes/edges of previous graphs to construct the memory, DeLoMe learns small lossless synthetic node representations as the memory. The learned memory can not only preserve the graph data privacy but also capture the holistic graph information, for which the sampling-based methods are not viable. Further, prior methods suffer from bias toward the current task due to the data imbalance between the classes in the memory data and the current data. A debiased GCL loss function is devised in DeLoMe to effectively alleviate this bias. Extensive experiments on four graph datasets show the effectiveness of DeLoMe under both class- and task-incremental learning settings.
comment: Accepted by ECAI 2024
♻ ☆ Variational Offline Multi-agent Skill Discovery
Skills are effective temporal abstractions established for sequential decision making, which enable efficient hierarchical learning for long-horizon tasks and facilitate multi-task learning through their transferability. Despite extensive research, research gaps remain in multi-agent scenarios, particularly for automatically extracting subgroup coordination patterns in a multi-agent task. In this case, we propose two novel auto-encoder schemes: VO-MASD-3D and VO-MASD-Hier, to simultaneously capture subgroup- and temporal-level abstractions and form multi-agent skills, which firstly solves the aforementioned challenge. An essential algorithm component of these schemes is a dynamic grouping function that can automatically detect latent subgroups based on agent interactions in a task. Our method can be applied to offline multi-task data, and the discovered subgroup skills can be transferred across relevant tasks without retraining. Empirical evaluations on StarCraft tasks indicate that our approach significantly outperforms existing hierarchical multi-agent reinforcement learning (MARL) methods. Moreover, skills discovered using our method can effectively reduce the learning difficulty in MARL scenarios with delayed and sparse reward signals.
Multimedia 10
☆ Improving Long-Text Alignment for Text-to-Image Diffusion Models
The rapid advancement of text-to-image (T2I) diffusion models has enabled them to generate unprecedented results from given texts. However, as text inputs become longer, existing encoding methods like CLIP face limitations, and aligning the generated images with long texts becomes challenging. To tackle these issues, we propose LongAlign, which includes a segment-level encoding method for processing long texts and a decomposed preference optimization method for effective alignment training. For segment-level encoding, long texts are divided into multiple segments and processed separately. This method overcomes the maximum input length limits of pretrained encoding models. For preference optimization, we provide decomposed CLIP-based preference models to fine-tune diffusion models. Specifically, to utilize CLIP-based preference models for T2I alignment, we delve into their scoring mechanisms and find that the preference scores can be decomposed into two components: a text-relevant part that measures T2I alignment and a text-irrelevant part that assesses other visual aspects of human preference. Additionally, we find that the text-irrelevant part contributes to a common overfitting problem during fine-tuning. To address this, we propose a reweighting strategy that assigns different weights to these two components, thereby reducing overfitting and enhancing alignment. After fine-tuning $512 \times 512$ Stable Diffusion (SD) v1.5 for about 20 hours using our method, the fine-tuned SD outperforms stronger foundation models in T2I alignment, such as PixArt-$\alpha$ and Kandinsky v2.2. The code is available at https://github.com/luping-liu/LongAlign.
☆ MLLM can see? Dynamic Correction Decoding for Hallucination Mitigation
Multimodal Large Language Models (MLLMs) frequently exhibit hallucination phenomena, but the underlying reasons remain poorly understood. In this paper, we present an empirical analysis and find that, although MLLMs incorrectly generate the objects in the final output, they are actually able to recognize visual objects in the preceding layers. We speculate that this may be due to the strong knowledge priors of the language model suppressing the visual information, leading to hallucinations. Motivated by this, we propose a novel dynamic correction decoding method for MLLMs (DeCo), which adaptively selects the appropriate preceding layers and proportionally integrates knowledge into the final layer to adjust the output logits. Note that DeCo is model agnostic and can be seamlessly incorporated with various classic decoding strategies and applied to different MLLMs. We evaluate DeCo on widely-used benchmarks, demonstrating that it can reduce hallucination rates by a large margin compared to baselines, highlighting its potential to mitigate hallucinations. Code is available at https://github.com/zjunlp/DeCo.
comment: Ongoing work
☆ Magnifier Prompt: Tackling Multimodal Hallucination via Extremely Simple Instructions
Hallucinations in multimodal large language models (MLLMs) hinder their practical applications. To address this, we propose a Magnifier Prompt (MagPrompt), a simple yet effective method to tackle hallucinations in MLLMs via extremely simple instructions. MagPrompt is based on the following two key principles, which guide the design of various effective prompts, demonstrating robustness: (1) MLLMs should focus more on the image. (2) When there are conflicts between the image and the model's inner knowledge, MLLMs should prioritize the image. MagPrompt is training-free and can be applied to open-source and closed-source models, such as GPT-4o and Gemini-pro. It performs well across many datasets and its effectiveness is comparable or even better than more complex methods like VCD. Furthermore, our prompt design principles and experimental analyses provide valuable insights into multimodal hallucination.
comment: 9 pages, 13 tables, 4 figures
☆ On-the-fly Modulation for Balanced Multimodal Learning
Multimodal learning is expected to boost model performance by integrating information from different modalities. However, its potential is not fully exploited because the widely-used joint training strategy, which has a uniform objective for all modalities, leads to imbalanced and under-optimized uni-modal representations. Specifically, we point out that there often exists modality with more discriminative information, e.g., vision of playing football and sound of blowing wind. They could dominate the joint training process, resulting in other modalities being significantly under-optimized. To alleviate this problem, we first analyze the under-optimized phenomenon from both the feed-forward and the back-propagation stages during optimization. Then, On-the-fly Prediction Modulation (OPM) and On-the-fly Gradient Modulation (OGM) strategies are proposed to modulate the optimization of each modality, by monitoring the discriminative discrepancy between modalities during training. Concretely, OPM weakens the influence of the dominant modality by dropping its feature with dynamical probability in the feed-forward stage, while OGM mitigates its gradient in the back-propagation stage. In experiments, our methods demonstrate considerable improvement across a variety of multimodal tasks. These simple yet effective strategies not only enhance performance in vanilla and task-oriented multimodal models, but also in more complex multimodal tasks, showcasing their effectiveness and flexibility. The source code is available at \url{https://github.com/GeWu-Lab/BML_TPAMI2024}.
comment: Accepted by T-PAMI 2024
☆ Leveraging LLM Embeddings for Cross Dataset Label Alignment and Zero Shot Music Emotion Prediction
In this work, we present a novel method for music emotion recognition that leverages Large Language Model (LLM) embeddings for label alignment across multiple datasets and zero-shot prediction on novel categories. First, we compute LLM embeddings for emotion labels and apply non-parametric clustering to group similar labels, across multiple datasets containing disjoint labels. We use these cluster centers to map music features (MERT) to the LLM embedding space. To further enhance the model, we introduce an alignment regularization that enables dissociation of MERT embeddings from different clusters. This further enhances the model's ability to better adaptation to unseen datasets. We demonstrate the effectiveness of our approach by performing zero-shot inference on a new dataset, showcasing its ability to generalize to unseen labels without additional training.
☆ VidCompress: Memory-Enhanced Temporal Compression for Video Understanding in Large Language Models
Video-based multimodal large language models (Video-LLMs) possess significant potential for video understanding tasks. However, most Video-LLMs treat videos as a sequential set of individual frames, which results in insufficient temporal-spatial interaction that hinders fine-grained comprehension and difficulty in processing longer videos due to limited visual token capacity. To address these challenges, we propose VidCompress, a novel Video-LLM featuring memory-enhanced temporal compression. VidCompress employs a dual-compressor approach: a memory-enhanced compressor captures both short-term and long-term temporal relationships in videos and compresses the visual tokens using a multiscale transformer with a memory-cache mechanism, while a text-perceived compressor generates condensed visual tokens by utilizing Q-Former and integrating temporal contexts into query embeddings with cross attention. Experiments on several VideoQA datasets and comprehensive benchmarks demonstrate that VidCompress efficiently models complex temporal-spatial relations and significantly outperforms existing Video-LLMs.
comment: 9 pages, 4 figures
☆ Enabling Data-Driven and Empathetic Interactions: A Context-Aware 3D Virtual Agent in Mixed Reality for Enhanced Financial Customer Experience
In this paper, we introduce a novel system designed to enhance customer service in the financial and retail sectors through a context-aware 3D virtual agent, utilizing Mixed Reality (MR) and Vision Language Models (VLMs). Our approach focuses on enabling data-driven and empathetic interactions that ensure customer satisfaction by introducing situational awareness of the physical location, personalized interactions based on customer profiles, and rigorous privacy and security standards. We discuss our design considerations critical for deployment in real-world customer service environments, addressing challenges in user data management and sensitive information handling. We also outline the system architecture and key features unique to banking and retail environments. Our work demonstrates the potential of integrating MR and VLMs in service industries, offering practical insights in customer service delivery while maintaining high standards of security and personalization.
comment: to appear at 1st Workshop on Intelligent XR: Harnessing AI for Next-Generation XR User Experiences at International Symposium on Mixed and Augmented Reality (ISMAR) 2024
☆ LocoMotion: Learning Motion-Focused Video-Language Representations ACCV 2024
This paper strives for motion-focused video-language representations. Existing methods to learn video-language representations use spatial-focused data, where identifying the objects and scene is often enough to distinguish the relevant caption. We instead propose LocoMotion to learn from motion-focused captions that describe the movement and temporal progression of local object motions. We achieve this by adding synthetic motions to videos and using the parameters of these motions to generate corresponding captions. Furthermore, we propose verb-variation paraphrasing to increase the caption variety and learn the link between primitive motions and high-level verbs. With this, we are able to learn a motion-focused video-language representation. Experiments demonstrate our approach is effective for a variety of downstream tasks, particularly when limited data is available for fine-tuning. Code is available: https://hazeldoughty.github.io/Papers/LocoMotion/
comment: ACCV 2024
♻ ☆ VIA: Unified Spatiotemporal Video Adaptation Framework for Global and Local Video Editing
Video editing is a cornerstone of digital media, from entertainment and education to professional communication. However, previous methods often overlook the necessity of comprehensively understanding both global and local contexts, leading to inaccurate and inconsistent edits in the spatiotemporal dimension, especially for long videos. In this paper, we introduce VIA, a unified spatiotemporal Video Adaptation framework for global and local video editing, pushing the limits of consistently editing minute-long videos. First, to ensure local consistency within individual frames, we designed test-time editing adaptation to adapt a pre-trained image editing model for improving consistency between potential editing directions and the text instruction, and adapt masked latent variables for precise local control. Furthermore, to maintain global consistency over the video sequence, we introduce spatiotemporal adaptation that recursively gather consistent attention variables in key frames and strategically applies them across the whole sequence to realize the editing effects. Extensive experiments demonstrate that, compared to baseline methods, our VIA approach produces edits that are more faithful to the source videos, more coherent in the spatiotemporal context, and more precise in local control. More importantly, we show that VIA can achieve consistent long video editing in minutes, unlocking the potential for advanced video editing tasks over long video sequences.
comment: 19 pages, 14 figures
♻ ☆ FineFake: A Knowledge-Enriched Dataset for Fine-Grained Multi-Domain Fake News Detection
Existing benchmarks for fake news detection have significantly contributed to the advancement of models in assessing the authenticity of news content. However, these benchmarks typically focus solely on news pertaining to a single semantic topic or originating from a single platform, thereby failing to capture the diversity of multi-domain news in real scenarios. In order to understand fake news across various domains, the external knowledge and fine-grained annotations are indispensable to provide precise evidence and uncover the diverse underlying strategies for fabrication, which are also ignored by existing benchmarks. To address this gap, we introduce a novel multi-domain knowledge-enhanced benchmark with fine-grained annotations, named \textbf{FineFake}. FineFake encompasses 16,909 data samples spanning six semantic topics and eight platforms. Each news item is enriched with multi-modal content, potential social context, semi-manually verified common knowledge, and fine-grained annotations that surpass conventional binary labels. Furthermore, we formulate three challenging tasks based on FineFake and propose a knowledge-enhanced domain adaptation network. Extensive experiments are conducted on FineFake under various scenarios, providing accurate and reliable benchmarks for future endeavors. The entire FineFake project is publicly accessible as an open-source repository at \url{https://github.com/Accuser907/FineFake}.
Computation and Language 176
☆ TemporalBench: Benchmarking Fine-grained Temporal Understanding for Multimodal Video Models
Understanding fine-grained temporal dynamics is crucial for multimodal video comprehension and generation. Due to the lack of fine-grained temporal annotations, existing video benchmarks mostly resemble static image benchmarks and are incompetent at evaluating models for temporal understanding. In this paper, we introduce TemporalBench, a new benchmark dedicated to evaluating fine-grained temporal understanding in videos. TemporalBench consists of ~10K video question-answer pairs, derived from ~2K high-quality human annotations detailing the temporal dynamics in video clips. As a result, our benchmark provides a unique testbed for evaluating various temporal understanding and reasoning abilities such as action frequency, motion magnitude, event order, etc. Moreover, it enables evaluations on various tasks like both video question answering and captioning, both short and long video understanding, as well as different models such as multimodal video embedding models and text generation models. Results show that state-of-the-art models like GPT-4o achieve only 38.5% question answering accuracy on TemporalBench, demonstrating a significant gap (~30%) between humans and AI in temporal understanding. Furthermore, we notice a critical pitfall for multi-choice QA where LLMs can detect the subtle changes in negative captions and find a centralized description as a cue for its prediction, where we propose Multiple Binary Accuracy (MBA) to correct such bias. We hope that TemporalBench can foster research on improving models' temporal reasoning capabilities. Both dataset and evaluation code will be made available.
comment: Project Page: https://temporalbench.github.io/
☆ DuoAttention: Efficient Long-Context LLM Inference with Retrieval and Streaming Heads
Deploying long-context large language models (LLMs) is essential but poses significant computational and memory challenges. Caching all Key and Value (KV) states across all attention heads consumes substantial memory. Existing KV cache pruning methods either damage the long-context capabilities of LLMs or offer only limited efficiency improvements. In this paper, we identify that only a fraction of attention heads, a.k.a, Retrieval Heads, are critical for processing long contexts and require full attention across all tokens. In contrast, all other heads, which primarily focus on recent tokens and attention sinks--referred to as Streaming Heads--do not require full attention. Based on this insight, we introduce DuoAttention, a framework that only applies a full KV cache to retrieval heads while using a light-weight, constant-length KV cache for streaming heads, which reduces both LLM's decoding and pre-filling memory and latency without compromising its long-context abilities. DuoAttention uses a lightweight, optimization-based algorithm with synthetic data to identify retrieval heads accurately. Our method significantly reduces long-context inference memory by up to 2.55x for MHA and 1.67x for GQA models while speeding up decoding by up to 2.18x and 1.50x and accelerating pre-filling by up to 1.73x and 1.63x for MHA and GQA models, respectively, with minimal accuracy loss compared to full attention. Notably, combined with quantization, DuoAttention enables Llama-3-8B decoding with 3.3 million context length on a single A100 GPU. Code is provided in https://github.com/mit-han-lab/duo-attention.
☆ LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory
Recent large language model (LLM)-driven chat assistant systems have integrated memory components to track user-assistant chat histories, enabling more accurate and personalized responses. However, their long-term memory capabilities in sustained interactions remain underexplored. This paper introduces LongMemEval, a comprehensive benchmark designed to evaluate five core long-term memory abilities of chat assistants: information extraction, multi-session reasoning, temporal reasoning, knowledge updates, and abstention. With 500 meticulously curated questions embedded within freely scalable user-assistant chat histories, LongMemEval presents a significant challenge to existing long-term memory systems, with commercial chat assistants and long-context LLMs showing 30% accuracy drop on memorizing information across sustained interactions. We then present a unified framework that breaks down the long-term memory design into four design choices across the indexing, retrieval, and reading stages. Built upon key experimental insights, we propose several memory designs including session decomposition for optimizing value granularity, fact-augmented key expansion for enhancing the index structure, and time-aware query expansion for refining the search scope. Experiment results show that these optimizations greatly improve both memory recall and downstream question answering on LongMemEval. Overall, our study provides valuable resources and guidance for advancing the long-term memory capabilities of LLM-based chat assistants, paving the way toward more personalized and reliable conversational AI.
☆ Your Mixture-of-Experts LLM Is Secretly an Embedding Model For Free
While large language models (LLMs) excel on generation tasks, their decoder-only architecture often limits their potential as embedding models if no further representation finetuning is applied. Does this contradict their claim of generalists? To answer the question, we take a closer look at Mixture-of-Experts (MoE) LLMs. Our study shows that the expert routers in MoE LLMs can serve as an off-the-shelf embedding model with promising performance on a diverse class of embedding-focused tasks, without requiring any finetuning. Moreover, our extensive analysis shows that the MoE routing weights (RW) is complementary to the hidden state (HS) of LLMs, a widely-used embedding. Compared to HS, we find that RW is more robust to the choice of prompts and focuses on high-level semantics. Motivated by the analysis, we propose MoEE combining RW and HS, which achieves better performance than using either separately. Our exploration of their combination and prompting strategy shed several novel insights, e.g., a weighted sum of RW and HS similarities outperforms the similarity on their concatenation. Our experiments are conducted on 6 embedding tasks with 20 datasets from the Massive Text Embedding Benchmark (MTEB). The results demonstrate the significant improvement brought by MoEE to LLM-based embedding without further finetuning.
comment: 10 pages, 5 figures
☆ Local and Global Decoding in Text Generation EMNLP 2024
Text generation, a key component in applications such as dialogue systems, relies on decoding algorithms that sample strings from a language model distribution. Traditional methods, such as top-$k$ and top-$\pi$, apply local normalisation to the model's output distribution, which can distort it. In this paper, we investigate the effect of this distortion by introducing globally-normalised versions of these decoding methods. Additionally, we propose an independent Metropolis-Hastings algorithm to approximate sampling from globally-normalised distributions without explicitly computing them. Our empirical analysis compares the performance of local and global normalisation across two decoding algorithms (top-$k$ and top-$\pi$) with various hyperparameters, using Pythia language models. Results show that, in most configurations, global decoding performs worse than the local decoding version of the same algorithms -- despite preserving the distribution's integrity. Our results suggest that distortion is an important feature of local decoding algorithms.
comment: Paper accepted in EMNLP 2024. Code is available in https://github.com/lowlypalace/global-decoding
☆ Mix Data or Merge Models? Optimizing for Diverse Multi-Task Learning
Large Language Models (LLMs) have been adopted and deployed worldwide for a broad variety of applications. However, ensuring their safe use remains a significant challenge. Preference training and safety measures often overfit to harms prevalent in Western-centric datasets, and safety protocols frequently fail to extend to multilingual settings. In this work, we explore model merging in a diverse multi-task setting, combining safety and general-purpose tasks within a multilingual context. Each language introduces unique and varied learning challenges across tasks. We find that objective-based merging is more effective than mixing data, with improvements of up to 8% and 10% in general performance and safety respectively. We also find that language-based merging is highly effective -- by merging monolingually fine-tuned models, we achieve a 4% increase in general performance and 7% reduction in harm across all languages on top of the data mixtures method using the same available data. Overall, our comprehensive study of merging approaches provides a useful framework for building strong and safe multilingual models.
☆ Context-Parametric Inversion: Why Instruction Finetuning May Not Actually Improve Context Reliance
Large language models are instruction-finetuned to enhance their ability to follow user instructions and process the input context. However, even state-of-the-art models often struggle to follow the instruction, especially when the input context is not aligned with the model's parametric knowledge. This manifests as various failures, such as hallucinations where the responses are outdated, biased or contain unverified facts. In this work, we try to understand the underlying reason for this poor context reliance, especially after instruction tuning. We observe an intriguing phenomenon: during instruction tuning, the context reliance initially increases as expected, but then gradually decreases as instruction finetuning progresses. We call this phenomenon context-parametric inversion and observe it across multiple general purpose instruction tuning datasets like TULU, Alpaca and Ultrachat, as well as model families such as Llama, Mistral and Pythia. In a simple theoretical setup, we isolate why context-parametric inversion occurs along the gradient descent trajectory of instruction finetuning. We tie this phenomena to examples in the instruction finetuning data mixture where the input context provides information that is already present in the model's parametric knowledge. Our analysis suggests natural mitigation strategies that provide some limited gains, while also validating our theoretical insights. We hope that our work serves as a starting point in addressing this failure mode in a staple part of LLM training.
comment: Under Review
☆ When Attention Sink Emerges in Language Models: An Empirical View
Language Models (LMs) assign significant attention to the first token, even if it is not semantically important, which is known as attention sink. This phenomenon has been widely adopted in applications such as streaming/long context generation, KV cache optimization, inference acceleration, model quantization, and others. Despite its widespread use, a deep understanding of attention sink in LMs is still lacking. In this work, we first demonstrate that attention sinks exist universally in LMs with various inputs, even in small models. Furthermore, attention sink is observed to emerge during the LM pre-training, motivating us to investigate how optimization, data distribution, loss function, and model architecture in LM pre-training influence its emergence. We highlight that attention sink emerges after effective optimization on sufficient training data. The sink position is highly correlated with the loss function and data distribution. Most importantly, we find that attention sink acts more like key biases, storing extra attention scores, which could be non-informative and not contribute to the value computation. We also observe that this phenomenon (at least partially) stems from tokens' inner dependence on attention scores as a result of softmax normalization. After relaxing such dependence by replacing softmax attention with other attention operations, such as sigmoid attention without normalization, attention sinks do not emerge in LMs up to 1B parameters. The code is available at https://github.com/sail-sg/Attention-Sink.
☆ AFlow: Automating Agentic Workflow Generation
Large language models (LLMs) have demonstrated remarkable potential in solving complex tasks across diverse domains, typically by employing agentic workflows that follow detailed instructions and operational sequences. However, constructing these workflows requires significant human effort, limiting scalability and generalizability. Recent research has sought to automate the generation and optimization of these workflows, but existing methods still rely on initial manual setup and fall short of achieving fully automated and effective workflow generation. To address this challenge, we reformulate workflow optimization as a search problem over code-represented workflows, where LLM-invoking nodes are connected by edges. We introduce AFlow, an automated framework that efficiently explores this space using Monte Carlo Tree Search, iteratively refining workflows through code modification, tree-structured experience, and execution feedback. Empirical evaluations across six benchmark datasets demonstrate AFlow's efficacy, yielding a 5.7% average improvement over state-of-the-art baselines. Furthermore, AFlow enables smaller models to outperform GPT-4o on specific tasks at 4.55% of its inference cost in dollars. The code will be available at https://github.com/geekan/MetaGPT.
☆ Denial-of-Service Poisoning Attacks against Large Language Models
Recent studies have shown that LLMs are vulnerable to denial-of-service (DoS) attacks, where adversarial inputs like spelling errors or non-semantic prompts trigger endless outputs without generating an [EOS] token. These attacks can potentially cause high latency and make LLM services inaccessible to other users or tasks. However, when there are speech-to-text interfaces (e.g., voice commands to a robot), executing such DoS attacks becomes challenging, as it is difficult to introduce spelling errors or non-semantic prompts through speech. A simple DoS attack in these scenarios would be to instruct the model to "Keep repeating Hello", but we observe that relying solely on natural instructions limits output length, which is bounded by the maximum length of the LLM's supervised finetuning (SFT) data. To overcome this limitation, we propose poisoning-based DoS (P-DoS) attacks for LLMs, demonstrating that injecting a single poisoned sample designed for DoS purposes can break the output length limit. For example, a poisoned sample can successfully attack GPT-4o and GPT-4o mini (via OpenAI's finetuning API) using less than $1, causing repeated outputs up to the maximum inference length (16K tokens, compared to 0.5K before poisoning). Additionally, we perform comprehensive ablation studies on open-source LLMs and extend our method to LLM agents, where attackers can control both the finetuning dataset and algorithm. Our findings underscore the urgent need for defenses against P-DoS attacks to secure LLMs. Our code is available at https://github.com/sail-sg/P-DoS.
☆ Use Random Selection for Now: Investigation of Few-Shot Selection Strategies in LLM-based Text Augmentation for Classification
The generative large language models (LLMs) are increasingly used for data augmentation tasks, where text samples are paraphrased (or generated anew) and then used for classifier fine-tuning. Existing works on augmentation leverage the few-shot scenarios, where samples are given to LLMs as part of prompts, leading to better augmentations. Yet, the samples are mostly selected randomly and a comprehensive overview of the effects of other (more ``informed'') sample selection strategies is lacking. In this work, we compare sample selection strategies existing in few-shot learning literature and investigate their effects in LLM-based textual augmentation. We evaluate this on in-distribution and out-of-distribution classifier performance. Results indicate, that while some ``informed'' selection strategies increase the performance of models, especially for out-of-distribution data, it happens only seldom and with marginal performance increases. Unless further advances are made, a default of random sample selection remains a good option for augmentation practitioners.
☆ Balancing Continuous Pre-Training and Instruction Fine-Tuning: Optimizing Instruction-Following in LLMs
Large Language Models (LLMs) for public use require continuous pre-training to remain up-to-date with the latest data. The models also need to be fine-tuned with specific instructions to maintain their ability to follow instructions accurately. Typically, LLMs are released in two versions: the Base LLM, pre-trained on diverse data, and the instruction-refined LLM, additionally trained with specific instructions for better instruction following. The question arises as to which model should undergo continuous pre-training to maintain its instruction-following abilities while also staying current with the latest data. In this study, we delve into the intricate relationship between continuous pre-training and instruction fine-tuning of the LLMs and investigate the impact of continuous pre-training on the instruction following abilities of both the base and its instruction finetuned model. Further, the instruction fine-tuning process is computationally intense and requires a substantial number of hand-annotated examples for the model to learn effectively. This study aims to find the most compute-efficient strategy to gain up-to-date knowledge and instruction-following capabilities without requiring any instruction data and fine-tuning. We empirically prove our findings on the LLaMa 3, 3.1 and Qwen 2, 2.5 family of base and instruction models, providing a comprehensive exploration of our hypotheses across varying sizes of pre-training data corpus and different LLMs settings.
☆ Embedding Self-Correction as an Inherent Ability in Large Language Models for Enhanced Mathematical Reasoning
Accurate mathematical reasoning with Large Language Models (LLMs) is crucial in revolutionizing domains that heavily rely on such reasoning. However, LLMs often encounter difficulties in certain aspects of mathematical reasoning, leading to flawed reasoning and erroneous results. To mitigate these issues, we introduce a novel mechanism, the Chain of Self-Correction (CoSC), specifically designed to embed self-correction as an inherent ability in LLMs, enabling them to validate and rectify their own results. The CoSC mechanism operates through a sequence of self-correction stages. In each stage, the LLMs generate a program to address a given problem, execute this program using program-based tools to obtain an output, subsequently verify this output. Based on the verification, the LLMs either proceed to the next correction stage or finalize the answer. This iterative self-correction process allows the LLMs to refine their reasoning steps and improve the accuracy of their mathematical reasoning. To enable the CoSC mechanism at a low cost, we employ a two-phase finetuning approach. In the first phase, the LLMs are trained with a relatively small volume of seeding data generated from GPT-4, establishing an initial CoSC capability. In the second phase, the CoSC capability is further enhanced by training with a larger volume of self-generated data using the trained model in the first phase, without relying on the paid GPT-4. Our comprehensive experiments demonstrate that CoSC significantly improves performance on traditional mathematical datasets among existing open-source LLMs. Notably, our CoSC-Code-34B model achieved a 53.5% score on MATH, the most challenging mathematical reasoning dataset in the public domain, surpassing the performance of well-established models such as ChatGPT, GPT-4, and even multi-modal LLMs like GPT-4V, Gemini-1.0 Pro, and Gemini-1.0 Ultra.
☆ Large Language Models Are Active Critics in NLG Evaluation ICLR2025
The conventional paradigm of using large language models (LLMs) for evaluating natural language generation (NLG) systems typically relies on two key inputs: (1) a clear definition of the NLG task to be evaluated and (2) a list of pre-defined evaluation criteria. This process treats LLMs as ''passive critics,'' strictly following human-defined criteria for evaluation. However, as new NLG tasks emerge, the criteria for assessing text quality can vary greatly. Consequently, these rigid evaluation methods struggle to adapt to diverse NLG tasks without extensive prompt engineering customized for each specific task. To address this limitation, we introduce Active-Critic, a novel LLM-based NLG evaluation protocol that enables LLMs to function as ''active critics.'' Specifically, our protocol comprises two key stages. In the first stage, the LLM is instructed to infer the target NLG task and establish relevant evaluation criteria from the data. Building on this self-inferred information, the second stage dynamically optimizes the prompt to guide the LLM toward more human-aligned scoring decisions, while also generating detailed explanations to justify its evaluations. Experiments across four NLG evaluation tasks show that our approach achieves stronger alignment with human judgments than state-of-the-art evaluation methods. Our comprehensive analysis further highlights the effectiveness and explainability of Active-Critic with only a small amount of labeled data. We will share our code and data on GitHub.
comment: Submitted to ICLR2025
☆ Derail Yourself: Multi-turn LLM Jailbreak Attack through Self-discovered Clues
This study exposes the safety vulnerabilities of Large Language Models (LLMs) in multi-turn interactions, where malicious users can obscure harmful intents across several queries. We introduce ActorAttack, a novel multi-turn attack method inspired by actor-network theory, which models a network of semantically linked actors as attack clues to generate diverse and effective attack paths toward harmful targets. ActorAttack addresses two main challenges in multi-turn attacks: (1) concealing harmful intents by creating an innocuous conversation topic about the actor, and (2) uncovering diverse attack paths towards the same harmful target by leveraging LLMs' knowledge to specify the correlated actors as various attack clues. In this way, ActorAttack outperforms existing single-turn and multi-turn attack methods across advanced aligned LLMs, even for GPT-o1. We will publish a dataset called SafeMTData, which includes multi-turn adversarial prompts and safety alignment data, generated by ActorAttack. We demonstrate that models safety-tuned using our safety dataset are more robust to multi-turn attacks. Code is available at https://github.com/renqibing/ActorAttack.
☆ Building a Multivariate Time Series Benchmarking Datasets Inspired by Natural Language Processing (NLP)
Time series analysis has become increasingly important in various domains, and developing effective models relies heavily on high-quality benchmark datasets. Inspired by the success of Natural Language Processing (NLP) benchmark datasets in advancing pre-trained models, we propose a new approach to create a comprehensive benchmark dataset for time series analysis. This paper explores the methodologies used in NLP benchmark dataset creation and adapts them to the unique challenges of time series data. We discuss the process of curating diverse, representative, and challenging time series datasets, highlighting the importance of domain relevance and data complexity. Additionally, we investigate multi-task learning strategies that leverage the benchmark dataset to enhance the performance of time series models. This research contributes to the broader goal of advancing the state-of-the-art in time series modeling by adopting successful strategies from the NLP domain.
☆ Large Language Model Evaluation via Matrix Nuclear-Norm
As large language models (LLMs) continue to evolve, efficient evaluation metrics are vital for assessing their ability to compress information and reduce redundancy. While traditional metrics like Matrix Entropy offer valuable insights, they are computationally intensive for large-scale models due to their \( O(n^3) \) time complexity with Singular Value Decomposition (SVD). To mitigate this issue, we introduce the Matrix Nuclear-Norm, which not only serves as a metric to quantify the data compression proficiency of LLM but also provides a convex approximation of matrix rank to capture both predictive discriminability and diversity. By employing the \( L_{1,2}\text{-norm} \) to further approximate the nuclear norm, we can effectively assess the model's information compression capabilities. This approach reduces the time complexity to \( O(n^2) \) and eliminates the need for SVD computation. Consequently, the Matrix Nuclear-Norm achieves speeds 8 to 24 times faster than Matrix Entropy for the CEREBRAS-GPT model as sizes increase from 111M to 6.7B. This performance gap becomes more pronounced with larger models, as validated in tests with other models like Pythia. Additionally, evaluations on benchmarks and model responses confirm that our proposed Matrix Nuclear-Norm is a reliable, scalable, and efficient tool for assessing LLMs' performance, striking a balance between accuracy and computational efficiency. The code is available at https://github.com/MLGroupJLU/MatrixNuclearNorm.
comment: 22 pages
☆ Double Jeopardy and Climate Impact in the Use of Large Language Models: Socio-economic Disparities and Reduced Utility for Non-English Speakers
Artificial Intelligence (AI), particularly large language models (LLMs), holds the potential to bridge language and information gaps, which can benefit the economies of developing nations. However, our analysis of FLORES-200, FLORES+, Ethnologue, and World Development Indicators data reveals that these benefits largely favor English speakers. Speakers of languages in low-income and lower-middle-income countries face higher costs when using OpenAI's GPT models via APIs because of how the system processes the input -- tokenization. Around 1.5 billion people, speaking languages primarily from lower-middle-income countries, could incur costs that are 4 to 6 times higher than those faced by English speakers. Disparities in LLM performance are significant, and tokenization in models priced per token amplifies inequalities in access, cost, and utility. Moreover, using the quality of translation tasks as a proxy measure, we show that LLMs perform poorly in low-resource languages, presenting a ``double jeopardy" of higher costs and poor performance for these users. We also discuss the direct impact of fragmentation in tokenizing low-resource languages on climate. This underscores the need for fairer algorithm development to benefit all linguistic groups.
comment: Project GitHub repository at https://github.com/worldbank/double-jeopardy-in-llms
☆ Generative AI and Its Impact on Personalized Intelligent Tutoring Systems
Generative Artificial Intelligence (AI) is revolutionizing educational technology by enabling highly personalized and adaptive learning environments within Intelligent Tutoring Systems (ITS). This report delves into the integration of Generative AI, particularly large language models (LLMs) like GPT-4, into ITS to enhance personalized education through dynamic content generation, real-time feedback, and adaptive learning pathways. We explore key applications such as automated question generation, customized feedback mechanisms, and interactive dialogue systems that respond to individual learner needs. The report also addresses significant challenges, including ensuring pedagogical accuracy, mitigating inherent biases in AI models, and maintaining learner engagement. Future directions highlight the potential advancements in multimodal AI integration, emotional intelligence in tutoring systems, and the ethical implications of AI-driven education. By synthesizing current research and practical implementations, this report underscores the transformative potential of Generative AI in creating more effective, equitable, and engaging educational experiences.
comment: Scientific Report (Under Review)
☆ Thinking LLMs: General Instruction Following with Thought Generation
LLMs are typically trained to answer user questions or follow instructions similarly to how human experts respond. However, in the standard alignment framework they lack the basic ability of explicit thinking before answering. Thinking is important for complex questions that require reasoning and planning -- but can be applied to any task. We propose a training method for equipping existing LLMs with such thinking abilities for general instruction following without use of additional human data. We achieve this by an iterative search and optimization procedure that explores the space of possible thought generations, allowing the model to learn how to think without direct supervision. For each instruction, the thought candidates are scored using a judge model to evaluate their responses only, and then optimized via preference optimization. We show that this procedure leads to superior performance on AlpacaEval and Arena-Hard, and shows gains from thinking on non-reasoning categories such as marketing, health and general knowledge, in addition to more traditional reasoning & problem-solving tasks.
☆ Efficiently Democratizing Medical LLMs for 50 Languages via a Mixture of Language Family Experts
Adapting medical Large Language Models to local languages can reduce barriers to accessing healthcare services, but data scarcity remains a significant challenge, particularly for low-resource languages. To address this, we first construct a high-quality medical dataset and conduct analysis to ensure its quality. In order to leverage the generalization capability of multilingual LLMs to efficiently scale to more resource-constrained languages, we explore the internal information flow of LLMs from a multilingual perspective using Mixture of Experts (MoE) modularity. Technically, we propose a novel MoE routing method that employs language-specific experts and cross-lingual routing. Inspired by circuit theory, our routing analysis revealed a Spread Out in the End information flow mechanism: while earlier layers concentrate cross-lingual information flow, the later layers exhibit language-specific divergence. This insight directly led to the development of the Post-MoE architecture, which applies sparse routing only in the later layers while maintaining dense others. Experimental results demonstrate that this approach enhances the generalization of multilingual models to other languages while preserving interpretability. Finally, to efficiently scale the model to 50 languages, we introduce the concept of language family experts, drawing on linguistic priors, which enables scaling the number of languages without adding additional parameters.
☆ SensorLLM: Aligning Large Language Models with Motion Sensors for Human Activity Recognition
In this work, we bridge the gap between wearable sensor technology and personalized AI assistants by enabling Large Language Models (LLMs) to understand time-series tasks like human activity recognition (HAR). Despite the strong reasoning and generalization capabilities of LLMs, leveraging them for sensor data tasks remains largely unexplored. This gap stems from challenges like the lack of semantic context in time-series data, computational limitations, and LLMs' difficulty processing numerical inputs. To address these issues, we introduce SensorLLM, a two-stage framework to unlock LLMs' potential for sensor data tasks. In the Sensor-Language Alignment Stage, we introduce special tokens for each sensor channel and automatically generate trend-descriptive text to align sensor data with textual inputs, enabling SensorLLM to capture numerical changes, channel-specific information, and sensor data of varying lengths-capabilities that existing LLMs typically struggle with, all without the need for human annotations. Next, in Task-Aware Tuning Stage, we refine the model for HAR classification using the frozen LLM and alignment module, achieving performance on par with or surpassing state-of-the-art models. We further demonstrate that SensorLLM evolves into an effective sensor learner, reasoner, and classifier through Sensor-Language Alignment, enabling it to generalize across diverse datasets for HAR tasks. We strongly believe our work lays the stepstone for future time-series and text alignment research, offering a path toward foundation models for sensor data.
☆ Modeling News Interactions and Influence for Financial Market Prediction EMNLP 2024
The diffusion of financial news into market prices is a complex process, making it challenging to evaluate the connections between news events and market movements. This paper introduces FININ (Financial Interconnected News Influence Network), a novel market prediction model that captures not only the links between news and prices but also the interactions among news items themselves. FININ effectively integrates multi-modal information from both market data and news articles. We conduct extensive experiments on two datasets, encompassing the S&P 500 and NASDAQ 100 indices over a 15-year period and over 2.7 million news articles. The results demonstrate FININ's effectiveness, outperforming advanced market prediction models with an improvement of 0.429 and 0.341 in the daily Sharpe ratio for the two markets respectively. Moreover, our results reveal insights into the financial news, including the delayed market pricing of news, the long memory effect of news, and the limitations of financial sentiment analysis in fully extracting predictive power from news data.
comment: Accepted by EMNLP 2024
☆ VisRAG: Vision-based Retrieval-augmented Generation on Multi-modality Documents
Retrieval-augmented generation (RAG) is an effective technique that enables large language models (LLMs) to utilize external knowledge sources for generation. However, current RAG systems are solely based on text, rendering it impossible to utilize vision information like layout and images that play crucial roles in real-world multi-modality documents. In this paper, we introduce VisRAG, which tackles this issue by establishing a vision-language model (VLM)-based RAG pipeline. In this pipeline, instead of first parsing the document to obtain text, the document is directly embedded using a VLM as an image and then retrieved to enhance the generation of a VLM. Compared to traditional text-based RAG, VisRAG maximizes the retention and utilization of the data information in the original documents, eliminating the information loss introduced during the parsing process. We collect both open-source and synthetic data to train the retriever in VisRAG and explore a variety of generation methods. Experiments demonstrate that VisRAG outperforms traditional RAG in both the retrieval and generation stages, achieving a 25--39\% end-to-end performance gain over traditional text-based RAG pipeline. Further analysis reveals that VisRAG is effective in utilizing training data and demonstrates strong generalization capability, positioning it as a promising solution for RAG on multi-modality documents. Our code and data are available at https://github.com/openbmb/visrag .
☆ Tübingen-CL at SemEval-2024 Task 1:Ensemble Learning for Semantic Relatedness Estimation
The paper introduces our system for SemEval-2024 Task 1, which aims to predict the relatedness of sentence pairs. Operating under the hypothesis that semantic relatedness is a broader concept that extends beyond mere similarity of sentences, our approach seeks to identify useful features for relatedness estimation. We employ an ensemble approach integrating various systems, including statistical textual features and outputs of deep learning models to predict relatedness scores. The findings suggest that semantic relatedness can be inferred from various sources and ensemble models outperform many individual systems in estimating semantic relatedness.
comment: 5 pages
☆ Multilingual Controlled Generation And Gold-Standard-Agnostic Evaluation of Code-Mixed Sentences COLING 2025
Code-mixing, the practice of alternating between two or more languages in an utterance, is a common phenomenon in multilingual communities. Due to the colloquial nature of code-mixing, there is no singular correct way to translate an English sentence into a code-mixed sentence. For this reason, standard n-gram-based MT evaluation metrics such as the BLEU score are not appropriate for code-mixed evaluation. To demonstrate this, we propose a novel method for code-mixed text generation: Controlled Generation, which parameterizes the code-mixing degree (CMD) and enables the generation of multiple semantically equivalent code-mixed sentences from a given English sentence. We introduce a robust new evaluation metric: GAME: A Gold-Standard Agnostic Measure for Evaluation of Code-Mixed Sentences. GAME is both language-agnostic and gold-standard-agnostic, i.e. unlike other metrics, GAME does not require gold-standard code-mixed sentences for evaluation, thus eliminating the need for human annotators in the code-mixed evaluation process. When used to evaluate semantically equivalent code-mixed sentences, we find that GAME scores have a lower standard deviation than BLEU scores. Further, we create and release a dataset containing gold-standard code-mixed sentences across 4 language pairs: English-{Hindi, Bengali, French, Spanish} to encourage more computational research on code-mixing.
comment: Manuscript submitted to COLING 2025
☆ Recipe for Zero-shot POS Tagging: Is It Useful in Realistic Scenarios? EMNLP 2024
POS tagging plays a fundamental role in numerous applications. While POS taggers are highly accurate in well-resourced settings, they lag behind in cases of limited or missing training data. This paper focuses on POS tagging for languages with limited data. We seek to identify the characteristics of datasets that make them favourable for training POS tagging models without using any labelled training data from the target language. This is a zero-shot approach. We compare the accuracies of a multilingual large language model (mBERT) fine-tuned on one or more languages related to the target language. Additionally, we compare these results with models trained directly on the target language itself. We do this for three target low-resource languages. Our research highlights the importance of accurate dataset selection for effective zero-shot POS tagging. Particularly, a strong linguistic relationship and high-quality datasets ensure optimal results. For extremely low-resource languages, zero-shot models prove to be a viable option.
comment: To appear at the 4th Multilingual NLP workshop collocated with EMNLP 2024
☆ Is Structure Dependence Shaped for Efficient Communication?: A Case Study on Coordination CoNLL 2024
Natural language exhibits various universal properties. But why do these universals exist? One explanation is that they arise from functional pressures to achieve efficient communication, a view which attributes cross-linguistic properties to domain-general cognitive abilities. This hypothesis has successfully addressed some syntactic universal properties such as compositionality and Greenbergian word order universals. However, more abstract syntactic universals have not been explored from the perspective of efficient communication. Among such universals, the most notable one is structure dependence, that is, the existence of grammar-internal operations that crucially depend on hierarchical representations. This property has traditionally been taken to be central to natural language and to involve domain-specific knowledge irreducible to communicative efficiency. In this paper, we challenge the conventional view by investigating whether structure dependence realizes efficient communication, focusing on coordinate structures. We design three types of artificial languages: (i) one with a structure-dependent reduction operation, which is similar to natural language, (ii) one without any reduction operations, and (iii) one with a linear (rather than structure-dependent) reduction operation. We quantify the communicative efficiency of these languages. The results demonstrate that the language with the structure-dependent reduction operation is significantly more communicatively efficient than the counterfactual languages. This suggests that the existence of structure-dependent properties can be explained from the perspective of efficient communication.
comment: CoNLL 2024
☆ SLaNC: Static LayerNorm Calibration NeurIPS 2024
The ever increasing sizes of Large Language Models (LLMs) beyond hundreds of billions of parameters have generated enormous pressure on the manufacturers of dedicated hardware accelerators and made the innovative design of the latter one of the most rapidly expanding fields of the AI industry. Various approaches have been explored to enable efficient and accurate processing of LLMs on the available accelerators given their computational and storage limitations. Among these, various quantization techniques have become the main focus of the community as a means of reducing the compute, communication and storage requirements. Quantization to lower precision formats naturally poses a number of challenges caused by the limited range of the available value representations. When it comes to processing the popular Transformer models on hardware, one of the main issues becomes calculation of the LayerNorm simply because accumulation of the variance requires a much wider dynamic range than the hardware enables. In this article, we address this matter and propose a computationally-efficient scaling technique that can be easily applied to Transformer models during inference. Our method suggests a straightforward way of scaling the LayerNorm inputs based on the static weights of the immediately preceding linear layers. The scaling factors are computed offline, based solely on the linear layer weights, hence no latency or computational overhead is added during inference. Most importantly, our technique ensures that no numerical issues such as overflow or underflow could happen during the compute. This approach offers smooth, accurate and resource-effective inference across a wide range of hardware architectures. The article provides theoretical justification as well as supporting numerical simulations.
comment: 9 pages, 3 figures, NeurIPS 2024 MLNCP Workshop
☆ Rethinking Legal Judgement Prediction in a Realistic Scenario in the Era of Large Language Models EMNLP 2024
This study investigates judgment prediction in a realistic scenario within the context of Indian judgments, utilizing a range of transformer-based models, including InLegalBERT, BERT, and XLNet, alongside LLMs such as Llama-2 and GPT-3.5 Turbo. In this realistic scenario, we simulate how judgments are predicted at the point when a case is presented for a decision in court, using only the information available at that time, such as the facts of the case, statutes, precedents, and arguments. This approach mimics real-world conditions, where decisions must be made without the benefit of hindsight, unlike retrospective analyses often found in previous studies. For transformer models, we experiment with hierarchical transformers and the summarization of judgment facts to optimize input for these models. Our experiments with LLMs reveal that GPT-3.5 Turbo excels in realistic scenarios, demonstrating robust performance in judgment prediction. Furthermore, incorporating additional legal information, such as statutes and precedents, significantly improves the outcome of the prediction task. The LLMs also provide explanations for their predictions. To evaluate the quality of these predictions and explanations, we introduce two human evaluation metrics: Clarity and Linking. Our findings from both automatic and human evaluations indicate that, despite advancements in LLMs, they are yet to achieve expert-level performance in judgment prediction and explanation tasks.
comment: Accepted on NLLP at EMNLP 2024
☆ Everyday Speech in the Indian Subcontinent ICASSP 2025
India has 1369 languages of which 22 are official. About 13 different scripts are used to represent these languages. A Common Label Set (CLS) was developed based on phonetics to address the issue of large vocabulary of units required in the End to End (E2E) framework for multilingual synthesis. This reduced the footprint of the synthesizer and also enabled fast adaptation to new languages which had similar phonotactics, provided language scripts belonged to the same family. In this paper, we provide new insights into speech synthesis, where the script belongs to one family, while the phonotactics comes from another. Indian language text is first converted to CLS, and then a synthesizer that matches the phonotactics of the language is used. Quality akin to that of a native speaker is obtained for Sanskrit and Konkani with zero adaptation data, using Kannada and Marathi synthesizers respectively. Further, this approach also lends itself seamless code switching across 13 Indian languages and English in a given native speaker's voice.
comment: 5 Pages, 1 Figure, Submitted to ICASSP 2025
☆ Cultural Fidelity in Large-Language Models: An Evaluation of Online Language Resources as a Driver of Model Performance in Value Representation
The training data for LLMs embeds societal values, increasing their familiarity with the language's culture. Our analysis found that 44% of the variance in the ability of GPT-4o to reflect the societal values of a country, as measured by the World Values Survey, correlates with the availability of digital resources in that language. Notably, the error rate was more than five times higher for the languages of the lowest resource compared to the languages of the highest resource. For GPT-4-turbo, this correlation rose to 72%, suggesting efforts to improve the familiarity with the non-English language beyond the web-scraped data. Our study developed one of the largest and most robust datasets in this topic area with 21 country-language pairs, each of which contain 94 survey questions verified by native speakers. Our results highlight the link between LLM performance and digital data availability in target languages. Weaker performance in low-resource languages, especially prominent in the Global South, may worsen digital divides. We discuss strategies proposed to address this, including developing multilingual LLMs from the ground up and enhancing fine-tuning on diverse linguistic datasets, as seen in African language initiatives.
☆ Will LLMs Replace the Encoder-Only Models in Temporal Relation Classification?
The automatic detection of temporal relations among events has been mainly investigated with encoder-only models such as RoBERTa. Large Language Models (LLM) have recently shown promising performance in temporal reasoning tasks such as temporal question answering. Nevertheless, recent studies have tested the LLMs' performance in detecting temporal relations of closed-source models only, limiting the interpretability of those results. In this work, we investigate LLMs' performance and decision process in the Temporal Relation Classification task. First, we assess the performance of seven open and closed-sourced LLMs experimenting with in-context learning and lightweight fine-tuning approaches. Results show that LLMs with in-context learning significantly underperform smaller encoder-only models based on RoBERTa. Then, we delve into the possible reasons for this gap by applying explainable methods. The outcome suggests a limitation of LLMs in this task due to their autoregressive nature, which causes them to focus only on the last part of the sequence. Additionally, we evaluate the word embeddings of these two models to better understand their pre-training differences. The code and the fine-tuned models can be found respectively on GitHub.
☆ Ada-K Routing: Boosting the Efficiency of MoE-based LLMs
In the era of Large Language Models (LLMs), Mixture-of-Experts (MoE) architectures offer a promising approach to managing computational costs while scaling up model parameters. Conventional MoE-based LLMs typically employ static Top-K routing, which activates a fixed and equal number of experts for each token regardless of their significance within the context. In this paper, we propose a novel Ada-K routing strategy that dynamically adjusts the number of activated experts for each token, thereby improving the balance between computational efficiency and model performance. Specifically, our strategy incorporates learnable and lightweight allocator modules that decide customized expert resource allocation tailored to the contextual needs for each token. These allocators are designed to be fully pluggable, making it broadly applicable across all mainstream MoE-based LLMs. We leverage the Proximal Policy Optimization (PPO) algorithm to facilitate an end-to-end learning process for this non-differentiable decision-making framework. Extensive evaluations on four popular baseline models demonstrate that our Ada-K routing method significantly outperforms conventional Top-K routing. Compared to Top-K, our method achieves over 25% reduction in FLOPs and more than 20% inference speedup while still improving performance across various benchmarks. Moreover, the training of Ada-K is highly efficient. Even for Mixtral-8x22B, a MoE-based LLM with more than 140B parameters, the training time is limited to 8 hours. Detailed analysis shows that harder tasks, middle layers, and content words tend to activate more experts, providing valuable insights for future adaptive MoE system designs. Both the training code and model checkpoints will be publicly available.
☆ Advancing Academic Knowledge Retrieval via LLM-enhanced Representation Similarity Fusion KDD
In an era marked by robust technological growth and swift information renewal, furnishing researchers and the populace with top-tier, avant-garde academic insights spanning various domains has become an urgent necessity. The KDD Cup 2024 AQA Challenge is geared towards advancing retrieval models to identify pertinent academic terminologies from suitable papers for scientific inquiries. This paper introduces the LLM-KnowSimFuser proposed by Robo Space, which wins the 2nd place in the competition. With inspirations drawed from the superior performance of LLMs on multiple tasks, after careful analysis of the provided datasets, we firstly perform fine-tuning and inference using LLM-enhanced pre-trained retrieval models to introduce the tremendous language understanding and open-domain knowledge of LLMs into this task, followed by a weighted fusion based on the similarity matrix derived from the inference results. Finally, experiments conducted on the competition datasets show the superiority of our proposal, which achieved a score of 0.20726 on the final leaderboard.
comment: The 2nd Place of KDD Cup 2024 OAG-Challenge AQA
☆ KBLaM: Knowledge Base augmented Language Model
In this paper, we propose Knowledge Base augmented Language Model (KBLaM), a new method for augmenting Large Language Models (LLMs) with external knowledge. KBLaM works with a knowledge base (KB) constructed from a corpus of documents, transforming each piece of knowledge in the KB into continuous key-value vector pairs via pre-trained sentence encoders with linear adapters and integrating them into pre-trained LLMs via a specialized rectangular attention mechanism. Unlike Retrieval-Augmented Generation, KBLaM eliminates external retrieval modules, and unlike in-context learning, its computational overhead scales linearly with KB size rather than quadratically. Our approach enables integrating a large KB of more than 10K triples into an 8B pre-trained LLM of only 8K context window on one single A100 80GB GPU and allows for dynamic updates without model fine-tuning or retraining. Experiments demonstrate KBLaM's effectiveness in various tasks, including question-answering and open-ended reasoning, while providing interpretable insights into its use of the augmented knowledge.
☆ QUITE: Quantifying Uncertainty in Natural Language Text in Bayesian Reasoning Scenarios EMNLP 2024
Reasoning is key to many decision making processes. It requires consolidating a set of rule-like premises that are often associated with degrees of uncertainty and observations to draw conclusions. In this work, we address both the case where premises are specified as numeric probabilistic rules and situations in which humans state their estimates using words expressing degrees of certainty. Existing probabilistic reasoning datasets simplify the task, e.g., by requiring the model to only rank textual alternatives, by including only binary random variables, or by making use of a limited set of templates that result in less varied text. In this work, we present QUITE, a question answering dataset of real-world Bayesian reasoning scenarios with categorical random variables and complex relationships. QUITE provides high-quality natural language verbalizations of premises together with evidence statements and expects the answer to a question in the form of an estimated probability. We conduct an extensive set of experiments, finding that logic-based models outperform out-of-the-box large language models on all reasoning types (causal, evidential, and explaining-away). Our results provide evidence that neuro-symbolic models are a promising direction for improving complex reasoning. We release QUITE and code for training and experiments on Github.
comment: accepted at EMNLP 2024 (main)
☆ On Calibration of LLM-based Guard Models for Reliable Content Moderation
Large language models (LLMs) pose significant risks due to the potential for generating harmful content or users attempting to evade guardrails. Existing studies have developed LLM-based guard models designed to moderate the input and output of threat LLMs, ensuring adherence to safety policies by blocking content that violates these protocols upon deployment. However, limited attention has been given to the reliability and calibration of such guard models. In this work, we empirically conduct comprehensive investigations of confidence calibration for 9 existing LLM-based guard models on 12 benchmarks in both user input and model output classification. Our findings reveal that current LLM-based guard models tend to 1) produce overconfident predictions, 2) exhibit significant miscalibration when subjected to jailbreak attacks, and 3) demonstrate limited robustness to the outputs generated by different types of response models. Additionally, we assess the effectiveness of post-hoc calibration methods to mitigate miscalibration. We demonstrate the efficacy of temperature scaling and, for the first time, highlight the benefits of contextual calibration for confidence calibration of guard models, particularly in the absence of validation sets. Our analysis and experiments underscore the limitations of current LLM-based guard models and provide valuable insights for the future development of well-calibrated guard models toward more reliable content moderation. We also advocate for incorporating reliability evaluation of confidence calibration when releasing future LLM-based guard models.
comment: 19 pages, 9 figures
☆ Medico: Towards Hallucination Detection and Correction with Multi-source Evidence Fusion EMNLP 2024
As we all know, hallucinations prevail in Large Language Models (LLMs), where the generated content is coherent but factually incorrect, which inflicts a heavy blow on the widespread application of LLMs. Previous studies have shown that LLMs could confidently state non-existent facts rather than answering ``I don't know''. Therefore, it is necessary to resort to external knowledge to detect and correct the hallucinated content. Since manual detection and correction of factual errors is labor-intensive, developing an automatic end-to-end hallucination-checking approach is indeed a needful thing. To this end, we present Medico, a Multi-source evidence fusion enhanced hallucination detection and correction framework. It fuses diverse evidence from multiple sources, detects whether the generated content contains factual errors, provides the rationale behind the judgment, and iteratively revises the hallucinated content. Experimental results on evidence retrieval (0.964 HR@5, 0.908 MRR@5), hallucination detection (0.927-0.951 F1), and hallucination correction (0.973-0.979 approval rate) manifest the great potential of Medico. A video demo of Medico can be found at https://youtu.be/RtsO6CSesBI.
comment: 12 pages, 3 figures, 6 tables. Accepted by EMNLP 2024's demo track
☆ MMCFND: Multimodal Multilingual Caption-aware Fake News Detection for Low-resource Indic Languages
The widespread dissemination of false information through manipulative tactics that combine deceptive text and images threatens the integrity of reliable sources of information. While there has been research on detecting fake news in high resource languages using multimodal approaches, methods for low resource Indic languages primarily rely on textual analysis. This difference highlights the need for robust methods that specifically address multimodal fake news in Indic languages, where the lack of extensive datasets and tools presents a significant obstacle to progress. To this end, we introduce the Multimodal Multilingual dataset for Indic Fake News Detection (MMIFND). This meticulously curated dataset consists of 28,085 instances distributed across Hindi, Bengali, Marathi, Malayalam, Tamil, Gujarati and Punjabi. We further propose the Multimodal Multilingual Caption-aware framework for Fake News Detection (MMCFND). MMCFND utilizes pre-trained unimodal encoders and pairwise encoders from a foundational model that aligns vision and language, allowing for extracting deep representations from visual and textual components of news articles. The multimodal fusion encoder in the foundational model integrates text and image representations derived from its pairwise encoders to generate a comprehensive cross modal representation. Furthermore, we generate descriptive image captions that provide additional context to detect inconsistencies and manipulations. The retrieved features are then fused and fed into a classifier to determine the authenticity of news articles. The curated dataset can potentially accelerate research and development in low resource environments significantly. Thorough experimentation on MMIFND demonstrates that our proposed framework outperforms established methods for extracting relevant fake news detection features.
☆ Optimizing Instruction Synthesis: Effective Exploration of Evolutionary Space with Tree Search
Instruction tuning is a crucial technique for aligning language models with humans' actual goals in the real world. Extensive research has highlighted the quality of instruction data is essential for the success of this alignment. However, creating high-quality data manually is labor-intensive and time-consuming, which leads researchers to explore using LLMs to synthesize data. Recent studies have focused on using a stronger LLM to iteratively enhance existing instruction data, showing promising results. Nevertheless, previous work often lacks control over the evolution direction, resulting in high uncertainty in the data synthesis process and low-quality instructions. In this paper, we introduce a general and scalable framework, IDEA-MCTS (Instruction Data Enhancement using Monte Carlo Tree Search), a scalable framework for efficiently synthesizing instructions. With tree search and evaluation models, it can efficiently guide each instruction to evolve into a high-quality form, aiding in instruction fine-tuning. Experimental results show that IDEA-MCTS significantly enhances the seed instruction data, raising the average evaluation scores of quality, diversity, and complexity from 2.19 to 3.81. Furthermore, in open-domain benchmarks, experimental results show that IDEA-MCTS improves the accuracy of real-world instruction-following skills in LLMs by an average of 5\% in low-resource settings.
☆ BookWorm: A Dataset for Character Description and Analysis EMNLP 2024
Characters are at the heart of every story, driving the plot and engaging readers. In this study, we explore the understanding of characters in full-length books, which contain complex narratives and numerous interacting characters. We define two tasks: character description, which generates a brief factual profile, and character analysis, which offers an in-depth interpretation, including character development, personality, and social context. We introduce the BookWorm dataset, pairing books from the Gutenberg Project with human-written descriptions and analyses. Using this dataset, we evaluate state-of-the-art long-context models in zero-shot and fine-tuning settings, utilizing both retrieval-based and hierarchical processing for book-length inputs. Our findings show that retrieval-based approaches outperform hierarchical ones in both tasks. Additionally, fine-tuned models using coreference-based retrieval produce the most factual descriptions, as measured by fact- and entailment-based metrics. We hope our dataset, experiments, and analysis will inspire further research in character-based narrative understanding.
comment: 30 pages, 2 figures, EMNLP 2024 Findings
☆ Parenting: Optimizing Knowledge Selection of Retrieval-Augmented Language Models with Parameter Decoupling and Tailored Tuning
Retrieval-Augmented Generation (RAG) offers an effective solution to the issues faced by Large Language Models (LLMs) in hallucination generation and knowledge obsolescence by incorporating externally retrieved knowledge. However, due to potential conflicts between internal and external knowledge, as well as retrieval noise, LLMs often struggle to effectively integrate external evidence, leading to a decline in performance. Although existing methods attempt to tackle these challenges, they often struggle to strike a balance between model adherence and robustness, resulting in significant learning variance. Inspired by human cognitive processes, we propose Parenting, a novel framework that decouples adherence and robustness within the parameter space of LLMs. Specifically, Parenting utilizes a key parameter mining method based on forward activation gain to identify and isolate the crucial parameter units that are strongly linked to adherence and robustness. Then, Parenting employs a type-guided tailored tuning strategy, applying specific and appropriate fine-tuning methods to parameter units representing different capabilities, aiming to achieve a balanced enhancement of adherence and robustness. Extensive experiments on various datasets and models validate the effectiveness and generalizability of our methods.
☆ LLM-based Code-Switched Text Generation for Grammatical Error Correction
With the rise of globalisation, code-switching (CSW) has become a ubiquitous part of multilingual conversation, posing new challenges for natural language processing (NLP), especially in Grammatical Error Correction (GEC). This work explores the complexities of applying GEC systems to CSW texts. Our objectives include evaluating the performance of state-of-the-art GEC systems on an authentic CSW dataset from English as a Second Language (ESL) learners, exploring synthetic data generation as a solution to data scarcity, and developing a model capable of correcting grammatical errors in monolingual and CSW texts. We generated synthetic CSW GEC data, resulting in one of the first substantial datasets for this task, and showed that a model trained on this data is capable of significant improvements over existing systems. This work targets ESL learners, aiming to provide educational technologies that aid in the development of their English grammatical correctness without constraining their natural multilingualism.
☆ Augmenting In-Context-Learning in LLMs via Automatic Data Labeling and Refinement
It has been shown that Large Language Models' (LLMs) performance can be improved for many tasks using Chain of Thought (CoT) or In-Context Learning (ICL), which involve demonstrating the steps needed to solve a task using a few examples. However, while datasets with input-output pairs are relatively easy to produce, providing demonstrations which include intermediate steps requires cumbersome manual work. These steps may be executable programs, as in agentic flows, or step-by-step reasoning as in CoT. In this work, we propose Automatic Data Labeling and Refinement (ADLR), a method to automatically generate and filter demonstrations which include the above intermediate steps, starting from a small seed of manually crafted examples. We demonstrate the advantage of ADLR in code-based table QA and mathematical reasoning, achieving up to a 5.5% gain. The code implementing our method is provided in the Supplementary material and will be made available.
☆ A Unified Approach to Routing and Cascading for LLMs
The widespread applicability of large language models (LLMs) has increased the availability of many fine-tuned models of various sizes targeting specific tasks. Given a set of such specialized models, to maximize overall performance, it is important to figure out the optimal strategy for selecting the right model for a given user query. An effective strategy could drastically increase overall performance and even offer improvements over a single large monolithic model. Existing approaches typically fall into two categories: routing, where a single model is selected for each query, and cascading, which runs a sequence of increasingly larger models until a satisfactory answer is obtained. However, both have notable limitations: routing commits to an initial model without flexibility, while cascading requires executing every model in sequence, which can be inefficient. Additionally, the conditions under which these strategies are provably optimal remain unclear. In this work, we derive optimal strategies for both routing and cascading. Building on this analysis, we propose a novel approach called cascade routing, which combines the adaptability of routing with the cost-efficiency of cascading. Our experiments demonstrate that cascade routing consistently outperforms both routing and cascading across a variety of settings, improving both output quality and lowering computational cost, thus offering a unified and efficient solution to the model selection problem.
☆ Locking Down the Finetuned LLMs Safety
Fine-tuning large language models (LLMs) on additional datasets is often necessary to optimize them for specific downstream tasks. However, existing safety alignment measures, which restrict harmful behavior during inference, are insufficient to mitigate safety risks during fine-tuning. Alarmingly, fine-tuning with just 10 toxic sentences can make models comply with harmful instructions. We introduce SafetyLock, a novel alignment intervention method that maintains robust safety post-fine-tuning through efficient and transferable mechanisms. SafetyLock leverages our discovery that fine-tuned models retain similar safety-related activation representations to their base models. This insight enables us to extract what we term the Meta-SafetyLock, a set of safety bias directions representing key activation patterns associated with safe responses in the original model. We can then apply these directions universally to fine-tuned models to enhance their safety. By searching for activation directions across multiple token dimensions, SafetyLock achieves enhanced robustness and transferability. SafetyLock re-aligns fine-tuned models in under 0.01 seconds without additional computational cost. Our experiments demonstrate that SafetyLock can reduce the harmful instruction response rate from 60% to below 1% in toxic fine-tuned models. It surpasses traditional methods in both performance and efficiency, offering a scalable, non-invasive solution for ensuring the safety of customized LLMs. Our analysis across various fine-tuning scenarios confirms SafetyLock's robustness, advocating its integration into safety protocols for aligned LLMs. The code is released at https://github.com/zhu-minjun/SafetyLock.
☆ CoMAT: Chain of Mathematically Annotated Thought Improves Mathematical Reasoning
Mathematical reasoning remains a significant challenge for large language models (LLMs), despite progress in prompting techniques such as Chain-of-Thought (CoT). We present Chain of Mathematically Annotated Thought (CoMAT), which enhances reasoning through two stages: Symbolic Conversion (converting natural language queries into symbolic form) and Reasoning Execution (deriving answers from symbolic representations). CoMAT operates entirely with a single LLM and without external solvers. Across four LLMs, CoMAT outperforms traditional CoT on six out of seven benchmarks, achieving gains of 4.48% on MMLU-Redux (MATH) and 4.58% on GaoKao MCQ. In addition to improved performance, CoMAT ensures faithfulness and verifiability, offering a transparent reasoning process for complex mathematical tasks
comment: 8 pages, 12 figures
☆ Disentangling Hate Across Target Identities
Hate speech (HS) classifiers do not perform equally well in detecting hateful expressions towards different target identities. They also demonstrate systematic biases in predicted hatefulness scores. Tapping on two recently proposed functionality test datasets for HS detection, we quantitatively analyze the impact of different factors on HS prediction. Experiments on popular industrial and academic models demonstrate that HS detectors assign a higher hatefulness score merely based on the mention of specific target identities. Besides, models often confuse hatefulness and the polarity of emotions. This result is worrisome as the effort to build HS detectors might harm the vulnerable identity groups we wish to protect: posts expressing anger or disapproval of hate expressions might be flagged as hateful themselves. We also carry out a study inspired by social psychology theory, which reveals that the accuracy of hatefulness prediction correlates strongly with the intensity of the stereotype.
☆ MentalGLM Series: Explainable Large Language Models for Mental Health Analysis on Chinese Social Media
As the prevalence of mental health challenges, social media has emerged as a key platform for individuals to express their emotions.Deep learning tends to be a promising solution for analyzing mental health on social media. However, black box models are often inflexible when switching between tasks, and their results typically lack explanations. With the rise of large language models (LLMs), their flexibility has introduced new approaches to the field. Also due to the generative nature, they can be prompted to explain decision-making processes. However, their performance on complex psychological analysis still lags behind deep learning. In this paper, we introduce the first multi-task Chinese Social Media Interpretable Mental Health Instructions (C-IMHI) dataset, consisting of 9K samples, which has been quality-controlled and manually validated. We also propose MentalGLM series models, the first open-source LLMs designed for explainable mental health analysis targeting Chinese social media, trained on a corpus of 50K instructions. The proposed models were evaluated on three downstream tasks and achieved better or comparable performance compared to deep learning models, generalized LLMs, and task fine-tuned LLMs. We validated a portion of the generated decision explanations with experts, showing promising results. We also evaluated the proposed models on a clinical dataset, where they outperformed other LLMs, indicating their potential applicability in the clinical field. Our models show strong performance, validated across tasks and perspectives. The decision explanations enhance usability and facilitate better understanding and practical application of the models. Both the constructed dataset and the models are publicly available via: https://github.com/zwzzzQAQ/MentalGLM.
☆ EasyRAG: Efficient Retrieval-Augmented Generation Framework for Network Automated Operations
This paper presents EasyRAG, a simple, lightweight, and efficient retrieval-augmented generation framework for network automated operations. The advantages of our solution are: 1.Accurate Question Answering: We designed a straightforward RAG scheme based on (1) a specific data processing workflow (2) dual-route sparse retrieval for coarse ranking (3) LLM Reranker for reranking (4) LLM answer generation and optimization. This approach achieved first place in the GLM4 track in the preliminary round and second place in the GLM4 track in the semifinals. 2.Simple Deployment: Our method primarily consists of BM25 retrieval and BGE-reranker reranking, requiring no fine-tuning of any models, occupying minimal VRAM, easy to deploy, and highly scalable; we provide a flexible code library with various search and generation strategies, facilitating custom process implementation. 3.Efficient Inference: We designed an efficient inference acceleration scheme for the entire coarse ranking, reranking, and generation process that significantly reduces the inference latency of RAG while maintaining a good level of accuracy; each acceleration scheme can be plug-and-play into any component of the RAG process, consistently enhancing the efficiency of the RAG system. Our code and data are released at https://github.com/BUAADreamer/EasyRAG.
comment: 10 pages, 2 figures
☆ A Comparative Study of Translation Bias and Accuracy in Multilingual Large Language Models for Cross-Language Claim Verification NeurIPS 2024
The rise of digital misinformation has heightened interest in using multilingual Large Language Models (LLMs) for fact-checking. This study systematically evaluates translation bias and the effectiveness of LLMs for cross-lingual claim verification across 15 languages from five language families: Romance, Slavic, Turkic, Indo-Aryan, and Kartvelian. Using the XFACT dataset to assess their impact on accuracy and bias, we investigate two distinct translation methods: pre-translation and self-translation. We use mBERT's performance on the English dataset as a baseline to compare language-specific accuracies. Our findings reveal that low-resource languages exhibit significantly lower accuracy in direct inference due to underrepresentation in the training data. Furthermore, larger models demonstrate superior performance in self-translation, improving translation accuracy and reducing bias. These results highlight the need for balanced multilingual training, especially in low-resource languages, to promote equitable access to reliable fact-checking tools and minimize the risk of spreading misinformation in different linguistic contexts.
comment: Accepted to ATTRIB @ NeurIPS 2024
☆ FunnelRAG: A Coarse-to-Fine Progressive Retrieval Paradigm for RAG
Retrieval-Augmented Generation (RAG) prevails in Large Language Models. It mainly consists of retrieval and generation. The retrieval modules (a.k.a. retrievers) aim to find useful information used to facilitate generation modules (a.k.a. generators). As such, generators' performance largely depends on the effectiveness and efficiency of retrievers. However, the retrieval paradigm that we design and use remains flat, which treats the retrieval procedures as a one-off deal with constant granularity. Despite effectiveness, we argue that they suffer from two limitations: (1) flat retrieval exerts a significant burden on one retriever; (2) constant granularity limits the ceiling of retrieval performance. In this work, we propose a progressive retrieval paradigm with coarse-to-fine granularity for RAG, termed FunnelRAG, so as to balance effectiveness and efficiency. Specifically, FunnelRAG establishes a progressive retrieval pipeline by collaborating coarse-to-fine granularity, large-to-small quantity, and low-to-high capacity, which can relieve the burden on one retriever and also promote the ceiling of retrieval performance. Extensive experiments manifest that FunnelRAG achieves comparable retrieval performance while the time overhead is reduced by nearly 40 percent.
comment: 18 pages, 6 figures, 13 tables
☆ Evaluating Semantic Variation in Text-to-Image Synthesis: A Causal Perspective
Accurate interpretation and visualization of human instructions are crucial for text-to-image (T2I) synthesis. However, current models struggle to capture semantic variations from word order changes, and existing evaluations, relying on indirect metrics like text-image similarity, fail to reliably assess these challenges. This often obscures poor performance on complex or uncommon linguistic patterns by the focus on frequent word combinations. To address these deficiencies, we propose a novel metric called SemVarEffect and a benchmark named SemVarBench, designed to evaluate the causality between semantic variations in inputs and outputs in T2I synthesis. Semantic variations are achieved through two types of linguistic permutations, while avoiding easily predictable literal variations. Experiments reveal that the CogView-3-Plus and Ideogram 2 performed the best, achieving a score of 0.2/1. Semantic variations in object relations are less understood than attributes, scoring 0.07/1 compared to 0.17-0.19/1. We found that cross-modal alignment in UNet or Transformers plays a crucial role in handling semantic variations, a factor previously overlooked by a focus on textual encoders. Our work establishes an effective evaluation framework that advances the T2I synthesis community's exploration of human instruction understanding.
comment: Our benchmark and code are available at https://github.com/zhuxiangru/SemVarBench
☆ A Multi-Task Text Classification Pipeline with Natural Language Explanations: A User-Centric Evaluation in Sentiment Analysis and Offensive Language Identification in Greek Tweets
Interpretability is a topic that has been in the spotlight for the past few years. Most existing interpretability techniques produce interpretations in the form of rules or feature importance. These interpretations, while informative, may be harder to understand for non-expert users and therefore, cannot always be considered as adequate explanations. To that end, explanations in natural language are often preferred, as they are easier to comprehend and also more presentable to end-users. This work introduces an early concept for a novel pipeline that can be used in text classification tasks, offering predictions and explanations in natural language. It comprises of two models: a classifier for labelling the text and an explanation generator which provides the explanation. The proposed pipeline can be adopted by any text classification task, given that ground truth rationales are available to train the explanation generator. Our experiments are centred around the tasks of sentiment analysis and offensive language identification in Greek tweets, using a Greek Large Language Model (LLM) to obtain the necessary explanations that can act as rationales. The experimental evaluation was performed through a user study based on three different metrics and achieved promising results for both datasets.
comment: Work In Progress
☆ Back-of-the-Book Index Automation for Arabic Documents
Back-of-the-book indexes are crucial for book readability. Their manual creation is laborious and error prone. In this paper, we consider automating back-of-the-book index extraction for Arabic books to help simplify both the creation and review tasks. Given a back-of-the-book index, we aim to check and identify the accurate occurrences of index terms relative to the associated pages. To achieve this, we first define a pool of candidates for each term by extracting all possible noun phrases from paragraphs appearing on the relevant index pages. These noun phrases, identified through part-of-speech analysis, are stored in a vector database for efficient retrieval. We use several metrics, including exact matches, lexical similarity, and semantic similarity, to determine the most appropriate occurrence. The candidate with the highest score based on these metrics is chosen as the occurrence of the term. We fine-tuned a heuristic method, that considers the above metrics and that achieves an F1-score of .966 (precision=.966, recall=.966). These excellent results open the door for future work related to automation of back-of-the-book index generation and checking.
☆ Machine Translation Evaluation Benchmark for Wu Chinese: Workflow and Analysis EMNLP
We introduce a FLORES+ dataset as an evaluation benchmark for modern Wu Chinese machine translation models and showcase its compatibility with existing Wu data. Wu Chinese is mutually unintelligible with other Sinitic languages such as Mandarin and Yue (Cantonese), but uses a set of Hanzi (Chinese characters) that profoundly overlaps with others. The population of Wu speakers is the second largest among languages in China, but the language has been suffering from significant drop in usage especially among the younger generations. We identify Wu Chinese as a textually low-resource language and address challenges for its machine translation models. Our contributions include: (1) an open-source, manually translated dataset, (2) full documentations on the process of dataset creation and validation experiments, (3) preliminary tools for Wu Chinese normalization and segmentation, and (4) benefits and limitations of our dataset, as well as implications to other low-resource languages.
comment: EMNLP WMT 24 Open Language Data Initiative Shared Task
☆ QUIS: Question-guided Insights Generation for Automated Exploratory Data Analysis
Discovering meaningful insights from a large dataset, known as Exploratory Data Analysis (EDA), is a challenging task that requires thorough exploration and analysis of the data. Automated Data Exploration (ADE) systems use goal-oriented methods with Large Language Models and Reinforcement Learning towards full automation. However, these methods require human involvement to anticipate goals that may limit insight extraction, while fully automated systems demand significant computational resources and retraining for new datasets. We introduce QUIS, a fully automated EDA system that operates in two stages: insight generation (ISGen) driven by question generation (QUGen). The QUGen module generates questions in iterations, refining them from previous iterations to enhance coverage without human intervention or manually curated examples. The ISGen module analyzes data to produce multiple relevant insights in response to each question, requiring no prior training and enabling QUIS to adapt to new datasets.
comment: 6 pages
☆ LoLCATs: On Low-Rank Linearizing of Large Language Models
Recent works show we can linearize large language models (LLMs) -- swapping the quadratic attentions of popular Transformer-based LLMs with subquadratic analogs, such as linear attention -- avoiding the expensive pretraining costs. However, linearizing LLMs often significantly degrades model quality, still requires training over billions of tokens, and remains limited to smaller 1.3B to 7B LLMs. We thus propose Low-rank Linear Conversion via Attention Transfer (LoLCATs), a simple two-step method that improves LLM linearizing quality with orders of magnitudes less memory and compute. We base these steps on two findings. First, we can replace an LLM's softmax attentions with closely-approximating linear attentions, simply by training the linear attentions to match their softmax counterparts with an output MSE loss ("attention transfer"). Then, this enables adjusting for approximation errors and recovering LLM quality simply with low-rank adaptation (LoRA). LoLCATs significantly improves linearizing quality, training efficiency, and scalability. We significantly reduce the linearizing quality gap and produce state-of-the-art subquadratic LLMs from Llama 3 8B and Mistral 7B v0.1, leading to 20+ points of improvement on 5-shot MMLU. Furthermore, LoLCATs does so with only 0.2% of past methods' model parameters and 0.4% of their training tokens. Finally, we apply LoLCATs to create the first linearized 70B and 405B LLMs (50x larger than prior work). When compared with prior approaches under the same compute budgets, LoLCATs significantly improves linearizing quality, closing the gap between linearized and original Llama 3.1 70B and 405B LLMs by 77.8% and 78.1% on 5-shot MMLU.
comment: 47 pages, 20 figures, 18 tables, preprint
☆ BanglaQuAD: A Bengali Open-domain Question Answering Dataset LREC
Bengali is the seventh most spoken language on earth, yet considered a low-resource language in the field of natural language processing (NLP). Question answering over unstructured text is a challenging NLP task as it requires understanding both question and passage. Very few researchers attempted to perform question answering over Bengali (natively pronounced as Bangla) text. Typically, existing approaches construct the dataset by directly translating them from English to Bengali, which produces noisy and improper sentence structures. Furthermore, they lack topics and terminologies related to the Bengali language and people. This paper introduces BanglaQuAD, a Bengali question answering dataset, containing 30,808 question-answer pairs constructed from Bengali Wikipedia articles by native speakers. Additionally, we propose an annotation tool that facilitates question-answering dataset construction on a local machine. A qualitative analysis demonstrates the quality of our proposed dataset.
comment: Accepted into LREC-COLING 2024, Turin, Italy
☆ QE-EBM: Using Quality Estimators as Energy Loss for Machine Translation
Reinforcement learning has shown great promise in aligning language models with human preferences in a variety of text generation tasks, including machine translation. For translation tasks, rewards can easily be obtained from quality estimation (QE) models which can generate rewards for unlabeled data. Despite its usefulness, reinforcement learning cannot exploit the gradients with respect to the QE score. We propose QE-EBM, a method of employing quality estimators as trainable loss networks that can directly backpropagate to the NMT model. We examine our method on several low and high resource target languages with English as the source language. QE-EBM outperforms strong baselines such as REINFORCE and proximal policy optimization (PPO) as well as supervised fine-tuning for all target languages, especially low-resource target languages. Most notably, for English-to-Mongolian translation, our method achieves improvements of 2.5 BLEU, 7.1 COMET-KIWI, 5.3 COMET, and 6.4 XCOMET relative to the supervised baseline.
☆ ChakmaNMT: A Low-resource Machine Translation On Chakma Language ACL
The geopolitical division between the indigenous Chakma population and mainstream Bangladesh creates a significant cultural and linguistic gap, as the Chakma community, mostly residing in the hill tracts of Bangladesh, maintains distinct cultural traditions and language. Developing a Machine Translation (MT) model or Chakma to Bangla could play a crucial role in alleviating this cultural-linguistic divide. Thus, we have worked on MT between CCP-BN(Chakma-Bangla) by introducing a novel dataset of 15,021 parallel samples and 42,783 monolingual samples of the Chakma Language. Moreover, we introduce a small set for Benchmarking containing 600 parallel samples between Chakma, Bangla, and English. We ran traditional and state-of-the-art models in NLP on the training set, where fine-tuning BanglaT5 with back-translation using transliteration of Chakma achieved the highest BLEU score of 17.8 and 4.41 in CCP-BN and BN-CCP respectively on the Benchmark Dataset. As far as we know, this is the first-ever work on MT for the Chakma Language. Hopefully, this research will help to bridge the gap in linguistic resources and contribute to preserving endangered languages. Our dataset link and codes will be published soon.
comment: to be submitted in ACL findings 2025
☆ SkillAggregation: Reference-free LLM-Dependent Aggregation
Large Language Models (LLMs) are increasingly used to assess NLP tasks due to their ability to generate human-like judgments. Single LLMs were used initially, however, recent work suggests using multiple LLMs as judges yields improved performance. An important step in exploiting multiple judgements is the combination stage, aggregation. Existing methods in NLP either assign equal weight to all LLM judgments or are designed for specific tasks such as hallucination detection. This work focuses on aggregating predictions from multiple systems where no reference labels are available. A new method called SkillAggregation is proposed, which learns to combine estimates from LLM judges without needing additional data or ground truth. It extends the Crowdlayer aggregation method, developed for image classification, to exploit the judge estimates during inference. The approach is compared to a range of standard aggregation methods on HaluEval-Dialogue, TruthfulQA and Chatbot Arena tasks. SkillAggregation outperforms Crowdlayer on all tasks, and yields the best performance over all approaches on the majority of tasks.
☆ Minimum Tuning to Unlock Long Output from LLMs with High Quality Data as the Key
As large language models rapidly evolve to support longer context, there is a notable disparity in their capability to generate output at greater lengths. Recent study suggests that the primary cause for this imbalance may arise from the lack of data with long-output during alignment training. In light of this observation, attempts are made to re-align foundation models with data that fills the gap, which result in models capable of generating lengthy output when instructed. In this paper, we explore the impact of data-quality in tuning a model for long output, and the possibility of doing so from the starting points of human-aligned (instruct or chat) models. With careful data curation, we show that it possible to achieve similar performance improvement in our tuned models, with only a small fraction of training data instances and compute. In addition, we assess the generalizability of such approaches by applying our tuning-recipes to several models. our findings suggest that, while capacities for generating long output vary across different models out-of-the-box, our approach to tune them with high-quality data using lite compute, consistently yields notable improvement across all models we experimented on. We have made public our curated dataset for tuning long-writing capability, the implementations of model tuning and evaluation, as well as the fine-tuned models, all of which can be openly-accessed.
☆ Effi-Code: Unleashing Code Efficiency in Language Models
As the use of large language models (LLMs) for code generation becomes more prevalent in software development, it is critical to enhance both the efficiency and correctness of the generated code. Existing methods and models primarily focus on the correctness of LLM-generated code, ignoring efficiency. In this work, we present Effi-Code, an approach to enhancing code generation in LLMs that can improve both efficiency and correctness. We introduce a Self-Optimization process based on Overhead Profiling that leverages open-source LLMs to generate a high-quality dataset of correct and efficient code samples. This dataset is then used to fine-tune various LLMs. Our method involves the iterative refinement of generated code, guided by runtime performance metrics and correctness checks. Extensive experiments demonstrate that models fine-tuned on the Effi-Code show significant improvements in both code correctness and efficiency across task types. For example, the pass@1 of DeepSeek-Coder-6.7B-Instruct generated code increases from \textbf{43.3\%} to \textbf{76.8\%}, and the average execution time for the same correct tasks decreases by \textbf{30.5\%}. Effi-Code offers a scalable and generalizable approach to improving code generation in AI systems, with potential applications in software development, algorithm design, and computational problem-solving. The source code of Effi-Code was released in \url{https://github.com/huangd1999/Effi-Code}.
comment: Under Review
☆ Scalable Multi-Domain Adaptation of Language Models using Modular Experts
Domain-specific adaptation is critical to maximizing the performance of pre-trained language models (PLMs) on one or multiple targeted tasks, especially under resource-constrained use cases, such as edge devices. However, existing methods often struggle to balance domain-specific performance, retention of general knowledge, and efficiency for training and inference. To address these challenges, we propose Modular Domain Experts (MoDE). MoDE is a mixture-of-experts architecture that augments a general PLMs with modular, domain-specialized experts. These experts are trained independently and composed together via a lightweight training process. In contrast to standard low-rank adaptation methods, each MoDE expert consists of several transformer layers which scale better with more training examples and larger parameter counts. Our evaluation demonstrates that MoDE achieves comparable target performances to full parameter fine-tuning while achieving 1.65% better retention performance. Moreover, MoDE's architecture enables flexible sharding configurations and improves training speeds by up to 38% over state-of-the-art distributed training configurations.
comment: 14 pages, 5 figures, 3 tables
☆ Is Parameter Collision Hindering Continual Learning in LLMs?
Large Language Models (LLMs) often suffer from catastrophic forgetting when learning multiple tasks sequentially, making continual learning (CL) essential for their dynamic deployment. Existing state-of-the-art (SOTA) methods, such as O-LoRA, typically focus on constructing orthogonality tasks to decouple parameter interdependence from various domains.In this paper, we reveal that building non-collision parameters is a more critical factor in addressing CL challenges. Our theoretical and experimental analyses demonstrate that non-collision parameters can provide better task orthogonality, which is a sufficient but unnecessary condition. Furthermore, knowledge from multiple domains will be preserved in non-collision parameter subspaces, making it more difficult to forget previously seen data. Leveraging this insight, we propose Non-collision Low-Rank Adaptation (N-LoRA), a simple yet effective approach leveraging low collision rates to enhance CL in LLMs. Experimental results on multiple CL benchmarks indicate that N-LoRA achieves superior performance (+2.9), higher task orthogonality (*4.1 times), and lower parameter collision (*58.1 times) than SOTA methods.
☆ HSR-Enhanced Sparse Attention Acceleration
Large Language Models (LLMs) have demonstrated remarkable capabilities across various applications, but their performance on long-context tasks is often limited by the computational complexity of attention mechanisms. This paper introduces a novel approach to accelerate attention computation in LLMs, particularly for long-context scenarios. We leverage the inherent sparsity within attention mechanisms, both in conventional Softmax attention and ReLU attention (with $\mathsf{ReLU}^\alpha$ activation, $\alpha \in \mathbb{N}_+$), to significantly reduce the running time complexity. Our method employs a Half-Space Reporting (HSR) data structure to rapidly identify non-zero or "massively activated" entries in the attention matrix. We present theoretical analyses for two key scenarios: attention generation and full attention computation with long input context. Our approach achieves a running time of $O(mn^{4/5})$ significantly faster than the naive approach $O(mn)$ for attention generation, where $n$ is the context length, $m$ is the query length, and $d$ is the hidden dimension. We can also reduce the running time of full attention computation from $O(mn)$ to $O(mn^{1 - 1 / \lfloor d/2\rfloor} + mn^{4/5})$. Importantly, our method introduces no error for ReLU attention and only provably negligible error for Softmax attention, where the latter is supported by our empirical validation. This work represents a significant step towards enabling efficient long-context processing in LLMs, potentially broadening their applicability across various domains.
☆ Diagnosing Hate Speech Classification: Where Do Humans and Machines Disagree, and Why?
This study uses the cosine similarity ratio, embedding regression, and manual re-annotation to diagnose hate speech classification. We begin by computing cosine similarity ratio on a dataset "Measuring Hate Speech" that contains 135,556 annotated comments on social media. This way, we show a basic use of cosine similarity as a description of hate speech content. We then diagnose hate speech classification starting from understanding the inconsistency of human annotation from the dataset. Using embedding regression as a basic diagnostic, we found that female annotators are more sensitive to racial slurs that target the black population. We perform with a more complicated diagnostic by training a hate speech classifier using a SoTA pre-trained large language model, NV-Embed-v2, to convert texts to embeddings and run a logistic regression. This classifier achieves a testing accuracy of 94%. In diagnosing where machines disagree with human annotators, we found that machines make fewer mistakes than humans despite the fact that human annotations are treated as ground truth in the training set. Machines perform better in correctly labeling long statements of facts, but perform worse in labeling short instances of swear words. We hypothesize that this is due to model alignment - while curating models at their creation prevents the models from producing obvious hate speech, it also reduces the model's ability to detect such content.
☆ Jailbreak Instruction-Tuned LLMs via end-of-sentence MLP Re-weighting
In this paper, we investigate the safety mechanisms of instruction fine-tuned large language models (LLMs). We discover that re-weighting MLP neurons can significantly compromise a model's safety, especially for MLPs in end-of-sentence inferences. We hypothesize that LLMs evaluate the harmfulness of prompts during end-of-sentence inferences, and MLP layers plays a critical role in this process. Based on this hypothesis, we develop 2 novel white-box jailbreak methods: a prompt-specific method and a prompt-general method. The prompt-specific method targets individual prompts and optimizes the attack on the fly, while the prompt-general method is pre-trained offline and can generalize to unseen harmful prompts. Our methods demonstrate robust performance across 7 popular open-source LLMs, size ranging from 2B to 72B. Furthermore, our study provides insights into vulnerabilities of instruction-tuned LLM's safety and deepens the understanding of the internal mechanisms of LLMs.
☆ $α$-DPO: Adaptive Reward Margin is What Direct Preference Optimization Needs
Aligning large language models (LLMs) with human values and intentions is crucial for their utility, honesty, and safety. Reinforcement learning from human feedback (RLHF) is a popular approach to achieve this alignment, but it faces challenges in computational efficiency and training stability. Recent methods like Direct Preference Optimization (DPO) and Simple Preference Optimization (SimPO) have proposed offline alternatives to RLHF, simplifying the process by reparameterizing the reward function. However, DPO depends on a potentially suboptimal reference model, and SimPO's assumption of a fixed target reward margin may lead to suboptimal decisions in diverse data settings. In this work, we propose $\alpha$-DPO, an adaptive preference optimization algorithm designed to address these limitations by introducing a dynamic reward margin. Specifically, $\alpha$-DPO employs an adaptive preference distribution, balancing the policy model and the reference model to achieve personalized reward margins. We provide theoretical guarantees for $\alpha$-DPO, demonstrating its effectiveness as a surrogate optimization objective and its ability to balance alignment and diversity through KL divergence control. Empirical evaluations on AlpacaEval 2 and Arena-Hard show that $\alpha$-DPO consistently outperforms DPO and SimPO across various model settings, establishing it as a robust approach for fine-tuning LLMs. Our method achieves significant improvements in win rates, highlighting its potential as a powerful tool for LLM alignment. The code is available at https://github.com/junkangwu/alpha-DPO
☆ Unified Representation of Genomic and Biomedical Concepts through Multi-Task, Multi-Source Contrastive Learning
We introduce GENomic Encoding REpresentation with Language Model (GENEREL), a framework designed to bridge genetic and biomedical knowledge bases. What sets GENEREL apart is its ability to fine-tune language models to infuse biological knowledge behind clinical concepts such as diseases and medications. This fine-tuning enables the model to capture complex biomedical relationships more effectively, enriching the understanding of how genomic data connects to clinical outcomes. By constructing a unified embedding space for biomedical concepts and a wide range of common SNPs from sources such as patient-level data, biomedical knowledge graphs, and GWAS summaries, GENEREL aligns the embeddings of SNPs and clinical concepts through multi-task contrastive learning. This allows the model to adapt to diverse natural language representations of biomedical concepts while bypassing the limitations of traditional code mapping systems across different data sources. Our experiments demonstrate GENEREL's ability to effectively capture the nuanced relationships between SNPs and clinical concepts. GENEREL also emerges to discern the degree of relatedness, potentially allowing for a more refined identification of concepts. This pioneering approach in constructing a unified embedding system for both SNPs and biomedical concepts enhances the potential for data integration and discovery in biomedical research.
comment: 15 pages, 2 figures, 5 tables
☆ Temperature-Centric Investigation of Speculative Decoding with Knowledge Distillation EMNLP 2024
Speculative decoding stands as a pivotal technique to expedite inference in autoregressive (large) language models. This method employs a smaller draft model to speculate a block of tokens, which the target model then evaluates for acceptance. Despite a wealth of studies aimed at increasing the efficiency of speculative decoding, the influence of generation configurations on the decoding process remains poorly understood, especially concerning decoding temperatures. This paper delves into the effects of decoding temperatures on speculative decoding's efficacy. Beginning with knowledge distillation (KD), we first highlight the challenge of decoding at higher temperatures, and demonstrate KD in a consistent temperature setting could be a remedy. We also investigate the effects of out-of-domain testing sets with out-of-range temperatures. Building upon these findings, we take an initial step to further the speedup for speculative decoding, particularly in a high-temperature generation setting. Our work offers new insights into how generation configurations drastically affect the performance of speculative decoding, and underscores the need for developing methods that focus on diverse decoding configurations. Code is publically available at https://github.com/ozyyshr/TempSpec.
comment: EMNLP 2024 Findings
☆ MMIE: Massive Multimodal Interleaved Comprehension Benchmark for Large Vision-Language Models
Interleaved multimodal comprehension and generation, enabling models to produce and interpret both images and text in arbitrary sequences, have become a pivotal area in multimodal learning. Despite significant advancements, the evaluation of this capability remains insufficient. Existing benchmarks suffer from limitations in data scale, scope, and evaluation depth, while current evaluation metrics are often costly or biased, lacking in reliability for practical applications. To address these challenges, we introduce MMIE, a large-scale knowledge-intensive benchmark for evaluating interleaved multimodal comprehension and generation in Large Vision-Language Models (LVLMs). MMIE comprises 20K meticulously curated multimodal queries, spanning 3 categories, 12 fields, and 102 subfields, including mathematics, coding, physics, literature, health, and arts. It supports both interleaved inputs and outputs, offering a mix of multiple-choice and open-ended question formats to evaluate diverse competencies. Moreover, we propose a reliable automated evaluation metric, leveraging a scoring model fine-tuned with human-annotated data and systematic evaluation criteria, aimed at reducing bias and improving evaluation accuracy. Extensive experiments demonstrate the effectiveness of our benchmark and metrics in providing a comprehensive evaluation of interleaved LVLMs. Specifically, we evaluate eight LVLMs, revealing that even the best models show significant room for improvement, with most achieving only moderate results. We believe MMIE will drive further advancements in the development of interleaved LVLMs. We publicly release our benchmark and code in https://mmie-bench.github.io/.
☆ Beyond-RAG: Question Identification and Answer Generation in Real-Time Conversations
In customer contact centers, human agents often struggle with long average handling times (AHT) due to the need to manually interpret queries and retrieve relevant knowledge base (KB) articles. While retrieval augmented generation (RAG) systems using large language models (LLMs) have been widely adopted in industry to assist with such tasks, RAG faces challenges in real-time conversations, such as inaccurate query formulation and redundant retrieval of frequently asked questions (FAQs). To address these limitations, we propose a decision support system that can look beyond RAG by first identifying customer questions in real time. If the query matches an FAQ, the system retrieves the answer directly from the FAQ database; otherwise, it generates answers via RAG. Our approach reduces reliance on manual queries, providing responses to agents within 2 seconds. Deployed in AI-powered human-agent assist solution at Minerva CQ, this system improves efficiency, reduces AHT, and lowers operational costs. We also introduce an automated LLM-agentic workflow to identify FAQs from historical transcripts when no predefined FAQs exist.
☆ FormalAlign: Automated Alignment Evaluation for Autoformalization
Autoformalization aims to convert informal mathematical proofs into machine-verifiable formats, bridging the gap between natural and formal languages. However, ensuring semantic alignment between the informal and formalized statements remains challenging. Existing approaches heavily rely on manual verification, hindering scalability. To address this, we introduce \textsc{FormalAlign}, the first automated framework designed for evaluating the alignment between natural and formal languages in autoformalization. \textsc{FormalAlign} trains on both the autoformalization sequence generation task and the representational alignment between input and output, employing a dual loss that combines a pair of mutually enhancing autoformalization and alignment tasks. Evaluated across four benchmarks augmented by our proposed misalignment strategies, \textsc{FormalAlign} demonstrates superior performance. In our experiments, \textsc{FormalAlign} outperforms GPT-4, achieving an Alignment-Selection Score 11.58\% higher on \forml-Basic (99.21\% vs. 88.91\%) and 3.19\% higher on MiniF2F-Valid (66.39\% vs. 64.34\%). This effective alignment evaluation significantly reduces the need for manual verification. Both the dataset and code can be accessed via~\url{https://github.com/rookie-joe/FormalAlign}.
comment: 23 pages, 13 tables, 3 figures
☆ How to Construct Random Unitaries
The existence of pseudorandom unitaries (PRUs) -- efficient quantum circuits that are computationally indistinguishable from Haar-random unitaries -- has been a central open question, with significant implications for cryptography, complexity theory, and fundamental physics. In this work, we close this question by proving that PRUs exist, assuming that any quantum-secure one-way function exists. We establish this result for both (1) the standard notion of PRUs, which are secure against any efficient adversary that makes queries to the unitary $U$, and (2) a stronger notion of PRUs, which are secure even against adversaries that can query both the unitary $U$ and its inverse $U^\dagger$. In the process, we prove that any algorithm that makes queries to a Haar-random unitary can be efficiently simulated on a quantum computer, up to inverse-exponential trace distance.
comment: 76 pages
☆ Mixture of Experts Made Personalized: Federated Prompt Learning for Vision-Language Models
Prompt learning for pre-trained Vision-Language Models (VLMs) like CLIP has demonstrated potent applicability across diverse downstream tasks. This lightweight approach has quickly gained traction from federated learning (FL) researchers who seek to efficiently adapt VLMs to heterogeneous scenarios. However, current federated prompt learning methods are habitually restricted to the traditional FL paradigm, where the participating clients are generally only allowed to download a single globally aggregated model from the server. While justifiable for training full-sized models under federated settings, in this work, we argue that this paradigm is ill-suited for lightweight prompts. By facilitating the clients to download multiple pre-aggregated prompts as fixed non-local experts, we propose Personalized Federated Mixture of Adaptive Prompts (pFedMoAP), a novel FL framework that personalizes the prompt learning process through the lens of Mixture of Experts (MoE). pFedMoAP implements a local attention-based gating network that learns to generate enhanced text features for better alignment with local image data on the client, benefiting from both local and downloaded non-local adaptive prompt experts. The non-local experts are sparsely selected from a server-maintained pool, fostering collaborative learning across clients. To evaluate the proposed algorithm, we conduct extensive experiments across 9 datasets under various heterogeneous federated settings. The results show that pFedMoAP consistently outperforms the state-of-the-art alternatives, underscoring its efficacy in personalizing prompt learning for CLIP within the federated learning paradigm.
comment: 16 pages, 4 figures
☆ Can We Predict Performance of Large Models across Vision-Language Tasks?
Evaluating large vision-language models (LVLMs) is very expensive, due to the high computational costs and the wide variety of tasks. The good news is that if we already have some observed performance scores, we may be able to infer unknown ones. In this study, we propose a new framework for predicting unknown performance scores based on observed ones from other LVLMs or tasks. We first formulate the performance prediction as a matrix completion task. Specifically, we construct a sparse performance matrix $\boldsymbol{R}$, where each entry $R_{mn}$ represents the performance score of the $m$-th model on the $n$-th dataset. By applying probabilistic matrix factorization (PMF) with Markov chain Monte Carlo (MCMC), we can complete the performance matrix, that is, predict unknown scores. Additionally, we estimate the uncertainty of performance prediction based on MCMC. Practitioners can evaluate their models on untested tasks with higher uncertainty first, quickly reducing errors in performance prediction. We further introduce several improvements to enhance PMF for scenarios with sparse observed performance scores. In experiments, we systematically evaluate 108 LVLMs on 176 datasets from 36 benchmarks, constructing training and testing sets for validating our framework. Our experiments demonstrate the accuracy of PMF in predicting unknown scores, the reliability of uncertainty estimates in ordering evaluations, and the effectiveness of our enhancements for handling sparse data.
comment: Under Review. Project page: https://github.com/Qinyu-Allen-Zhao/CrossPred-LVLM
☆ Learning Linear Attention in Polynomial Time
Previous research has explored the computational expressivity of Transformer models in simulating Boolean circuits or Turing machines. However, the learnability of these simulators from observational data has remained an open question. Our study addresses this gap by providing the first polynomial-time learnability results (specifically strong, agnostic PAC learning) for single-layer Transformers with linear attention. We show that linear attention may be viewed as a linear predictor in a suitably defined RKHS. As a consequence, the problem of learning any linear transformer may be converted into the problem of learning an ordinary linear predictor in an expanded feature space, and any such predictor may be converted back into a multiheaded linear transformer. Moving to generalization, we show how to efficiently identify training datasets for which every empirical risk minimizer is equivalent (up to trivial symmetries) to the linear Transformer that generated the data, thereby guaranteeing the learned model will correctly generalize across all inputs. Finally, we provide examples of computations expressible via linear attention and therefore polynomial-time learnable, including associative memories, finite automata, and a class of Universal Turing Machine (UTMs) with polynomially bounded computation histories. We empirically validate our theoretical findings on three tasks: learning random linear attention networks, key--value associations, and learning to execute finite automata. Our findings bridge a critical gap between theoretical expressivity and learnability of Transformers, and show that flexible and general models of computation are efficiently learnable.
☆ How to Leverage Demonstration Data in Alignment for Large Language Model? A Self-Imitation Learning Perspective EMNLP 2024
This paper introduces a novel generalized self-imitation learning ($\textbf{GSIL}$) framework, which effectively and efficiently aligns large language models with offline demonstration data. We develop $\textbf{GSIL}$ by deriving a surrogate objective of imitation learning with density ratio estimates, facilitating the use of self-generated data and optimizing the imitation learning objective with simple classification losses. $\textbf{GSIL}$ eliminates the need for complex adversarial training in standard imitation learning, achieving lightweight and efficient fine-tuning for large language models. In addition, $\textbf{GSIL}$ encompasses a family of offline losses parameterized by a general class of convex functions for density ratio estimation and enables a unified view for alignment with demonstration data. Extensive experiments show that $\textbf{GSIL}$ consistently and significantly outperforms baselines in many challenging benchmarks, such as coding (HuamnEval), mathematical reasoning (GSM8K) and instruction-following benchmark (MT-Bench).
comment: EMNLP 2024 Main
☆ RoCoFT: Efficient Finetuning of Large Language Models with Row-Column Updates
We propose RoCoFT, a parameter-efficient fine-tuning method for large-scale language models (LMs) based on updating only a few rows and columns of the weight matrices in transformers. Through extensive experiments with medium-size LMs like BERT and RoBERTa, and larger LMs like Bloom-7B, Llama2-7B, and Llama2-13B, we show that our method gives comparable or better accuracies than state-of-art PEFT methods while also being more memory and computation-efficient. We also study the reason behind the effectiveness of our method with tools from neural tangent kernel theory. We empirically demonstrate that our kernel, constructed using a restricted set of row and column parameters, are numerically close to the full-parameter kernel and gives comparable classification performance. Ablation studies are conducted to investigate the impact of different algorithmic choices, including the selection strategy for rows and columns as well as the optimal rank for effective implementation of our method.
comment: RoCoFT is a parameter-efficient method
☆ Divide, Reweight, and Conquer: A Logit Arithmetic Approach for In-Context Learning
In-Context Learning (ICL) emerges as a key feature for Large Language Models (LLMs), allowing them to adapt to new tasks by leveraging task-specific examples without updating model parameters. However, ICL faces challenges with increasing numbers of examples due to performance degradation and quadratic computational costs. In this paper, we propose Logit Arithmetic Reweighting Approach (LARA), a novel framework that enhances ICL by using logit-based ensembling of multiple demonstrations. Our approach divides long input demonstrations into parallelizable shorter inputs to significantly reduce memory requirements, and then effectively aggregate the information by reweighting logits of each group via a non-gradient optimization approach. We further introduce Binary LARA (B-LARA), a variant that constrains weights to binary values to simplify the search space and reduces memory usage by filtering out less informative demonstration groups. Experiments on BBH and MMLU demonstrate that LARA and B-LARA outperform all baseline methods in both accuracy and memory efficiency. We also conduct extensive analysis to show that LARA generalizes well to scenarios of varying numbers of examples from limited to many-shot demonstrations.
☆ Ukrainian-to-English folktale corpus: Parallel corpus creation and augmentation for machine translation in low-resource languages
Folktales are linguistically very rich and culturally significant in understanding the source language. Historically, only human translation has been used for translating folklore. Therefore, the number of translated texts is very sparse, which limits access to knowledge about cultural traditions and customs. We have created a new Ukrainian-To-English parallel corpus of familiar Ukrainian folktales based on available English translations and suggested several new ones. We offer a combined domain-specific approach to building and augmenting this corpus, considering the nature of the domain and differences in the purpose of human versus machine translation. Our corpus is word and sentence-aligned, allowing for the best curation of meaning, specifically tailored for use as training data for machine translation models.
☆ AlphaLoRA: Assigning LoRA Experts Based on Layer Training Quality
Parameter-efficient fine-tuning methods, such as Low-Rank Adaptation (LoRA), are known to enhance training efficiency in Large Language Models (LLMs). Due to the limited parameters of LoRA, recent studies seek to combine LoRA with Mixture-of-Experts (MoE) to boost performance across various tasks. However, inspired by the observed redundancy in traditional MoE structures, previous studies identify similar redundancy among LoRA experts within the MoE architecture, highlighting the necessity for non-uniform allocation of LoRA experts across different layers. In this paper, we leverage Heavy-Tailed Self-Regularization (HT-SR) Theory to design a fine-grained allocation strategy. Our analysis reveals that the number of experts per layer correlates with layer training quality, which exhibits significant variability across layers. Based on this, we introduce AlphaLoRA, a theoretically principled and training-free method for allocating LoRA experts to further mitigate redundancy. Experiments on three models across ten language processing and reasoning benchmarks demonstrate that AlphaLoRA achieves comparable or superior performance over all baselines. Our code is available at https://github.com/morelife2017/alphalora.
comment: The 2024 Conference on Empirical Methods in Natural Language Processing
☆ LLM Unlearning via Loss Adjustment with Only Forget Data
Unlearning in Large Language Models (LLMs) is essential for ensuring ethical and responsible AI use, especially in addressing privacy leak, bias, safety, and evolving regulations. Existing approaches to LLM unlearning often rely on retain data or a reference LLM, yet they struggle to adequately balance unlearning performance with overall model utility. This challenge arises because leveraging explicit retain data or implicit knowledge of retain data from a reference LLM to fine-tune the model tends to blur the boundaries between the forgotten and retain data, as different queries often elicit similar responses. In this work, we propose eliminating the need to retain data or the reference LLM for response calibration in LLM unlearning. Recognizing that directly applying gradient ascent on the forget data often leads to optimization instability and poor performance, our method guides the LLM on what not to respond to, and importantly, how to respond, based on the forget data. Hence, we introduce Forget data only Loss AjustmenT (FLAT), a "flat" loss adjustment approach which addresses these issues by maximizing f-divergence between the available template answer and the forget answer only w.r.t. the forget data. The variational form of the defined f-divergence theoretically provides a way of loss adjustment by assigning different importance weights for the learning w.r.t. template responses and the forgetting of responses subject to unlearning. Empirical results demonstrate that our approach not only achieves superior unlearning performance compared to existing methods but also minimizes the impact on the model's retained capabilities, ensuring high utility across diverse tasks, including copyrighted content unlearning on Harry Potter dataset and MUSE Benchmark, and entity unlearning on the TOFU dataset.
comment: Paper under review
☆ Mimetic Initialization Helps State Space Models Learn to Recall
Recent work has shown that state space models such as Mamba are significantly worse than Transformers on recall-based tasks due to the fact that their state size is constant with respect to their input sequence length. But in practice, state space models have fairly large state sizes, and we conjecture that they should be able to perform much better at these tasks than previously reported. We investigate whether their poor copying and recall performance could be due in part to training difficulties rather than fundamental capacity constraints. Based on observations of their "attention" maps, we propose a structured initialization technique that allows state space layers to more readily mimic attention. Across a variety of architecture settings, our initialization makes it substantially easier for Mamba to learn to copy and do associative recall from scratch.
☆ IsoChronoMeter: A simple and effective isochronic translation evaluation metric EMNLP24
Machine translation (MT) has come a long way and is readily employed in production systems to serve millions of users daily. With the recent advances in generative AI, a new form of translation is becoming possible - video dubbing. This work motivates the importance of isochronic translation, especially in the context of automatic dubbing, and introduces `IsoChronoMeter' (ICM). ICM is a simple yet effective metric to measure isochrony of translations in a scalable and resource-efficient way without the need for gold data, based on state-of-the-art text-to-speech (TTS) duration predictors. We motivate IsoChronoMeter and demonstrate its effectiveness. Using ICM we demonstrate the shortcomings of state-of-the-art translation systems and show the need for new methods. We release the code at this URL: \url{https://github.com/braskai/isochronometer}.
comment: WMT24 (co-located with EMNLP24), Accepted to Shared Task Track, 6 pages, 2 figures, 4 tables, 2 pages references
☆ A Systematic Review on Prompt Engineering in Large Language Models for K-12 STEM Education
Large language models (LLMs) have the potential to enhance K-12 STEM education by improving both teaching and learning processes. While previous studies have shown promising results, there is still a lack of comprehensive understanding regarding how LLMs are effectively applied, specifically through prompt engineering-the process of designing prompts to generate desired outputs. To address this gap, our study investigates empirical research published between 2021 and 2024 that explores the use of LLMs combined with prompt engineering in K-12 STEM education. Following the PRISMA protocol, we screened 2,654 papers and selected 30 studies for analysis. Our review identifies the prompting strategies employed, the types of LLMs used, methods of evaluating effectiveness, and limitations in prior work. Results indicate that while simple and zero-shot prompting are commonly used, more advanced techniques like few-shot and chain-of-thought prompting have demonstrated positive outcomes for various educational tasks. GPT-series models are predominantly used, but smaller and fine-tuned models (e.g., Blender 7B) paired with effective prompt engineering outperform prompting larger models (e.g., GPT-3) in specific contexts. Evaluation methods vary significantly, with limited empirical validation in real-world settings.
☆ ChuLo: Chunk-Level Key Information Representation for Long Document Processing ICLR 2025
Transformer-based models have achieved remarkable success in various Natural Language Processing (NLP) tasks, yet their ability to handle long documents is constrained by computational limitations. Traditional approaches, such as truncating inputs, sparse self-attention, and chunking, attempt to mitigate these issues, but they often lead to information loss and hinder the model's ability to capture long-range dependencies. In this paper, we introduce ChuLo, a novel chunk representation method for long document classification that addresses these limitations. Our ChuLo groups input tokens using unsupervised keyphrase extraction, emphasizing semantically important keyphrase based chunk to retain core document content while reducing input length. This approach minimizes information loss and improves the efficiency of Transformer-based models. Preserving all tokens in long document understanding, especially token classification tasks, is especially important to ensure that fine-grained annotations, which depend on the entire sequence context, are not lost. We evaluate our method on multiple long document classification tasks and long document token classification tasks, demonstrating its effectiveness through comprehensive qualitative and quantitative analyses.
comment: Submitted to ICLR 2025
☆ Active Learning for Robust and Representative LLM Generation in Safety-Critical Scenarios
Ensuring robust safety measures across a wide range of scenarios is crucial for user-facing systems. While Large Language Models (LLMs) can generate valuable data for safety measures, they often exhibit distributional biases, focusing on common scenarios and neglecting rare but critical cases. This can undermine the effectiveness of safety protocols developed using such data. To address this, we propose a novel framework that integrates active learning with clustering to guide LLM generation, enhancing their representativeness and robustness in safety scenarios. We demonstrate the effectiveness of our approach by constructing a dataset of 5.4K potential safety violations through an iterative process involving LLM generation and an active learner model's feedback. Our results show that the proposed framework produces a more representative set of safety scenarios without requiring prior knowledge of the underlying data distribution. Additionally, data acquired through our method improves the accuracy and F1 score of both the active learner model as well models outside the scope of active learning process, highlighting its broad applicability.
☆ JOOCI: a Framework for Learning Comprehensive Speech Representations ICLR 2025
Information in speech can be divided into two categories: what is being said (content) and how it is expressed (other). Current state-of-the-art (SOTA) techniques model speech at fixed segments, usually 10-25 ms, using a single embedding. Given the orthogonal nature of other and content information, attempting to optimize both within a single embedding results in suboptimal solutions. This approach divides the models capacity, limiting its ability to build complex hierarchical features effectively. In this work, we present an end-to-end speech representation learning framework designed to jointly optimize the other and content information (JOOCI) in speech. By using separate learnable parameters, JOOCI addresses this optimization challenge by modeling other and content information independently. Our results show that JOOCI consistently outperforms other SOTA models of similar size (100 million parameters) and pre-training data used (960 hours) by a significant margin when evaluated on a range of speech downstream tasks in the SUPERB benchmark, as shown in Table 1.
comment: Submitted to ICLR 2025
☆ Gender Bias in Decision-Making with Large Language Models: A Study of Relationship Conflicts EMNLP
Large language models (LLMs) acquire beliefs about gender from training data and can therefore generate text with stereotypical gender attitudes. Prior studies have demonstrated model generations favor one gender or exhibit stereotypes about gender, but have not investigated the complex dynamics that can influence model reasoning and decision-making involving gender. We study gender equity within LLMs through a decision-making lens with a new dataset, DeMET Prompts, containing scenarios related to intimate, romantic relationships. We explore nine relationship configurations through name pairs across three name lists (men, women, neutral). We investigate equity in the context of gender roles through numerous lenses: typical and gender-neutral names, with and without model safety enhancements, same and mixed-gender relationships, and egalitarian versus traditional scenarios across various topics. While all models exhibit the same biases (women favored, then those with gender-neutral names, and lastly men), safety guardrails reduce bias. In addition, models tend to circumvent traditional male dominance stereotypes and side with 'traditionally female' individuals more often, suggesting relationships are viewed as a female domain by the models.
comment: EMNLP Findings 2024
☆ Code-Mixer Ya Nahi: Novel Approaches to Measuring Multilingual LLMs' Code-Mixing Capabilities COLING 2025
Multilingual Large Language Models (LLMs) have demonstrated exceptional performance in Machine Translation (MT) tasks. However, their MT abilities in the context of code-switching (the practice of mixing two or more languages in an utterance) remain under-explored. In this paper, we introduce Rule-Based Prompting, a novel prompting technique to generate code-mixed sentences. We measure and compare the code-mixed MT abilities of 3 popular multilingual LLMs: GPT-3.5-turbo, GPT-4, and Gemini Pro across five language pairs: English-{Hindi, Bengali, Gujarati, French, Spanish} using $k$-shot prompting ($k\in\{0, 1, 10, 20\}$) and Rule-Based Prompting. Our findings suggest that though $k$-shot prompting often leads to the best results, Rule-Based prompting shows promise in generating unique code-mixed sentences that vary in their style of code-mixing. We also use $k$-shot prompting to gauge the code-mixed to English translation abilities of multilingual LLMs. For this purpose, we create a gold-standard code-mixed dataset spanning five language pairs: English-{Hindi, Bengali, Gujarati, French, Spanish}. As a real-world application of our work, we create a code-mixed chatbot.
comment: Manuscript submitted to COLING 2025
☆ PRACTIQ: A Practical Conversational Text-to-SQL dataset with Ambiguous and Unanswerable Queries
Previous text-to-SQL datasets and systems have primarily focused on user questions with clear intentions that can be answered. However, real user questions can often be ambiguous with multiple interpretations or unanswerable due to a lack of relevant data. In this work, we construct a practical conversational text-to-SQL dataset called PRACTIQ, consisting of ambiguous and unanswerable questions inspired by real-world user questions. We first identified four categories of ambiguous questions and four categories of unanswerable questions by studying existing text-to-SQL datasets. Then, we generate conversations with four turns: the initial user question, an assistant response seeking clarification, the user's clarification, and the assistant's clarified SQL response with the natural language explanation of the execution results. For some ambiguous queries, we also directly generate helpful SQL responses, that consider multiple aspects of ambiguity, instead of requesting user clarification. To benchmark the performance on ambiguous, unanswerable, and answerable questions, we implemented large language model (LLM)-based baselines using various LLMs. Our approach involves two steps: question category classification and clarification SQL prediction. Our experiments reveal that state-of-the-art systems struggle to handle ambiguous and unanswerable questions effectively. We will release our code for data generation and experiments on GitHub.
☆ An Annotated Dataset of Errors in Premodern Greek and Baselines for Detecting Them
As premodern texts are passed down over centuries, errors inevitably accrue. These errors can be challenging to identify, as some have survived undetected for so long precisely because they are so elusive. While prior work has evaluated error detection methods on artificially-generated errors, we introduce the first dataset of real errors in premodern Greek, enabling the evaluation of error detection methods on errors that genuinely accumulated at some stage in the centuries-long copying process. To create this dataset, we use metrics derived from BERT conditionals to sample 1,000 words more likely to contain errors, which are then annotated and labeled by a domain expert as errors or not. We then propose and evaluate new error detection methods and find that our discriminator-based detector outperforms all other methods, improving the true positive rate for classifying real errors by 5%. We additionally observe that scribal errors are more difficult to detect than print or digitization errors. Our dataset enables the evaluation of error detection methods on real errors in premodern texts for the first time, providing a benchmark for developing more effective error detection algorithms to assist scholars in restoring premodern works.
☆ Assessing Bias in Metric Models for LLM Open-Ended Generation Bias Benchmarks NeurIPS 2024
Open-generation bias benchmarks evaluate social biases in Large Language Models (LLMs) by analyzing their outputs. However, the classifiers used in analysis often have inherent biases, leading to unfair conclusions. This study examines such biases in open-generation benchmarks like BOLD and SAGED. Using the MGSD dataset, we conduct two experiments. The first uses counterfactuals to measure prediction variations across demographic groups by altering stereotype-related prefixes. The second applies explainability tools (SHAP) to validate that the observed biases stem from these counterfactuals. Results reveal unequal treatment of demographic descriptors, calling for more robust bias metric models.
comment: NeurIPS 2024 EvalEval Workshop
☆ Beyond Human-Only: Evaluating Human-Machine Collaboration for Collecting High-Quality Translation Data
Collecting high-quality translations is crucial for the development and evaluation of machine translation systems. However, traditional human-only approaches are costly and slow. This study presents a comprehensive investigation of 11 approaches for acquiring translation data, including human-only, machineonly, and hybrid approaches. Our findings demonstrate that human-machine collaboration can match or even exceed the quality of human-only translations, while being more cost-efficient. Error analysis reveals the complementary strengths between human and machine contributions, highlighting the effectiveness of collaborative methods. Cost analysis further demonstrates the economic benefits of human-machine collaboration methods, with some approaches achieving top-tier quality at around 60% of the cost of traditional methods. We release a publicly available dataset containing nearly 18,000 segments of varying translation quality with corresponding human ratings to facilitate future research.
☆ Varying Shades of Wrong: Aligning LLMs with Wrong Answers Only
In the absence of abundant reliable annotations for challenging tasks and contexts, how can we expand the frontier of LLM capabilities with potentially wrong answers? We focus on two research questions: (1) Can LLMs generate reliable preferences among wrong options? And if so, (2) Would alignment with such wrong-over-wrong preferences be helpful? We employ methods based on self-consistency, token probabilities, and LLM-as-a-judge to elicit wrong-over-wrong preferences, and fine-tune language models with preference optimization approaches using these synthesized preferences. Extensive experiments with seven LLMs and eight datasets demonstrate that (1) LLMs do have preliminary capability in distinguishing various shades of wrong, achieving up to 20.9% higher performance than random guess; (2) Alignment with wrong-over-wrong preferences helps LLMs to produce less wrong and sometimes even outright correct answers, while overall improving model calibration.
☆ Personality Differences Drive Conversational Dynamics: A High-Dimensional NLP Approach EMNLP 2024
This paper investigates how the topical flow of dyadic conversations emerges over time and how differences in interlocutors' personality traits contribute to this topical flow. Leveraging text embeddings, we map the trajectories of $N = 1655$ conversations between strangers into a high-dimensional space. Using nonlinear projections and clustering, we then identify when each interlocutor enters and exits various topics. Differences in conversational flow are quantified via $\textit{topic entropy}$, a summary measure of the "spread" of topics covered during a conversation, and $\textit{linguistic alignment}$, a time-varying measure of the cosine similarity between interlocutors' embeddings. Our findings suggest that interlocutors with a larger difference in the personality dimension of openness influence each other to spend more time discussing a wider range of topics and that interlocutors with a larger difference in extraversion experience a larger decrease in linguistic alignment throughout their conversation. We also examine how participants' affect (emotion) changes from before to after a conversation, finding that a larger difference in extraversion predicts a larger difference in affect change and that a greater topic entropy predicts a larger affect increase. This work demonstrates how communication research can be advanced through the use of high-dimensional NLP methods and identifies personality difference as an important driver of social influence.
comment: To be published in the Proceedings of the Second Workshop on Social Influence in Conversations (SICon 2024), co-located with EMNLP 2024
☆ Persistent Topological Features in Large Language Models
Understanding the decision-making processes of large language models (LLMs) is critical given their widespread applications. Towards this goal, describing the topological and geometrical properties of internal representations has recently provided valuable insights. For a more comprehensive characterization of these inherently complex spaces, we present a novel framework based on zigzag persistence, a method in topological data analysis (TDA) well-suited for describing data undergoing dynamic transformations across layers. Within this framework, we introduce persistence similarity, a new metric that quantifies the persistence and transformation of topological features such as $p$-cycles throughout the model layers. Unlike traditional similarity measures, our approach captures the entire evolutionary trajectory of these features, providing deeper insights into the internal workings of LLMs. As a practical application, we leverage persistence similarity to identify and prune redundant layers, demonstrating comparable performance to state-of-the-art methods across several benchmark datasets. Additionally, our analysis reveals consistent topological behaviors across various models and hyperparameter settings, suggesting a universal structure in LLM internal representations.
comment: 10+6 pages, 7 figures, 1 table. All comments welcome!
☆ Improving the Language Understanding Capabilities of Large Language Models Using Reinforcement Learning
Large language models (LLMs), built on decoder-only transformers, excel in natural language generation and adapt to diverse tasks using zero-shot and few-shot prompting. However, these prompting methods often struggle on natural language understanding (NLU) tasks, where encoder-only models like BERT-base outperform LLMs on benchmarks like GLUE and SuperGLUE. This paper explores two approaches-supervised fine-tuning (SFT) and proximal policy optimization (PPO)-to enhance LLMs' NLU abilities. To reduce the cost of full-model fine-tuning, we integrate low-rank adaptation (LoRA) layers, limiting updates to these layers during both SFT and PPO. In SFT, task-specific prompts are concatenated with input queries and ground-truth labels, optimizing with next-token prediction. Despite this, LLMs still underperform compared to models like BERT-base on several NLU tasks. To close this gap, we apply PPO, a reinforcement learning technique that treats each token generation as an action and uses a reward function based on alignment with ground-truth answers. PPO then updates the model to maximize these rewards, aligning outputs with correct labels. Our experiments with LLAMA2-7B show that PPO improves performance, with a 6.3-point gain over SFT on GLUE. PPO exceeds zero-shot by 38.7 points and few-shot by 26.1 points on GLUE, while surpassing these by 28.8 and 28.5 points on SuperGLUE. Additionally, PPO outperforms BERT-large by 2.7 points on GLUE and 9.3 points on SuperGLUE. The improvements are consistent across models like Qwen2.5-7B and MPT-7B, highlighting PPO's robustness in enhancing LLMs' NLU capabilities.
☆ Enhancing AI Assisted Writing with One-Shot Implicit Negative Feedback EMNLP 2024
AI-mediated communication enables users to communicate more quickly and efficiently. Various systems have been proposed such as smart reply and AI-assisted writing. Yet, the heterogeneity of the forms of inputs and architectures often renders it challenging to combine insights from user behaviour in one system to improve performance in another. In this work, we consider the case where the user does not select any of the suggested replies from a smart reply system, and how this can be used as one-shot implicit negative feedback to enhance the accuracy of an AI writing model. We introduce Nifty, an approach that uses classifier guidance to controllably integrate implicit user feedback into the text generation process. Empirically, we find up to 34% improvement in Rouge-L, 89% improvement in generating the correct intent, and an 86% win-rate according to human evaluators compared to a vanilla AI writing system on the MultiWOZ and Schema-Guided Dialog datasets.
comment: Accepted to appear at EMNLP 2024
☆ Assessing the Human Likeness of AI-Generated Counterspeech
Counterspeech is a targeted response to counteract and challenge abusive or hateful content. It can effectively curb the spread of hatred and foster constructive online communication. Previous studies have proposed different strategies for automatically generated counterspeech. Evaluations, however, focus on the relevance, surface form, and other shallow linguistic characteristics. In this paper, we investigate the human likeness of AI-generated counterspeech, a critical factor influencing effectiveness. We implement and evaluate several LLM-based generation strategies, and discover that AI-generated and human-written counterspeech can be easily distinguished by both simple classifiers and humans. Further, we reveal differences in linguistic characteristics, politeness, and specificity.
☆ Effective Self-Mining of In-Context Examples for Unsupervised Machine Translation with LLMs
Large Language Models (LLMs) have demonstrated impressive performance on a wide range of natural language processing (NLP) tasks, primarily through in-context learning (ICL). In ICL, the LLM is provided with examples that represent a given task such that it learns to generate answers for test inputs. However, access to these in-context examples is not guaranteed especially for low-resource or massively multilingual tasks. In this work, we propose an unsupervised approach to mine in-context examples for machine translation (MT), enabling unsupervised MT (UMT) across different languages. Our approach begins with word-level mining to acquire word translations that are then used to perform sentence-level mining. As the quality of mined parallel pairs may not be optimal due to noise or mistakes, we introduce a filtering criterion to select the optimal in-context examples from a pool of unsupervised parallel sentences. We evaluate our approach using two multilingual LLMs on 288 directions from the FLORES-200 dataset and analyze the impact of various linguistic features on performance. Our findings demonstrate the effectiveness of our unsupervised approach in mining in-context examples for MT, leading to better or comparable translation performance as translation with regular in-context samples (extracted from human-annotated data), while also outperforming the other state-of-the-art UMT methods by an average of $7$ BLEU points.
☆ One Language, Many Gaps: Evaluating Dialect Fairness and Robustness of Large Language Models in Reasoning Tasks
Language is not monolithic. While many benchmarks are used as proxies to systematically estimate Large Language Models' (LLM) performance in real-life tasks, they tend to ignore the nuances of within-language variation and thus fail to model the experience of speakers of minority dialects. Focusing on African American Vernacular English (AAVE), we present the first study on LLMs' fairness and robustness to a dialect in canonical reasoning tasks (algorithm, math, logic, and comprehensive reasoning). We hire AAVE speakers, including experts with computer science backgrounds, to rewrite seven popular benchmarks, such as HumanEval and GSM8K. The result of this effort is ReDial, a dialectal benchmark comprising $1.2K+$ parallel query pairs in Standardized English and AAVE. We use ReDial to evaluate state-of-the-art LLMs, including GPT-4o/4/3.5-turbo, LLaMA-3.1/3, Mistral, and Phi-3. We find that, compared to Standardized English, almost all of these widely used models show significant brittleness and unfairness to queries in AAVE. Furthermore, AAVE queries can degrade performance more substantially than misspelled texts in Standardized English, even when LLMs are more familiar with the AAVE queries. Finally, asking models to rephrase questions in Standardized English does not close the performance gap but generally introduces higher costs. Overall, our findings indicate that LLMs provide unfair service to dialect users in complex reasoning tasks. Code can be found at https://github.com/fangru-lin/redial_dialect_robustness_fairness.git.
☆ Graph of Records: Boosting Retrieval Augmented Generation for Long-context Summarization with Graphs
Retrieval-augmented generation (RAG) has revitalized Large Language Models (LLMs) by injecting non-parametric factual knowledge. Compared with long-context LLMs, RAG is considered an effective summarization tool in a more concise and lightweight manner, which can interact with LLMs multiple times using diverse queries to get comprehensive responses. However, the LLM-generated historical responses, which contain potentially insightful information, are largely neglected and discarded by existing approaches, leading to suboptimal results. In this paper, we propose \textit{graph of records} (\textbf{GoR}), which leverages historical responses generated by LLMs to enhance RAG for long-context global summarization. Inspired by the \textit{retrieve-then-generate} paradigm of RAG, we construct a graph by establishing an edge between the retrieved text chunks and the corresponding LLM-generated response. To further uncover the intricate correlations between them, GoR further features a \textit{graph neural network} and an elaborately designed \textit{BERTScore}-based objective for self-supervised model training, enabling seamless supervision signal backpropagation between reference summaries and node embeddings. We comprehensively compare GoR with 12 baselines across four long-context summarization datasets, and the results indicate that our proposed method reaches the best performance e.g., 15\%, 8\%, and 19\% improvement over retrievers w.r.t. Rouge-L, Rouge-1, and Rouge-2 on the WCEP dataset). Extensive experiments further demonstrate the effectiveness of GoR. Code is available at https://github.com/ulab-uiuc/GoR
☆ Watching the Watchers: Exposing Gender Disparities in Machine Translation Quality Estimation
The automatic assessment of translation quality has recently become crucial for many stages of the translation pipeline, from data curation to training and decoding. However, while quality estimation metrics have been optimized to align with human judgments, no attention has been given to these metrics' potential biases, particularly in reinforcing visibility and usability for some demographic groups over others. This paper is the first to investigate gender bias in quality estimation (QE) metrics and its downstream impact on machine translation (MT). We focus on out-of-English translations where the target language uses grammatical gender. We ask: (RQ1) Do contemporary QE metrics exhibit gender bias? (RQ2) Can the use of contextual information mitigate this bias? (RQ3) How does QE influence gender bias in MT outputs? Experiments with state-of-the-art QE metrics across multiple domains, datasets, and languages reveal significant bias. Masculine-inflected translations score higher than feminine-inflected ones, and gender-neutral translations are penalized. Moreover, context-aware QE metrics reduce errors for masculine-inflected references but fail to address feminine referents, exacerbating gender disparities. Additionally, we show that QE metrics can perpetuate gender bias in MT systems when used in quality-aware decoding. Our findings highlight the need to address gender bias in QE metrics to ensure equitable and unbiased MT systems.
comment: Work in progress
☆ Performance in a dialectal profiling task of LLMs for varieties of Brazilian Portuguese
Different of biases are reproduced in LLM-generated responses, including dialectal biases. A study based on prompt engineering was carried out to uncover how LLMs discriminate varieties of Brazilian Portuguese, specifically if sociolinguistic rules are taken into account in four LLMs: GPT 3.5, GPT-4o, Gemini, and Sabi.-2. The results offer sociolinguistic contributions for an equity fluent NLP technology.
comment: 8 pages, XI Jornada de Descri\c{c}\~ao do Portugu\^es
☆ Liger Kernel: Efficient Triton Kernels for LLM Training
Training Large Language Models (LLMs) efficiently at scale presents a formidable challenge, driven by their ever-increasing computational demands and the need for enhanced performance. In this work, we introduce Liger-Kernel, an open-sourced set of Triton kernels developed specifically for LLM training. With kernel optimization techniques like kernel operation fusing and input chunking, our kernels achieve on average a 20% increase in training throughput and a 60% reduction in GPU memory usage for popular LLMs compared to HuggingFace implementations. In addition, Liger-Kernel is designed with modularity, accessibility, and adaptability in mind, catering to both casual and expert users. Comprehensive benchmarks and integration tests are built in to ensure compatibility, performance, correctness, and convergence across diverse computing environments and model architectures. The source code is available under a permissive license at: github.com/linkedin/Liger-Kernel.
comment: 17 pages, 12 figures
♻ ☆ Reuse Your Rewards: Reward Model Transfer for Zero-Shot Cross-Lingual Alignment EMNLP 2024
Aligning language models (LMs) based on human-annotated preference data is a crucial step in obtaining practical and performant LM-based systems. However, multilingual human preference data are difficult to obtain at scale, making it challenging to extend this framework to diverse languages. In this work, we evaluate a simple approach for zero-shot cross-lingual alignment, where a reward model is trained on preference data in one source language and directly applied to other target languages. On summarization and open-ended dialog generation, we show that this method is consistently successful under comprehensive evaluation settings, including human evaluation: cross-lingually aligned models are preferred by humans over unaligned models on up to >70% of evaluation instances. We moreover find that a different-language reward model sometimes yields better aligned models than a same-language reward model. We also identify best practices when there is no language-specific data for even supervised finetuning, another component in alignment.
comment: EMNLP 2024
♻ ☆ Designing a Dashboard for Transparency and Control of Conversational AI
Conversational LLMs function as black box systems, leaving users guessing about why they see the output they do. This lack of transparency is potentially problematic, especially given concerns around bias and truthfulness. To address this issue, we present an end-to-end prototype-connecting interpretability techniques with user experience design-that seeks to make chatbots more transparent. We begin by showing evidence that a prominent open-source LLM has a "user model": examining the internal state of the system, we can extract data related to a user's age, gender, educational level, and socioeconomic status. Next, we describe the design of a dashboard that accompanies the chatbot interface, displaying this user model in real time. The dashboard can also be used to control the user model and the system's behavior. Finally, we discuss a study in which users conversed with the instrumented system. Our results suggest that users appreciate seeing internal states, which helped them expose biased behavior and increased their sense of control. Participants also made valuable suggestions that point to future directions for both design and machine learning research. The project page and video demo of our TalkTuner system are available at https://bit.ly/talktuner-project-page
comment: Project page: https://bit.ly/talktuner-project-page, 38 pages, 23 figures
♻ ☆ Enhancing Indonesian Automatic Speech Recognition: Evaluating Multilingual Models with Diverse Speech Variabilities
An ideal speech recognition model has the capability to transcribe speech accurately under various characteristics of speech signals, such as speaking style (read and spontaneous), speech context (formal and informal), and background noise conditions (clean and moderate). Building such a model requires a significant amount of training data with diverse speech characteristics. Currently, Indonesian data is dominated by read, formal, and clean speech, leading to a scarcity of Indonesian data with other speech variabilities. To develop Indonesian automatic speech recognition (ASR), we present our research on state-of-the-art speech recognition models, namely Massively Multilingual Speech (MMS) and Whisper, as well as compiling a dataset comprising Indonesian speech with variabilities to facilitate our study. We further investigate the models' predictive ability to transcribe Indonesian speech data across different variability groups. The best results were achieved by the Whisper fine-tuned model across datasets with various characteristics, as indicated by the decrease in word error rate (WER) and character error rate (CER). Moreover, we found that speaking style variability affected model performance the most.
comment: Accepted at O-COCOSDA 2024
♻ ☆ SimpleStrat: Diversifying Language Model Generation with Stratification
Generating diverse responses from large language models (LLMs) is crucial for applications such as planning/search and synthetic data generation, where diversity provides distinct answers across generations. Prior approaches rely on increasing temperature to increase diversity. However, contrary to popular belief, we show not only does this approach produce lower quality individual generations as temperature increases, but it depends on model's next-token probabilities being similar to the true distribution of answers. We propose SimpleStrat, an alternative approach that uses the language model itself to partition the space into strata. At inference, a random stratum is selected and a sample drawn from within the strata. To measure diversity, we introduce CoverageQA, a dataset of underspecified questions with multiple equally plausible answers, and assess diversity by measuring KL Divergence between the output distribution and uniform distribution over valid ground truth answers. As computing probability per response/solution for proprietary models is infeasible, we measure recall on ground truth solutions. Our evaluation show using SimpleStrat achieves higher recall by 0.05 compared to GPT-4o and 0.36 average reduction in KL Divergence compared to Llama 3.
♻ ☆ AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents
The robustness of LLMs to jailbreak attacks, where users design prompts to circumvent safety measures and misuse model capabilities, has been studied primarily for LLMs acting as simple chatbots. Meanwhile, LLM agents -- which use external tools and can execute multi-stage tasks -- may pose a greater risk if misused, but their robustness remains underexplored. To facilitate research on LLM agent misuse, we propose a new benchmark called AgentHarm. The benchmark includes a diverse set of 110 explicitly malicious agent tasks (440 with augmentations), covering 11 harm categories including fraud, cybercrime, and harassment. In addition to measuring whether models refuse harmful agentic requests, scoring well on AgentHarm requires jailbroken agents to maintain their capabilities following an attack to complete a multi-step task. We evaluate a range of leading LLMs, and find (1) leading LLMs are surprisingly compliant with malicious agent requests without jailbreaking, (2) simple universal jailbreak templates can be adapted to effectively jailbreak agents, and (3) these jailbreaks enable coherent and malicious multi-step agent behavior and retain model capabilities. To enable simple and reliable evaluation of attacks and defenses for LLM-based agents, we publicly release AgentHarm at https://huggingface.co/datasets/ai-safety-institute/AgentHarm.
♻ ☆ Sign Stitching: A Novel Approach to Sign Language Production BMVC 2024
Sign Language Production (SLP) is a challenging task, given the limited resources available and the inherent diversity within sign data. As a result, previous works have suffered from the problem of regression to the mean, leading to under-articulated and incomprehensible signing. In this paper, we propose using dictionary examples to create expressive sign language sequences. However, simply concatenating the signs would create robotic and unnatural sequences. Therefore, we present a 7-step approach to effectively stitch the signs together. First, by normalising each sign into a canonical pose, cropping and stitching we create a continuous sequence. Then by applying filtering in the frequency domain and resampling each sign we create cohesive natural sequences, that mimic the prosody found in the original data. We leverage the SignGAN model to map the output to a photo-realistic signer and present a complete Text-to-Sign (T2S) SLP pipeline. Our evaluation demonstrates the effectiveness of this approach, showcasing state-of-the-art performance across all datasets.
comment: 18 pages, 3 figures, 4 tables. Accepted to BMVC 2024
♻ ☆ Building Knowledge-Guided Lexica to Model Cultural Variation NAACL 2024
Cultural variation exists between nations (e.g., the United States vs. China), but also within regions (e.g., California vs. Texas, Los Angeles vs. San Francisco). Measuring this regional cultural variation can illuminate how and why people think and behave differently. Historically, it has been difficult to computationally model cultural variation due to a lack of training data and scalability constraints. In this work, we introduce a new research problem for the NLP community: How do we measure variation in cultural constructs across regions using language? We then provide a scalable solution: building knowledge-guided lexica to model cultural variation, encouraging future work at the intersection of NLP and cultural understanding. We also highlight modern LLMs' failure to measure cultural variation or generate culturally varied language.
comment: Accepted at NAACL 2024
♻ ☆ Translating Across Cultures: LLMs for Intralingual Cultural Adaptation CoNLL 2024
LLMs are increasingly being deployed for multilingual applications and have demonstrated impressive translation capabilities between several low and high-resource languages. An aspect of translation that often gets overlooked is that of cultural adaptation, or modifying source culture references to suit the target culture. While specialized translation models still outperform LLMs on the machine translation task when viewed from the lens of correctness, they are not sensitive to cultural differences often requiring manual correction. LLMs on the other hand have a rich reservoir of cultural knowledge embedded within its parameters that can be potentially exploited for such applications. In this paper, we define the task of cultural adaptation and create an evaluation framework to evaluate the performance of modern LLMs for cultural adaptation and analyze their cross-cultural knowledge while connecting related concepts across different cultures. We also analyze possible issues with automatic adaptation. We hope that this task will offer more insight into the cultural understanding of LLMs and their creativity in cross-cultural scenarios.
comment: Accepted to CoNLL 2024
♻ ☆ Mitigating the Risk of Health Inequity Exacerbated by Large Language Models
Recent advancements in large language models have demonstrated their potential in numerous medical applications, particularly in automating clinical trial matching for translational research and enhancing medical question answering for clinical decision support. However, our study shows that incorporating non decisive sociodemographic factors such as race, sex, income level, LGBT+ status, homelessness, illiteracy, disability, and unemployment into the input of LLMs can lead to incorrect and harmful outputs for these populations. These discrepancies risk exacerbating existing health disparities if LLMs are widely adopted in healthcare. To address this issue, we introduce EquityGuard, a novel framework designed to detect and mitigate the risk of health inequities in LLM based medical applications. Our evaluation demonstrates its efficacy in promoting equitable outcomes across diverse populations.
♻ ☆ BA-LoRA: Bias-Alleviating Low-Rank Adaptation to Mitigate Catastrophic Inheritance in Large Language Models
Large language models (LLMs) have demonstrated remarkable proficiency across various natural language processing (NLP) tasks. However, adapting LLMs to downstream applications requires computationally intensive and memory-demanding fine-tuning procedures. To alleviate these burdens, parameter-efficient fine-tuning (PEFT) techniques have emerged as a promising approach to tailor LLMs with minimal computational overhead. While PEFT methods offer substantial advantages, they do not fully address the pervasive issue of bias propagation from pre-training data. This work introduces Bias-Alleviating Low-Rank Adaptation (BA-LoRA), a novel PEFT method designed to counteract bias inheritance. BA-LoRA incorporates three distinct regularization terms: (1) a consistency regularizer, (2) a diversity regularizer, and (3) a singular value decomposition regularizer. These regularizers aim to enhance the models' consistency, diversity, and generalization capabilities during fine-tuning. We conduct extensive experiments on natural language understanding (NLU) and natural language generation (NLG) tasks using prominent LLMs such as LLaMA, Mistral, and Gemma. The results demonstrate that BA-LoRA outperforms LoRA and its state-of-the-art variants. Moreover, our method effectively mitigates the adverse effects of pre-training bias, leading to more reliable and robust model outputs. The code is available at https://github.com/cyp-jlu-ai/BA-LoRA.
comment: 23 pages
♻ ☆ Towards a Psychology of Machines: Large Language Models Predict Human Memory
Large language models (LLMs) are excelling across various tasks despite not being based on human cognition, prompting an investigation into their potential to offer insights into human cognitive mechanisms. This study examines ChatGPT's ability to predict human performance in a language-based memory task. Following theories of text comprehension, we hypothesized that recognizing ambiguous sentences is easier with relevant preceding context. Participants, including humans and ChatGPT, were given pairs of sentences: the second always a garden-path sentence, and the first providing either fitting or unfitting context. We measured their ratings of sentence relatedness and memorability. Results showed a strong alignment between ChatGPT's assessments and human memory performance. Sentences in the fitting context were rated as being more related and memorable by ChatGPT and were better remembered by humans, highlighting LLMs' potential to predict human performance and contribute to psychological theories.
comment: 33 pages, 3 figures, 2 tables
♻ ☆ Which Demographics do LLMs Default to During Annotation?
Demographics and cultural background of annotators influence the labels they assign in text annotation -- for instance, an elderly woman might find it offensive to read a message addressed to a "bro", but a male teenager might find it appropriate. It is therefore important to acknowledge label variations to not under-represent members of a society. Two research directions developed out of this observation in the context of using large language models (LLM) for data annotations, namely (1) studying biases and inherent knowledge of LLMs and (2) injecting diversity in the output by manipulating the prompt with demographic information. We combine these two strands of research and ask the question to which demographics an LLM resorts to when no demographics is given. To answer this question, we evaluate which attributes of human annotators LLMs inherently mimic. Furthermore, we compare non-demographic conditioned prompts and placebo-conditioned prompts (e.g., "you are an annotator who lives in house number 5") to demographics-conditioned prompts ("You are a 45 year old man and an expert on politeness annotation. How do you rate {instance}"). We study these questions for politeness and offensiveness annotations on the POPQUORN data set, a corpus created in a controlled manner to investigate human label variations based on demographics which has not been used for LLM-based analyses so far. We observe notable influences related to gender, race, and age in demographic prompting, which contrasts with previous studies that found no such effects.
♻ ☆ Multi-Agent Collaborative Data Selection for Efficient LLM Pretraining
Efficient data selection is crucial to accelerate the pretraining of large language models (LLMs). While various methods have been proposed to enhance data efficiency, limited research has addressed the inherent conflicts between these approaches to achieve optimal data selection for LLM pretraining. To tackle this problem, we propose a novel multi-agent collaborative data selection mechanism. In this framework, each data selection method serves as an independent agent, and an agent console is designed to dynamically integrate the information from all agents throughout the LLM training process. We conduct extensive empirical studies to evaluate our multi-agent framework. The experimental results demonstrate that our approach significantly improves data efficiency, accelerates convergence in LLM training, and achieves an average performance gain up to 10.5% across multiple language model benchmarks compared to the state-of-the-art methods.
♻ ☆ Uplifting Lower-Income Data: Strategies for Socioeconomic Perspective Shifts in Large Multi-modal Models
Recent work has demonstrated that the unequal representation of cultures and socioeconomic groups in training data leads to biased Large Multi-modal (LMM) models. To improve LMM model performance on underrepresented data, we propose and evaluate several prompting strategies using non-English, geographic, and socioeconomic attributes. We show that these geographic and socioeconomic integrated prompts favor retrieving topic appearances commonly found in data from low-income households across different countries leading to improved LMM model performance on lower-income data. Our analyses identify and highlight contexts where these strategies yield the most improvements.
♻ ☆ Let Me Speak Freely? A Study on the Impact of Format Restrictions on Performance of Large Language Models
Structured generation, the process of producing content in standardized formats like JSON and XML, is widely utilized in real-world applications to extract key output information from large language models (LLMs). This study investigates whether such constraints on generation space impact LLMs abilities, including reasoning and domain knowledge comprehension. Specifically, we evaluate LLMs performance when restricted to adhere to structured formats versus generating free-form responses across various common tasks. Surprisingly, we observe a significant decline in LLMs reasoning abilities under format restrictions. Furthermore, we find that stricter format constraints generally lead to greater performance degradation in reasoning tasks.
comment: 18 pages
♻ ☆ PAD: Personalized Alignment at Decoding-Time
Aligning with personalized preferences, which vary significantly across cultural, educational, and political differences, poses a significant challenge due to the computational costs and data demands of traditional alignment methods. In response, this paper presents Personalized Alignment at Decoding-time (PAD), a novel framework designed to align LLM outputs with diverse personalized preferences during the inference phase, eliminating the need for additional training. By introducing a unique personalized reward modeling strategy, this framework decouples the text generation process from personalized preferences, facilitating the generation of generalizable token-level personalized rewards. The PAD algorithm leverages these rewards to guide the decoding process, dynamically tailoring the base model's predictions to personalized preferences. Extensive experimental results demonstrate that PAD not only outperforms existing training-based alignment methods in terms of aligning with diverse preferences but also shows significant generalizability to preferences unseen during training and scalability across different base models. This work advances the capability of LLMs to meet user needs in real-time applications, presenting a substantial step forward in personalized LLM alignment.
comment: This paper presents Personalized Alignment at Decoding-time (PAD), a novel framework designed to align LLM outputs with diverse personalized preferences during the inference phase
♻ ☆ TACT: Advancing Complex Aggregative Reasoning with Information Extraction Tools NeurIPS 2024
Large Language Models (LLMs) often do not perform well on queries that require the aggregation of information across texts. To better evaluate this setting and facilitate modeling efforts, we introduce TACT - Text And Calculations through Tables, a dataset crafted to evaluate LLMs' reasoning and computational abilities using complex instructions. TACT contains challenging instructions that demand stitching information scattered across one or more texts, and performing complex integration on this information to generate the answer. We construct this dataset by leveraging an existing dataset of texts and their associated tables. For each such tables, we formulate new queries, and gather their respective answers. We demonstrate that all contemporary LLMs perform poorly on this dataset, achieving an accuracy below 38%. To pinpoint the difficulties and thoroughly dissect the problem, we analyze model performance across three components: table-generation, Pandas command-generation, and execution. Unexpectedly, we discover that each component presents substantial challenges for current LLMs. These insights lead us to propose a focused modeling framework, which we refer to as IE as a tool. Specifically, we propose to add "tools" for each of the above steps, and implement each such tool with few-shot prompting. This approach shows an improvement over existing prompting techniques, offering a promising direction for enhancing model capabilities in these tasks.
comment: Accepted to NeurIPS 2024. Website (https://tact-benchmark.github.io), Huggingface (https://huggingface.co/datasets/google/TACT)
♻ ☆ Benchmarking Vision Language Models for Cultural Understanding EMNLP 2024
Foundation models and vision-language pre-training have notably advanced Vision Language Models (VLMs), enabling multimodal processing of visual and linguistic data. However, their performance has been typically assessed on general scene understanding - recognizing objects, attributes, and actions - rather than cultural comprehension. This study introduces CulturalVQA, a visual question-answering benchmark aimed at assessing VLM's geo-diverse cultural understanding. We curate a collection of 2,378 image-question pairs with 1-5 answers per question representing cultures from 11 countries across 5 continents. The questions probe understanding of various facets of culture such as clothing, food, drinks, rituals, and traditions. Benchmarking VLMs on CulturalVQA, including GPT-4V and Gemini, reveals disparity in their level of cultural understanding across regions, with strong cultural understanding capabilities for North America while significantly lower performance for Africa. We observe disparity in their performance across cultural facets too, with clothing, rituals, and traditions seeing higher performances than food and drink. These disparities help us identify areas where VLMs lack cultural understanding and demonstrate the potential of CulturalVQA as a comprehensive evaluation set for gauging VLM progress in understanding diverse cultures.
comment: Accepted to EMNLP 2024 Main Conference
♻ ☆ Language Model Alignment in Multilingual Trolley Problems
We evaluate the moral alignment of large language models (LLMs) with human preferences in multilingual trolley problems. Building on the Moral Machine experiment, which captures over 40 million human judgments across 200+ countries, we develop a cross-lingual corpus of moral dilemma vignettes in over 100 languages called $\mathrm{MultiTP}$. This dataset enables the assessment of LLMs' decision-making processes in diverse linguistic contexts. Our analysis explores the alignment of 19 different LLMs with human judgments, capturing preferences across six moral dimensions: species, gender, fitness, status, age, and the number of lives involved. By correlating these preferences with the demographic distribution of language speakers and examining the consistency of LLM responses to various prompt paraphrasings, our findings provide insights into cross-lingual and ethical biases of LLMs and their intersection. We discover significant variance in alignment across languages, challenging the assumption of uniform moral reasoning in AI systems and highlighting the importance of incorporating diverse perspectives in AI ethics. The results underscore the need for further research on the integration of multilingual dimensions in responsible AI research to ensure fair and equitable AI interactions worldwide.
♻ ☆ CompAct: Compressing Retrieved Documents Actively for Question Answering EMNLP 2024
Retrieval-augmented generation supports language models to strengthen their factual groundings by providing external contexts. However, language models often face challenges when given extensive information, diminishing their effectiveness in solving questions. Context compression tackles this issue by filtering out irrelevant information, but current methods still struggle in realistic scenarios where crucial information cannot be captured with a single-step approach. To overcome this limitation, we introduce CompAct, a novel framework that employs an active strategy to condense extensive documents without losing key information. Our experiments demonstrate that CompAct brings significant improvements in both performance and compression rate on multi-hop question-answering benchmarks. CompAct flexibly operates as a cost-efficient plug-in module with various off-the-shelf retrievers or readers, achieving exceptionally high compression rates (47x).
comment: Accepted to the main conference at EMNLP 2024
♻ ☆ Should We Respect LLMs? A Cross-Lingual Study on the Influence of Prompt Politeness on LLM Performance
We investigate the impact of politeness levels in prompts on the performance of large language models (LLMs). Polite language in human communications often garners more compliance and effectiveness, while rudeness can cause aversion, impacting response quality. We consider that LLMs mirror human communication traits, suggesting they align with human cultural norms. We assess the impact of politeness in prompts on LLMs across English, Chinese, and Japanese tasks. We observed that impolite prompts often result in poor performance, but overly polite language does not guarantee better outcomes. The best politeness level is different according to the language. This phenomenon suggests that LLMs not only reflect human behavior but are also influenced by language, particularly in different cultural contexts. Our findings highlight the need to factor in politeness for cross-cultural natural language processing and LLM usage.
comment: SICon 2024
♻ ☆ mGTE: Generalized Long-Context Text Representation and Reranking Models for Multilingual Text Retrieval EMNLP 2024
We present systematic efforts in building long-context multilingual text representation model (TRM) and reranker from scratch for text retrieval. We first introduce a text encoder (base size) enhanced with RoPE and unpadding, pre-trained in a native 8192-token context (longer than 512 of previous multilingual encoders). Then we construct a hybrid TRM and a cross-encoder reranker by contrastive learning. Evaluations show that our text encoder outperforms the same-sized previous state-of-the-art XLM-R. Meanwhile, our TRM and reranker match the performance of large-sized state-of-the-art BGE-M3 models and achieve better results on long-context retrieval benchmarks. Further analysis demonstrate that our proposed models exhibit higher efficiency during both training and inference. We believe their efficiency and effectiveness could benefit various researches and industrial applications.
comment: Camera-ready version of EMNLP 2024: Industry Track
♻ ☆ Summarizing long regulatory documents with a multi-step pipeline
Due to their length and complexity, long regulatory texts are challenging to summarize. To address this, a multi-step extractive-abstractive architecture is proposed to handle lengthy regulatory documents more effectively. In this paper, we show that the effectiveness of a two-step architecture for summarizing long regulatory texts varies significantly depending on the model used. Specifically, the two-step architecture improves the performance of decoder-only models. For abstractive encoder-decoder models with short context lengths, the effectiveness of an extractive step varies, whereas for long-context encoder-decoder models, the extractive step worsens their performance. This research also highlights the challenges of evaluating generated texts, as evidenced by the differing results from human and automated evaluations. Most notably, human evaluations favoured language models pretrained on legal text, while automated metrics rank general-purpose language models higher. The results underscore the importance of selecting the appropriate summarization strategy based on model architecture and context length.
comment: Published in: Proceedings of the 6th Workshop on Natural Legal Language Processing (NLLP 2024)
♻ ☆ An Adaptive Placement and Parallelism Framework for Accelerating RLHF Training
Recently, ChatGPT or InstructGPT like large language models (LLM) has made a significant impact in the AI world. Many works have attempted to reproduce the complex InstructGPT's training pipeline, namely Reinforcement Learning with Human Feedback (RLHF). However, the mainstream distributed RLHF training methods typically adopt a fixed model placement strategy, referred to as the Co-located strategy. This strategy treats all four interdependent models involved in RLHF as a single entity, distributing them across all devices and applying parallelism techniques designed for a single model, regardless of the workload heterogeneity inherent to each model. As a result, this strategy exacerbates the generation bottlenecks in the RLHF training and degrades the overall training efficiency. To address these issues, we propose a flexible model placement framework that offers two general and agile model placement strategies. The Interleaving strategy helps reduce memory redundancy and communication costs of RLHF training by placing models without dependencies on exclusive devices with careful orchestration. On the other hand, the Disaggregated strategy improves the throughput of model training by separating the training and inference runtime of the RLHF pipeline with additional shadow models. Furthermore, our framework provides a simple user interface and guidelines to easily and flexibly configure these strategies in various training scenarios. Our experiments have shown that our strategy can achieve notable improvements up to 11x, compared to the current state-of-the-art (SOTA) approaches. The results highlight the effectiveness and adaptability of our methods in accelerating the training of distributed RLHF.
♻ ☆ LexEval: A Comprehensive Chinese Legal Benchmark for Evaluating Large Language Models
Large language models (LLMs) have made significant progress in natural language processing tasks and demonstrate considerable potential in the legal domain. However, legal applications demand high standards of accuracy, reliability, and fairness. Applying existing LLMs to legal systems without careful evaluation of their potential and limitations could pose significant risks in legal practice. To this end, we introduce a standardized comprehensive Chinese legal benchmark LexEval. This benchmark is notable in the following three aspects: (1) Ability Modeling: We propose a new taxonomy of legal cognitive abilities to organize different tasks. (2) Scale: To our knowledge, LexEval is currently the largest Chinese legal evaluation dataset, comprising 23 tasks and 14,150 questions. (3) Data: we utilize formatted existing datasets, exam datasets and newly annotated datasets by legal experts to comprehensively evaluate the various capabilities of LLMs. LexEval not only focuses on the ability of LLMs to apply fundamental legal knowledge but also dedicates efforts to examining the ethical issues involved in their application. We evaluated 38 open-source and commercial LLMs and obtained some interesting findings. The experiments and findings offer valuable insights into the challenges and potential solutions for developing Chinese legal systems and LLM evaluation pipelines. The LexEval dataset and leaderboard are publicly available at \url{https://github.com/CSHaitao/LexEval} and will be continuously updated.
comment: NeurIPs 2024
♻ ☆ Induction Heads as an Essential Mechanism for Pattern Matching in In-context Learning
Large language models (LLMs) have shown a remarkable ability to learn and perform complex tasks through in-context learning (ICL). However, a comprehensive understanding of its internal mechanisms is still lacking. This paper explores the role of induction heads in a few-shot ICL setting. We analyse two state-of-the-art models, Llama-3-8B and InternLM2-20B on abstract pattern recognition and NLP tasks. Our results show that even a minimal ablation of induction heads leads to ICL performance decreases of up to ~32% for abstract pattern recognition tasks, bringing the performance close to random. For NLP tasks, this ablation substantially decreases the model's ability to benefit from examples, bringing few-shot ICL performance close to that of zero-shot prompts. We further use attention knockout to disable specific induction patterns, and present fine-grained evidence for the role that the induction mechanism plays in ICL.
comment: 9 pages, 7 figures; mean-ablation experiments instead of zero-ablations
♻ ☆ On Giant's Shoulders: Effortless Weak to Strong by Dynamic Logits Fusion NeurIPS 2024
Efficient fine-tuning of large language models for task-specific applications is imperative, yet the vast number of parameters in these models makes their training increasingly challenging. Despite numerous proposals for effective methods, a substantial memory overhead remains for gradient computations during updates. \thm{Can we fine-tune a series of task-specific small models and transfer their knowledge directly to a much larger model without additional training?} In this paper, we explore weak-to-strong specialization using logit arithmetic, facilitating a direct answer to this question. Existing weak-to-strong methods often employ a static knowledge transfer ratio and a single small model for transferring complex knowledge, which leads to suboptimal performance. % To address this, To surmount these limitations, we propose a dynamic logit fusion approach that works with a series of task-specific small models, each specialized in a different task. This method adaptively allocates weights among these models at each decoding step, learning the weights through Kullback-Leibler divergence constrained optimization problems. We conduct extensive experiments across various benchmarks in both single-task and multi-task settings, achieving leading results. By transferring expertise from the 7B model to the 13B model, our method closes the performance gap by 96.4\% in single-task scenarios and by 86.3\% in multi-task scenarios compared to full fine-tuning of the 13B model. Notably, we achieve surpassing performance on unseen tasks. Moreover, we further demonstrate that our method can effortlessly integrate in-context learning for single tasks and task arithmetic for multi-task scenarios.
comment: Accepted by NeurIPS 2024
♻ ☆ SAGED: A Holistic Bias-Benchmarking Pipeline for Language Models with Customisable Fairness Calibration COLING 2025
The development of unbiased large language models is widely recognized as crucial, yet existing benchmarks fall short in detecting biases due to limited scope, contamination, and lack of a fairness baseline. SAGED(-Bias) is the first holistic benchmarking pipeline to address these problems. The pipeline encompasses five core stages: scraping materials, assembling benchmarks, generating responses, extracting numeric features, and diagnosing with disparity metrics. SAGED includes metrics for max disparity, such as impact ratio, and bias concentration, such as Max Z-scores. Noticing that assessment tool bias and contextual bias in prompts can distort evaluation, SAGED implements counterfactual branching and baseline calibration for mitigation. For demonstration, we use SAGED on G20 Countries with popular 8b-level models including Gemma2, Llama3.1, Mistral, and Qwen2. With sentiment analysis, we find that while Mistral and Qwen2 show lower max disparity and higher bias concentration than Gemma2 and Llama3.1, all models are notably biased against countries like Russia and (except for Qwen2) China. With further experiments to have models role-playing U.S. (vice-/former-) presidents, we see bias amplifies and shifts in heterogeneous directions. Moreover, we see Qwen2 and Mistral not engage in role-playing, while Llama3.1 and Gemma2 role-play Trump notably more intensively than Biden and Harris, indicating role-playing performance bias in these models.
comment: Submitted to COLING 2025 Main Conference
♻ ☆ Demonstration Organization via Logit Separability: Advancing In-Context Learning through Multiple Class-Related Words
Effective organization of in-context learning (ICL) demonstrations is key to improving the quality of large language model (LLM) responses. To create better demonstrations that guide LLM understanding, we introduce logit separability, a criterion to assess the clarity of both samples and class-related words at the logit level. This facilitates the optimization of sample and label selection, enhancing the precision of information provided in ICL demonstrations. Additionally, we find that incorporating multiple class-related words for each sample, rather than relying on a single class name, improves performance by offering a broader range of label information. Building on these insights, we propose LICL, a logit separability-based method that jointly organizes samples and integrates multiple class-related words into each sample-label pair. Evaluations across seven classification datasets show that this approach significantly improves ICL performance by providing clearer instructions and richer label information.
comment: 20 pages, 10 figures
♻ ☆ FlipGuard: Defending Preference Alignment against Update Regression with Constrained Optimization EMNLP 2024
Recent breakthroughs in preference alignment have significantly improved Large Language Models' ability to generate texts that align with human preferences and values. However, current alignment metrics typically emphasize the post-hoc overall improvement, while overlooking a critical aspect: regression, which refers to the backsliding on previously correctly-handled data after updates. This potential pitfall may arise from excessive fine-tuning on already well-aligned data, which subsequently leads to over-alignment and degeneration. To address this challenge, we propose FlipGuard, a constrained optimization approach to detect and mitigate update regression with focal attention. Specifically, FlipGuard identifies performance degradation using a customized reward characterization and strategically enforces a constraint to encourage conditional congruence with the pre-aligned model during training. Comprehensive experiments demonstrate that FlipGuard effectively alleviates update regression while demonstrating excellent overall performance, with the added benefit of knowledge preservation while aligning preferences.
comment: Accepted by EMNLP 2024 Main track
♻ ☆ NoiseBench: Benchmarking the Impact of Real Label Noise on Named Entity Recognition EMNLP2024
Available training data for named entity recognition (NER) often contains a significant percentage of incorrect labels for entity types and entity boundaries. Such label noise poses challenges for supervised learning and may significantly deteriorate model quality. To address this, prior work proposed various noise-robust learning approaches capable of learning from data with partially incorrect labels. These approaches are typically evaluated using simulated noise where the labels in a clean dataset are automatically corrupted. However, as we show in this paper, this leads to unrealistic noise that is far easier to handle than real noise caused by human error or semi-automatic annotation. To enable the study of the impact of various types of real noise, we introduce NoiseBench, an NER benchmark consisting of clean training data corrupted with 6 types of real noise, including expert errors, crowdsourcing errors, automatic annotation errors and LLM errors. We present an analysis that shows that real noise is significantly more challenging than simulated noise, and show that current state-of-the-art models for noise-robust learning fall far short of their theoretically achievable upper bound. We release NoiseBench to the research community.
comment: data available at https://github.com/elenamer/NoiseBench; to appear at EMNLP2024 main conference
♻ ☆ Polyrating: A Cost-Effective and Bias-Aware Rating System for LLM Evaluation
Rating-based human evaluation has become an essential tool to accurately evaluate the impressive performance of large language models (LLMs). However, current rating systems suffer from several important limitations: first, they fail to account for biases that significantly influence evaluation results, second, they require large and expensive preference datasets to obtain accurate ratings, and third, they do not facilitate meaningful comparisons of model ratings across different tasks. To address these issues, we introduce Polyrating, an expressive and flexible rating system based on maximum a posteriori estimation that enables a more nuanced and thorough analysis of model performance at lower costs. Polyrating can detect and quantify biases affecting human preferences, ensuring fairer model comparisons. Further, Polyrating can reduce the cost of human evaluations by up to $41\%$ for new models and up to $77\%$ for new tasks by leveraging existing benchmark scores. Lastly, Polyrating enables direct comparisons of ratings across different tasks, providing a comprehensive understanding of an LLMs' strengths, weaknesses, and relative performance across different applications.
♻ ☆ Embedded Named Entity Recognition using Probing Classifiers EMNLP 2024
Streaming text generation has become a common way of increasing the responsiveness of language model powered applications, such as chat assistants. At the same time, extracting semantic information from generated text is a useful tool for applications such as automated fact checking or retrieval augmented generation. Currently, this requires either separate models during inference, which increases computational cost, or destructive fine-tuning of the language model. Instead, we propose an approach called EMBER which enables streaming named entity recognition in decoder-only language models without fine-tuning them and while incurring minimal additional computational cost at inference time. Specifically, our experiments show that EMBER maintains high token generation rates, with only a negligible decrease in speed of around 1% compared to a 43.64% slowdown measured for a baseline. We make our code and data available online, including a toolkit for training, testing, and deploying efficient token classification models optimized for streaming text generation.
comment: EMNLP 2024 (main)
♻ ☆ Hopping Too Late: Exploring the Limitations of Large Language Models on Multi-Hop Queries EMNLP 2024
Large language models (LLMs) can solve complex multi-step problems, but little is known about how these computations are implemented internally. Motivated by this, we study how LLMs answer multi-hop queries such as "The spouse of the performer of Imagine is". These queries require two information extraction steps: a latent one for resolving the first hop ("the performer of Imagine") into the bridge entity (John Lennon), and another for resolving the second hop ("the spouse of John Lennon") into the target entity (Yoko Ono). Understanding how the latent step is computed internally is key to understanding the overall computation. By carefully analyzing the internal computations of transformer-based LLMs, we discover that the bridge entity is resolved in the early layers of the model. Then, only after this resolution, the two-hop query is solved in the later layers. Because the second hop commences in later layers, there could be cases where these layers no longer encode the necessary knowledge for correctly predicting the answer. Motivated by this, we propose a novel "back-patching" analysis method whereby a hidden representation from a later layer is patched back to an earlier layer. We find that in up to 66% of previously incorrect cases there exists a back-patch that results in the correct generation of the answer, showing that the later layers indeed sometimes lack the needed functionality. Overall, our methods and findings open further opportunities for understanding and improving latent reasoning in transformer-based LLMs.
comment: Accepted at EMNLP 2024
♻ ☆ EmphAssess : a Prosodic Benchmark on Assessing Emphasis Transfer in Speech-to-Speech Models EMNLP 2024
We introduce EmphAssess, a prosodic benchmark designed to evaluate the capability of speech-to-speech models to encode and reproduce prosodic emphasis. We apply this to two tasks: speech resynthesis and speech-to-speech translation. In both cases, the benchmark evaluates the ability of the model to encode emphasis in the speech input and accurately reproduce it in the output, potentially across a change of speaker and language. As part of the evaluation pipeline, we introduce EmphaClass, a new model that classifies emphasis at the frame or word level.
comment: Accepted at EMNLP 2024 (Main)
♻ ☆ Learning Video Temporal Dynamics with Cross-Modal Attention for Robust Audio-Visual Speech Recognition
Audio-visual speech recognition (AVSR) aims to transcribe human speech using both audio and video modalities. In practical environments with noise-corrupted audio, the role of video information becomes crucial. However, prior works have primarily focused on enhancing audio features in AVSR, overlooking the importance of video features. In this study, we strengthen the video features by learning three temporal dynamics in video data: context order, playback direction, and the speed of video frames. Cross-modal attention modules are introduced to enrich video features with audio information so that speech variability can be taken into account when training on the video temporal dynamics. Based on our approach, we achieve the state-of-the-art performance on the LRS2 and LRS3 AVSR benchmarks for the noise-dominant settings. Our approach excels in scenarios especially for babble and speech noise, indicating the ability to distinguish the speech signal that should be recognized from lip movements in the video modality. We support the validity of our methodology by offering the ablation experiments for the temporal dynamics losses and the cross-modal attention architecture design.
comment: Accepted at SLT 2024 Main Conference; Code is available at https://github.com/sungnyun/avsr-temporal-dynamics
♻ ☆ Buffer of Thoughts: Thought-Augmented Reasoning with Large Language Models NeurIPS 2024
We introduce Buffer of Thoughts (BoT), a novel and versatile thought-augmented reasoning approach for enhancing accuracy, efficiency and robustness of large language models (LLMs). Specifically, we propose meta-buffer to store a series of informative high-level thoughts, namely thought-template, distilled from the problem-solving processes across various tasks. Then for each problem, we retrieve a relevant thought-template and adaptively instantiate it with specific reasoning structures to conduct efficient reasoning. To guarantee the scalability and stability, we further propose buffer-manager to dynamically update the meta-buffer, thus enhancing the capacity of meta-buffer as more tasks are solved. We conduct extensive experiments on 10 challenging reasoning-intensive tasks, and achieve significant performance improvements over previous SOTA methods: 11% on Game of 24, 20% on Geometric Shapes and 51% on Checkmate-in-One. Further analysis demonstrate the superior generalization ability and model robustness of our BoT, while requiring only 12% of the cost of multi-query prompting methods (e.g., tree/graph of thoughts) on average. Notably, we find that our Llama3-8B+BoT has the potential to surpass Llama3-70B model. Our project is available at: https://github.com/YangLing0818/buffer-of-thought-llm
comment: NeurIPS 2024 Spotlight. Project: https://github.com/YangLing0818/buffer-of-thought-llm
♻ ☆ Bridging Local Details and Global Context in Text-Attributed Graphs EMNLP 2024
Representation learning on text-attributed graphs (TAGs) is vital for real-world applications, as they combine semantic textual and contextual structural information. Research in this field generally consist of two main perspectives: local-level encoding and global-level aggregating, respectively refer to textual node information unification (e.g., using Language Models) and structure-augmented modeling (e.g., using Graph Neural Networks). Most existing works focus on combining different information levels but overlook the interconnections, i.e., the contextual textual information among nodes, which provides semantic insights to bridge local and global levels. In this paper, we propose GraphBridge, a multi-granularity integration framework that bridges local and global perspectives by leveraging contextual textual information, enhancing fine-grained understanding of TAGs. Besides, to tackle scalability and efficiency challenges, we introduce a graphaware token reduction module. Extensive experiments across various models and datasets show that our method achieves state-of-theart performance, while our graph-aware token reduction module significantly enhances efficiency and solves scalability issues.
comment: Accepted by EMNLP 2024(Main)
♻ ☆ Packing Analysis: Packing Is More Appropriate for Large Models or Datasets in Supervised Fine-tuning
Packing, initially utilized in the pre-training phase, is an optimization technique designed to maximize hardware resource efficiency by combining different training sequences to fit the model's maximum input length. Although it has demonstrated effectiveness during pre-training, there remains a lack of comprehensive analysis for the supervised fine-tuning (SFT) stage on the following points: (1) whether packing can effectively enhance training efficiency while maintaining performance, (2) the suitable size of the model and dataset for fine-tuning with the packing method, and (3) whether packing unrelated or related training samples might cause the model to either excessively disregard or over-rely on the context. In this paper, we perform extensive comparisons between SFT methods using padding and packing, covering SFT datasets ranging from 69K to 1.2M and models from 8B to 70B. This provides the first comprehensive analysis of the advantages and limitations of packing versus padding, as well as practical considerations for implementing packing in various training scenarios. Our analysis covers various benchmarks, including knowledge, reasoning, and coding, as well as GPT-based evaluations, time efficiency, and other fine-tuning parameters. We also open-source our code for fine-tuning and evaluation and provide checkpoints fine-tuned on datasets of different sizes, aiming to advance future research on packing methods. Code is available at: https://github.com/ShuheWang1998/Packing-Analysis?tab=readme-ov-file.
♻ ☆ FC-KAN: Function Combinations in Kolmogorov-Arnold Networks
In this paper, we introduce FC-KAN, a Kolmogorov-Arnold Network (KAN) that leverages combinations of popular mathematical functions such as B-splines, wavelets, and radial basis functions on low-dimensional data through element-wise operations. We explore several methods for combining the outputs of these functions, including sum, element-wise product, the addition of sum and element-wise product, quadratic function representation, and concatenation. In our experiments, we compare FC-KAN with multi-layer perceptron network (MLP) and other existing KANs, such as BSRBF-KAN, EfficientKAN, FastKAN, and FasterKAN, on the MNIST and Fashion-MNIST datasets. A variant of FC-KAN, which uses a combination of outputs from B-splines and Difference of Gaussians (DoG) in the form of a quadratic function, outperformed all other models on the average of 5 independent training runs. We expect that FC-KAN can leverage function combinations to design future KANs. Our repository is publicly available at: https://github.com/hoangthangta/FC_KAN.
comment: 15 pages
♻ ☆ Improving Non-autoregressive Translation Quality with Pretrained Language Model, Embedding Distillation and Upsampling Strategy for CTC
Non-autoregressive approaches aim to improve the inference speed of translation models, particularly those that generate output in a one-pass forward manner. However, these approaches often suffer from a significant drop in translation quality compared to autoregressive models. This paper introduces a series of innovative techniques to enhance the translation quality of Non-Autoregressive Translation (NAT) models while maintaining a substantial acceleration in inference speed. We propose fine-tuning Pretrained Multilingual Language Models (PMLMs) with the CTC loss to train NAT models effectively. Furthermore, we adopt the MASK insertion scheme for up-sampling instead of token duplication, and we present an embedding distillation method to further enhance performance. In our experiments, our model outperforms the baseline autoregressive model (Transformer \textit{base}) on multiple datasets, including WMT'14 DE$\leftrightarrow$EN, WMT'16 RO$\leftrightarrow$EN, and IWSLT'14 DE$\leftrightarrow$EN. Notably, our model achieves better performance than the baseline autoregressive model on the IWSLT'14 En$\leftrightarrow$De and WMT'16 En$\leftrightarrow$Ro datasets, even without using distillation data during training. It is worth highlighting that on the IWSLT'14 DE$\rightarrow$EN dataset, our model achieves an impressive BLEU score of 39.59, setting a new state-of-the-art performance. Additionally, our model exhibits a remarkable speed improvement of 16.35 times compared to the autoregressive model.
comment: 12 pages, 6 figures
♻ ☆ $\textbf{Only-IF}$:Revealing the Decisive Effect of Instruction Diversity on Generalization
Understanding and accurately following instructions is critical for large language models (LLMs) to be effective across diverse tasks. In this work, we rigorously examine the key factors that enable models to generalize to unseen instructions, providing insights to guide the collection of data for instruction-tuning. Through controlled experiments, inspired by the Turing-complete Markov algorithm, we demonstrate that such generalization $\textbf{only emerges}$ when training data is diversified enough across semantic domains. Our findings also reveal that merely diversifying within limited domains fails to ensure robust generalization. In contrast, cross-domain data diversification, even under constrained data budgets, significantly enhances a model's adaptability. We further extend our analysis to real-world scenarios, including fine-tuning of $\textit{$\textbf{specialist}$}$ and $\textit{$\textbf{generalist}$}$ models. In both cases, we demonstrate that 1) better performance can be achieved by increasing the diversity of an established dataset while keeping the data size constant, and 2) when scaling up the data, diversifying the semantics of instructions is more effective than simply increasing the quantity of similar data. Our research provides important insights for dataset collation, particularly when optimizing model performance by expanding training data for both specialist and generalist scenarios. We show that careful consideration of data diversification is key: training specialist models with data extending beyond their core domain leads to significant performance improvements, while generalist models benefit from diverse data mixtures that enhance their overall instruction-following capabilities across a wide range of applications. Our results highlight the critical role of strategic diversification and offer clear guidelines for improving data quality.
comment: Template change
♻ ☆ Diff-eRank: A Novel Rank-Based Metric for Evaluating Large Language Models NeurIPS 2024
Large Language Models (LLMs) have transformed natural language processing and extended their powerful capabilities to multi-modal domains. As LLMs continue to advance, it is crucial to develop diverse and appropriate metrics for their evaluation. In this paper, we introduce a novel rank-based metric, Diff-eRank, grounded in information theory and geometry principles. Diff-eRank assesses LLMs by analyzing their hidden representations, providing a quantitative measure of how efficiently they eliminate redundant information during training. We demonstrate the applicability of Diff-eRank in both single-modal (e.g., language) and multi-modal settings. For language models, our results show that Diff-eRank increases with model size and correlates well with conventional metrics such as loss and accuracy. In the multi-modal context, we propose an alignment evaluation method based on the eRank, and verify that contemporary multi-modal LLMs exhibit strong alignment performance based on our method. Our code is publicly available at https://github.com/waltonfuture/Diff-eRank.
comment: Accepted by NeurIPS 2024
♻ ☆ Transformer Block Coupling and its Correlation with Generalization in LLMs
Large Language Models (LLMs) have made significant strides in natural language processing, and a precise understanding of the internal mechanisms driving their success is essential. In this work, we trace the trajectories of individual tokens as they pass through transformer blocks, and linearize the system along these trajectories through their Jacobian matrices. By examining the relationships between these Jacobians, we uncover a $\textbf{transformer block coupling}$ phenomenon in a variety of LLMs, characterized by the coupling of their top singular vectors across tokens and depth. Our findings reveal that coupling $\textit{positively correlates}$ with model performance, and that this relationship is stronger than with other hyperparameters, namely parameter budget, model depth, and embedding dimension. We further investigate the emergence of these properties through training, noting the development of coupling, as well as an increase in linearity and layer-wise exponential growth in the token trajectories. These collective insights provide a novel perspective on the interactions between token embeddings, and prompt further approaches to study training and generalization in LLMs.
♻ ☆ Vocabulary-Defined Semantics: Latent Space Clustering for Improving In-Context Learning
In-context learning enables language models (LM) to adapt to downstream data or tasks by incorporating few samples as demonstrations within the prompts. It offers strong performance without the expense of fine-tuning. However, the performance of in-context learning can be unstable depending on the quality, format, or order of demonstrations, which in turn exacerbates the difficulty of optimization. Prior work, such as Knn Prompting, index samples based on the similarities of logits at the output-side, in addition to the regular retrieval operation at the input-side. They improve in-context learning by leveraging the core ability of next-token prediction, rather than relying solely on the emergent capacity to make analogies. Despite this, the hard-to-optimize issue of in-context learning still exists. In our view, it stems from the process of selecting demonstrations. To address this, we propose complementing in-context learning with an additional clustering operation. We propose a novel approach "vocabulary-defined semantics". Grounded in LM vocabulary, which is the label space of model outputs, the proposed approach computes semantically equivalent latent representations for output labels. Then, taking the representations as centroids, a clustering operation is performed to align the semantic properties between the language model and the downstream data/tasks. Based on extensive experiments across diverse textual understanding datasets and multiple models, our approach outperforms the state-of-the-art in terms of effectiveness and efficiency. On average, it achieves $3\%-49\%$ improvements while requiring only half of the computation time.
comment: under peer-review
♻ ☆ Twin-Merging: Dynamic Integration of Modular Expertise in Model Merging NeurIPS 2024
In the era of large language models, model merging is a promising way to combine multiple task-specific models into a single multitask model without extra training. However, two challenges remain: (a) interference between different models and (b) heterogeneous data during testing. Traditional model merging methods often show significant performance gaps compared to fine-tuned models due to these issues. Additionally, a one-size-fits-all model lacks flexibility for diverse test data, leading to performance degradation. We show that both shared and exclusive task-specific knowledge are crucial for merging performance, but directly merging exclusive knowledge hinders overall performance. In view of this, we propose Twin-Merging, a method that encompasses two principal stages: (1) modularizing knowledge into shared and exclusive components, with compression to reduce redundancy and enhance efficiency; (2) dynamically merging shared and task-specific knowledge based on the input. This approach narrows the performance gap between merged and fine-tuned models and improves adaptability to heterogeneous data. Extensive experiments on $20$ datasets for both language and vision tasks demonstrate the effectiveness of our method, showing an average improvement of $28.34\%$ in absolute normalized score for discriminative tasks and even surpassing the fine-tuned upper bound on the generative tasks. Our implementation is available in \url{https://github.com/LZY-the-boys/Twin-Merging}
comment: NeurIPS 2024 poster
♻ ☆ Tensor Attention Training: Provably Efficient Learning of Higher-order Transformers
Tensor Attention, a multi-view attention that is able to capture high-order correlations among multiple modalities, can overcome the representational limitations of classical matrix attention. However, the $O(n^3)$ time complexity of tensor attention poses a significant obstacle to its utilization in transformers, where $n$ is the input sequence length. In this work, we prove that the backward gradient of tensor attention training can be computed in almost linear time $n^{1+o(1)}$, the same complexity as its forward computation under the bounded entries assumption. We provide a closed-form solution for the gradient and propose a fast computation method utilizing polynomial approximation methods and tensor algebraic techniques. Furthermore, we prove the necessity and tightness of our assumption through hardness analysis, showing that slightly weakening it renders the gradient problem unsolvable in truly subcubic time. Our theoretical results establish the feasibility of efficient higher-order transformer training and may facilitate practical applications of tensor attention architectures.
♻ ☆ Hyper-multi-step: The Truth Behind Difficult Long-context Tasks
Long-context language models (LCLM), characterized by their extensive context window, is becoming increasingly popular. Meanwhile, many long-context benchmarks present challenging tasks that even the most advanced LCLMs struggle to complete. However, the underlying sources of various challenging long-context tasks have seldom been studied. To bridge this gap, we conduct experiments to indicate their difficulty stems primarily from two basic issues: "multi-matching retrieval," which requires the simultaneous retrieval of multiple items, and "logic-based retrieval," which necessitates logical judgment within retrieval criteria. These two problems, while seemingly straightforward, actually exceed the capabilities of LCLMs because they are proven to be hyper-multi-step (demanding numerous steps to solve) in nature. This finding could explain why LLMs struggle with more advanced long-context tasks, providing a more accurate perspective for rethinking solutions for them.
comment: Our code and datasets are publicly available at https://github.com/yuyijiong/hard_retrieval_for_llm
♻ ☆ Process-Driven Autoformalization in Lean 4
Autoformalization, the conversion of natural language mathematics into formal languages, offers significant potential for advancing mathematical reasoning. However, existing efforts are limited to formal languages with substantial online corpora and struggle to keep pace with rapidly evolving languages like Lean 4. To bridge this gap, we propose a new benchmark \textbf{Form}alization for \textbf{L}ean~\textbf{4} (\textbf{\name}) designed to evaluate the autoformalization capabilities of large language models (LLMs). This benchmark encompasses a comprehensive assessment of questions, answers, formal statements, and proofs. Additionally, we introduce a \textbf{P}rocess-\textbf{S}upervised \textbf{V}erifier (\textbf{PSV}) model that leverages the precise feedback from Lean 4 compilers to enhance autoformalization. Our experiments demonstrate that the PSV method improves autoformalization, enabling higher accuracy using less filtered training data. Furthermore, when fine-tuned with data containing detailed process information, PSV can leverage the data more effectively, leading to more significant improvements in autoformalization for Lean 4. Our dataset and code are available at \url{https://github.com/rookie-joe/PDA}.
comment: 32 pages, 1 figures, 15 tables
♻ ☆ Connecting the Dots: Evaluating Abstract Reasoning Capabilities of LLMs Using the New York Times Connections Word Game
The New York Times Connections game has emerged as a popular and challenging pursuit for word puzzle enthusiasts. We collect 438 Connections games to evaluate the performance of state-of-the-art large language models (LLMs) against expert and novice human players. Our results show that even the best performing LLM, Claude 3.5 Sonnet, which has otherwise shown impressive reasoning abilities on a wide variety of benchmarks, can only fully solve 18% of the games. Novice and expert players perform better than Claude 3.5 Sonnet, with expert human players significantly outperforming it. We create a taxonomy of the knowledge types required to successfully cluster and categorize words in the Connections game. We find that while LLMs perform relatively well on categorizing words based on semantic relations they struggle with other types of knowledge such as Encyclopedic Knowledge, Multiword Expressions or knowledge that combines both Word Form and Meaning. Our results establish the New York Times Connections game as a challenging benchmark for evaluating abstract reasoning capabilities in AI systems.
♻ ☆ LARA: Linguistic-Adaptive Retrieval-Augmentation for Multi-Turn Intent Classification EMNLP'24
Multi-turn intent classification is notably challenging due to the complexity and evolving nature of conversational contexts. This paper introduces LARA, a Linguistic-Adaptive Retrieval-Augmentation framework to enhance accuracy in multi-turn classification tasks across six languages, accommodating a large number of intents in chatbot interactions. LARA combines a fine-tuned smaller model with a retrieval-augmented mechanism, integrated within the architecture of LLMs. The integration allows LARA to dynamically utilize past dialogues and relevant intents, thereby improving the understanding of the context. Furthermore, our adaptive retrieval techniques bolster the cross-lingual capabilities of LLMs without extensive retraining and fine-tuning. Comprehensive experiments demonstrate that LARA achieves state-of-the-art performance on multi-turn intent classification tasks, enhancing the average accuracy by 3.67\% from state-of-the-art single-turn intent classifiers.
comment: Accepted to EMNLP'24
♻ ☆ PSST: A Benchmark for Evaluation-driven Text Public-Speaking Style Transfer EMNLP 2024
Language style is necessary for AI systems to understand and generate diverse human language accurately. However, previous text style transfer primarily focused on sentence-level data-driven approaches, limiting exploration of potential problems in large language models (LLMs) and the ability to meet complex application needs. To overcome these limitations, we introduce a novel task called Public-Speaking Style Transfer (PSST), which aims to simulate humans to transform passage-level, official texts into a public-speaking style. Grounded in the analysis of real-world data from a linguistic perspective, we decompose public-speaking style into key sub-styles to pose challenges and quantify the style modeling capability of LLMs. For such intricate text style transfer, we further propose a fine-grained evaluation framework to analyze the characteristics and identify the problems of stylized texts. Comprehensive experiments suggest that current LLMs struggle to generate public speaking texts that align with human preferences, primarily due to excessive stylization and loss of semantic information.
comment: EMNLP 2024 Findings
♻ ☆ From Artificial Needles to Real Haystacks: Improving Retrieval Capabilities in LLMs by Finetuning on Synthetic Data
Recent studies have shown that Large Language Models (LLMs) struggle to accurately retrieve information and maintain reasoning capabilities when processing long-context inputs. To address these limitations, we propose a finetuning approach utilizing a carefully designed synthetic dataset comprising numerical key-value retrieval tasks. Our experiments on models like GPT-3.5 Turbo and Mistral 7B demonstrate that finetuning LLMs on this dataset significantly improves LLMs' information retrieval and reasoning capabilities in longer-context settings. We present an analysis of the finetuned models, illustrating the transfer of skills from synthetic to real task evaluations (e.g., $10.5\%$ improvement on $20$ documents MDQA at position $10$ for GPT-3.5 Turbo). We also find that finetuned LLMs' performance on general benchmarks remains almost constant while LLMs finetuned on other baseline long-context augmentation data can encourage hallucination (e.g., on TriviaQA, Mistral 7B finetuned on our synthetic data cause no performance drop while other baseline data can cause a drop that ranges from $2.33\%$ to $6.19\%$). Our study highlights the potential of finetuning on synthetic data for improving the performance of LLMs on longer-context tasks.
♻ ☆ EMS-SD: Efficient Multi-sample Speculative Decoding for Accelerating Large Language Models
Speculative decoding emerges as a pivotal technique for enhancing the inference speed of Large Language Models (LLMs). Despite recent research aiming to improve prediction efficiency, multi-sample speculative decoding has been overlooked due to varying numbers of accepted tokens within a batch in the verification phase. Vanilla method adds padding tokens in order to ensure that the number of new tokens remains consistent across samples. However, this increases the computational and memory access overhead, thereby reducing the speedup ratio. We propose a novel method that can resolve the issue of inconsistent tokens accepted by different samples without necessitating an increase in memory or computing overhead. Furthermore, our proposed method can handle the situation where the prediction tokens of different samples are inconsistent without the need to add padding tokens. Sufficient experiments demonstrate the efficacy of our method. Our code is available at https://github.com/niyunsheng/EMS-SD.
♻ ☆ Unintentional Unalignment: Likelihood Displacement in Direct Preference Optimization
Direct Preference Optimization (DPO) and its variants are increasingly used for aligning language models with human preferences. Although these methods are designed to teach a model to generate preferred responses more frequently relative to dispreferred responses, prior work has observed that the likelihood of preferred responses often decreases during training. The current work sheds light on the causes and implications of this counter-intuitive phenomenon, which we term likelihood displacement. We demonstrate that likelihood displacement can be catastrophic, shifting probability mass from preferred responses to responses with an opposite meaning. As a simple example, training a model to prefer $\texttt{No}$ over $\texttt{Never}$ can sharply increase the probability of $\texttt{Yes}$. Moreover, when aligning the model to refuse unsafe prompts, we show that such displacement can unintentionally lead to unalignment, by shifting probability mass from preferred refusal responses to harmful responses (e.g., reducing the refusal rate of Llama-3-8B-Instruct from 74.4% to 33.4%). We theoretically characterize that likelihood displacement is driven by preferences that induce similar embeddings, as measured by a centered hidden embedding similarity (CHES) score. Empirically, the CHES score enables identifying which training samples contribute most to likelihood displacement in a given dataset. Filtering out these samples effectively mitigated unintentional unalignment in our experiments. More broadly, our results highlight the importance of curating data with sufficiently distinct preferences, for which we believe the CHES score may prove valuable.
comment: Code available at https://github.com/princeton-nlp/unintentional-unalignment
♻ ☆ Inadequacies of Large Language Model Benchmarks in the Era of Generative Artificial Intelligence
The rapid rise in popularity of Large Language Models (LLMs) with emerging capabilities has spurred public curiosity to evaluate and compare different LLMs, leading many researchers to propose their own LLM benchmarks. Noticing preliminary inadequacies in those benchmarks, we embarked on a study to critically assess 23 state-of-the-art LLM benchmarks, using our novel unified evaluation framework through the lenses of people, process, and technology, under the pillars of benchmark functionality and integrity. Our research uncovered significant limitations, including biases, difficulties in measuring genuine reasoning, adaptability, implementation inconsistencies, prompt engineering complexity, evaluator diversity, and the overlooking of cultural and ideological norms in one comprehensive assessment. Our discussions emphasized the urgent need for standardized methodologies, regulatory certainties, and ethical guidelines in light of Artificial Intelligence (AI) advancements, including advocating for an evolution from static benchmarks to dynamic behavioral profiling to accurately capture LLMs' complex behaviors and potential risks. Our study highlighted the necessity for a paradigm shift in LLM evaluation methodologies, underlining the importance of collaborative efforts for the development of universally accepted benchmarks and the enhancement of AI systems' integration into society.
♻ ☆ Instruction Fine-Tuning: Does Prompt Loss Matter? EMNLP 2024
We present a novel study analyzing the effects of various prompt loss token weights (PLW) for supervised instruction fine-tuning (SIFT). While prompt-masking (PLW = 0) is common for SIFT, some fine-tuning APIs support fractional PLWs and suggest that using a small non-zero PLW can help stabilize learning when fine-tuning on short-completion data. However, there has never been a study confirming this claim, and OpenAI, a major cloud-based SIFT provider, recently removed this parameter from their fine-tuning API. We found that performance of models fine-tuned on short-completion data had a statistically-significant negative quadratic relationship with PLW. Using small values (0.01 - 0.5) of PLW produced better results on multiple-choice and short-generation benchmarks (outperforming models fine-tuned on long-completion data) while large values (~ 1.0) of PLW produced better results on long-generation benchmarks. We explained this effect and verified its importance through additional experiments. This research serves as a warning to API providers about the importance of providing a PLW parameter for SIFT.
comment: EMNLP 2024: Camera-ready version
♻ ☆ Impact of Large Language Model Assistance on Patients Reading Clinical Notes: A Mixed-Methods Study
Large language models (LLMs) have immense potential to make information more accessible, particularly in medicine, where complex medical jargon can hinder patient comprehension of clinical notes. We developed a patient-facing tool using LLMs to make clinical notes more readable by simplifying, extracting information from, and adding context to the notes. We piloted the tool with clinical notes donated by patients with a history of breast cancer and synthetic notes from a clinician. Participants (N=200, healthy, female-identifying patients) were randomly assigned three clinical notes in our tool with varying levels of augmentations and answered quantitative and qualitative questions evaluating their understanding of follow-up actions. Augmentations significantly increased their quantitative understanding scores. In-depth interviews were conducted with participants (N=7, patients with a history of breast cancer), revealing both positive sentiments about the augmentations and concerns about AI. We also performed a qualitative clinician-driven analysis of the model's error modes.
♻ ☆ Words in Motion: Extracting Interpretable Control Vectors for Motion Transformers NeurIPS'24
Transformer-based models generate hidden states that are difficult to interpret. In this work, we aim to interpret these hidden states and control them at inference, with a focus on motion forecasting. We leverage the phenomenon of neural collapse and use linear probes to measure interpretable features in hidden states. Our experiments reveal meaningful directions and distances between hidden states of opposing features, which we use to fit control vectors for activation steering. Consequently, our method enables controlling transformer-based motion forecasting models with interpretable features, providing a unique interface to interact with and understand these models. Our implementation is available at https://github.com/kit-mrt/future-motion
comment: NeurIPS'24 workshop camera-ready
♻ ☆ Defining Boundaries: A Spectrum of Task Feasibility for Large Language Models
Large language models (LLMs) have shown remarkable performance in various tasks but often fail to handle queries that exceed their knowledge and capabilities, leading to incorrect or fabricated responses. This paper addresses the need for LLMs to recognize and refuse infeasible tasks due to the required skills surpassing their capabilities. We first conceptualize infeasible tasks for LLMs and provide categorizations that cover a spectrum of related hallucinations over existing literature. We develop and benchmark a new dataset comprising diverse infeasible and feasible tasks to evaluate multiple LLMs' abilities to reject infeasible tasks. Furthermore, we explore the potential of increasing LLMs' refusal capabilities with fine-tuning. Experiments validate the effectiveness of our trained models, offering promising directions for refining the operational boundaries of LLMs in real applications.
comment: 19 pages, 7 tables, 16 Figures
♻ ☆ RoRA-VLM: Robust Retrieval-Augmented Vision Language Models
Current vision-language models (VLMs) still exhibit inferior performance on knowledge-intensive tasks, primarily due to the challenge of accurately encoding all the associations between visual objects and scenes to their corresponding entities and background knowledge. While retrieval augmentation methods offer an efficient way to integrate external knowledge, extending them to vision-language domain presents unique challenges in (1) precisely retrieving relevant information from external sources due to the inherent discrepancy within the multimodal queries, and (2) being resilient to the irrelevant, extraneous and noisy information contained in the retrieved multimodal knowledge snippets. In this work, we introduce RORA-VLM, a novel and robust retrieval augmentation framework specifically tailored for VLMs, with two key innovations: (1) a 2-stage retrieval process with image-anchored textual-query expansion to synergistically combine the visual and textual information in the query and retrieve the most relevant multimodal knowledge snippets; and (2) a robust retrieval augmentation method that strengthens the resilience of VLMs against irrelevant information in the retrieved multimodal knowledge by injecting adversarial noises into the retrieval-augmented training process, and filters out extraneous visual information, such as unrelated entities presented in images, via a query-oriented visual token refinement strategy. We conduct extensive experiments to validate the effectiveness and robustness of our proposed methods on three widely adopted benchmark datasets. Our results demonstrate that with a minimal amount of training instance, RORA-VLM enables the base model to achieve significant performance improvement and constantly outperform state-of-the-art retrieval-augmented VLMs on all benchmarks while also exhibiting a novel zero-shot domain transfer capability.
♻ ☆ Group Preference Optimization: Few-Shot Alignment of Large Language Models ICLR 2024
Many applications of large language models (LLMs), ranging from chatbots to creative writing, require nuanced subjective judgments that can differ significantly across different groups. Existing alignment algorithms can be expensive to align for each group, requiring prohibitive amounts of group-specific preference data and computation for real-world use cases. We introduce Group Preference Optimization (GPO), an alignment framework that steers language models to preferences of individual groups in a few-shot manner. In GPO, we augment the base LLM with an independent transformer module trained to predict the preferences of a group for the LLM generations. For few-shot learning, we parameterize this module as an in-context autoregressive transformer and train it via meta-learning on several groups. We empirically validate the efficacy of GPO through rigorous evaluations using LLMs with varied sizes on three human opinion adaptation tasks. These tasks involve adapting to the preferences of US demographic groups, global countries, and individual users. Our results demonstrate that GPO not only aligns models more accurately but also requires fewer group-specific preferences, and less training and inference computing resources, outperforming existing strategies such as in-context steering and fine-tuning methods.
comment: accepted at ICLR 2024, code at https://github.com/jamqd/Group-Preference-Optimization
♻ ☆ Multi-step Problem Solving Through a Verifier: An Empirical Analysis on Model-induced Process Supervision
Process supervision, using a trained verifier to evaluate the intermediate steps generated by a reasoner, has demonstrated significant improvements in multi-step problem solving. In this paper, to avoid the expensive effort of human annotation on the verifier training data, we introduce Model-induced Process Supervision (MiPS), a novel method for automating data curation. MiPS annotates an intermediate step by sampling completions of this solution through the reasoning model, and obtaining an accuracy defined as the proportion of correct completions. Inaccuracies of the reasoner would cause MiPS underestimating the accuracy of intermediate steps, therefore, we suggest and empirically show that verification focusing on high predicted scores of the verifier shall be preferred over that of low predicted scores, contrary to prior observations on human curated data. Our approach significantly improves the performance of PaLM 2 on math and coding tasks (accuracy +0.67% on GSM8K, +4.16% on MATH, +0.92% on MBPP compared with an output supervision trained verifier). Additionally, our study demonstrates that the verifier exhibits strong generalization ability across different reasoning models.
♻ ☆ Weakly Supervised Veracity Classification with LLM-Predicted Credibility Signals
Credibility signals represent a wide range of heuristics typically used by journalists and fact-checkers to assess the veracity of online content. Automating the extraction of credibility signals presents significant challenges due to the necessity of training high-accuracy, signal-specific extractors, coupled with the lack of sufficiently large annotated datasets. This paper introduces Pastel (Prompted weAk Supervision wiTh crEdibility signaLs), a weakly supervised approach that leverages large language models (LLMs) to extract credibility signals from web content, and subsequently combines them to predict the veracity of content without relying on human supervision. We validate our approach using four article-level misinformation detection datasets, demonstrating that Pastel outperforms zero-shot veracity detection by 38.3% and achieves 86.7% of the performance of the state-of-the-art system trained with human supervision. Moreover, in cross-domain settings where training and testing datasets originate from different domains, Pastel significantly outperforms the state-of-the-art supervised model by 63%. We further study the association between credibility signals and veracity, and perform an ablation study showing the impact of each signal on model performance. Our findings reveal that 12 out of the 19 proposed signals exhibit strong associations with veracity across all datasets, while some signals show domain-specific strengths.
♻ ☆ From Crowdsourced Data to High-Quality Benchmarks: Arena-Hard and BenchBuilder Pipeline
The rapid evolution of Large Language Models (LLMs) has outpaced the development of model evaluation, highlighting the need for continuous curation of new, challenging benchmarks. However, manual curation of high-quality, human-aligned benchmarks is expensive and time-consuming. To address this, we introduce BenchBuilder, an automated pipeline that leverages LLMs to curate high-quality, open-ended prompts from large, crowd-sourced datasets, enabling continuous benchmark updates without human in the loop. We apply BenchBuilder to datasets such as Chatbot Arena and WildChat-1M, extracting challenging prompts and utilizing LLM-as-a-Judge for automatic model evaluation. To validate benchmark quality, we propose new metrics to measure a benchmark's alignment with human preferences and ability to separate models. We release Arena-Hard-Auto, a benchmark consisting 500 challenging prompts curated by BenchBuilder. Arena-Hard-Auto provides 3x higher separation of model performances compared to MT-Bench and achieves 98.6% correlation with human preference rankings, all at a cost of $20. Our work sets a new framework for the scalable curation of automated benchmarks from extensive data.
♻ ☆ Human-centered NLP Fact-checking: Co-Designing with Fact-checkers using Matchmaking for AI SC
While many Natural Language Processing (NLP) techniques have been proposed for fact-checking, both academic research and fact-checking organizations report limited adoption of such NLP work due to poor alignment with fact-checker practices, values, and needs. To address this, we investigate a co-design method, Matchmaking for AI, to enable fact-checkers, designers, and NLP researchers to collaboratively identify what fact-checker needs should be addressed by technology, and to brainstorm ideas for potential solutions. Co-design sessions we conducted with 22 professional fact-checkers yielded a set of 11 design ideas that offer a "north star", integrating fact-checker criteria into novel NLP design concepts. These concepts range from pre-bunking misinformation, efficient and personalized monitoring misinformation, proactively reducing fact-checker potential biases, and collaborative writing fact-check reports. Our work provides new insights into both human-centered fact-checking research and practice and AI co-design research.
comment: Accepted at CSCW 2024
Computer Vision and Pattern Recognition 201
☆ Tex4D: Zero-shot 4D Scene Texturing with Video Diffusion Models
3D meshes are widely used in computer vision and graphics for their efficiency in animation and minimal memory use, playing a crucial role in movies, games, AR, and VR. However, creating temporally consistent and realistic textures for mesh sequences remains labor-intensive for professional artists. On the other hand, while video diffusion models excel at text-driven video generation, they often lack 3D geometry awareness and struggle with achieving multi-view consistent texturing for 3D meshes. In this work, we present Tex4D, a zero-shot approach that integrates inherent 3D geometry knowledge from mesh sequences with the expressiveness of video diffusion models to produce multi-view and temporally consistent 4D textures. Given an untextured mesh sequence and a text prompt as inputs, our method enhances multi-view consistency by synchronizing the diffusion process across different views through latent aggregation in the UV space. To ensure temporal consistency, we leverage prior knowledge from a conditional video generation model for texture synthesis. However, straightforwardly combining the video diffusion model and the UV texture aggregation leads to blurry results. We analyze the underlying causes and propose a simple yet effective modification to the DDIM sampling process to address this issue. Additionally, we introduce a reference latent texture to strengthen the correlation between frames during the denoising process. To the best of our knowledge, Tex4D is the first method specifically designed for 4D scene texturing. Extensive experiments demonstrate its superiority in producing multi-view and multi-frame consistent videos based on untextured mesh sequences.
comment: Project page: https://tex4d.github.io/
☆ When Does Perceptual Alignment Benefit Vision Representations?
Humans judge perceptual similarity according to diverse visual attributes, including scene layout, subject location, and camera pose. Existing vision models understand a wide range of semantic abstractions but improperly weigh these attributes and thus make inferences misaligned with human perception. While vision representations have previously benefited from alignment in contexts like image generation, the utility of perceptually aligned representations in more general-purpose settings remains unclear. Here, we investigate how aligning vision model representations to human perceptual judgments impacts their usability across diverse computer vision tasks. We finetune state-of-the-art models on human similarity judgments for image triplets and evaluate them across standard vision benchmarks. We find that aligning models to perceptual judgments yields representations that improve upon the original backbones across many downstream tasks, including counting, segmentation, depth estimation, instance retrieval, and retrieval-augmented generation. In addition, we find that performance is widely preserved on other tasks, including specialized out-of-distribution domains such as in medical imaging and 3D environment frames. Our results suggest that injecting an inductive bias about human perceptual knowledge into vision models can contribute to better representations.
comment: S.S. and S.F. contributed equally. Website: percep-align.github.io
☆ TemporalBench: Benchmarking Fine-grained Temporal Understanding for Multimodal Video Models
Understanding fine-grained temporal dynamics is crucial for multimodal video comprehension and generation. Due to the lack of fine-grained temporal annotations, existing video benchmarks mostly resemble static image benchmarks and are incompetent at evaluating models for temporal understanding. In this paper, we introduce TemporalBench, a new benchmark dedicated to evaluating fine-grained temporal understanding in videos. TemporalBench consists of ~10K video question-answer pairs, derived from ~2K high-quality human annotations detailing the temporal dynamics in video clips. As a result, our benchmark provides a unique testbed for evaluating various temporal understanding and reasoning abilities such as action frequency, motion magnitude, event order, etc. Moreover, it enables evaluations on various tasks like both video question answering and captioning, both short and long video understanding, as well as different models such as multimodal video embedding models and text generation models. Results show that state-of-the-art models like GPT-4o achieve only 38.5% question answering accuracy on TemporalBench, demonstrating a significant gap (~30%) between humans and AI in temporal understanding. Furthermore, we notice a critical pitfall for multi-choice QA where LLMs can detect the subtle changes in negative captions and find a centralized description as a cue for its prediction, where we propose Multiple Binary Accuracy (MBA) to correct such bias. We hope that TemporalBench can foster research on improving models' temporal reasoning capabilities. Both dataset and evaluation code will be made available.
comment: Project Page: https://temporalbench.github.io/
☆ LVD-2M: A Long-take Video Dataset with Temporally Dense Captions NeurIPS 2024
The efficacy of video generation models heavily depends on the quality of their training datasets. Most previous video generation models are trained on short video clips, while recently there has been increasing interest in training long video generation models directly on longer videos. However, the lack of such high-quality long videos impedes the advancement of long video generation. To promote research in long video generation, we desire a new dataset with four key features essential for training long video generation models: (1) long videos covering at least 10 seconds, (2) long-take videos without cuts, (3) large motion and diverse contents, and (4) temporally dense captions. To achieve this, we introduce a new pipeline for selecting high-quality long-take videos and generating temporally dense captions. Specifically, we define a set of metrics to quantitatively assess video quality including scene cuts, dynamic degrees, and semantic-level quality, enabling us to filter high-quality long-take videos from a large amount of source videos. Subsequently, we develop a hierarchical video captioning pipeline to annotate long videos with temporally-dense captions. With this pipeline, we curate the first long-take video dataset, LVD-2M, comprising 2 million long-take videos, each covering more than 10 seconds and annotated with temporally dense captions. We further validate the effectiveness of LVD-2M by fine-tuning video generation models to generate long videos with dynamic motions. We believe our work will significantly contribute to future research in long video generation.
comment: NeurIPS 2024 Dataset and Benchmark Track. Project page: https://silentview.github.io/LVD-2M/ . Code: https://github.com/SilentView/LVD-2M
☆ Depth Any Video with Scalable Synthetic Data
Video depth estimation has long been hindered by the scarcity of consistent and scalable ground truth data, leading to inconsistent and unreliable results. In this paper, we introduce Depth Any Video, a model that tackles the challenge through two key innovations. First, we develop a scalable synthetic data pipeline, capturing real-time video depth data from diverse synthetic environments, yielding 40,000 video clips of 5-second duration, each with precise depth annotations. Second, we leverage the powerful priors of generative video diffusion models to handle real-world videos effectively, integrating advanced techniques such as rotary position encoding and flow matching to further enhance flexibility and efficiency. Unlike previous models, which are limited to fixed-length video sequences, our approach introduces a novel mixed-duration training strategy that handles videos of varying lengths and performs robustly across different frame rates-even on single frames. At inference, we propose a depth interpolation method that enables our model to infer high-resolution video depth across sequences of up to 150 frames. Our model outperforms all previous generative depth models in terms of spatial accuracy and temporal consistency.
comment: Project Page: https://depthanyvideo.github.io/
☆ HART: Efficient Visual Generation with Hybrid Autoregressive Transformer
We introduce Hybrid Autoregressive Transformer (HART), an autoregressive (AR) visual generation model capable of directly generating 1024x1024 images, rivaling diffusion models in image generation quality. Existing AR models face limitations due to the poor image reconstruction quality of their discrete tokenizers and the prohibitive training costs associated with generating 1024px images. To address these challenges, we present the hybrid tokenizer, which decomposes the continuous latents from the autoencoder into two components: discrete tokens representing the big picture and continuous tokens representing the residual components that cannot be represented by the discrete tokens. The discrete component is modeled by a scalable-resolution discrete AR model, while the continuous component is learned with a lightweight residual diffusion module with only 37M parameters. Compared with the discrete-only VAR tokenizer, our hybrid approach improves reconstruction FID from 2.11 to 0.30 on MJHQ-30K, leading to a 31% generation FID improvement from 7.85 to 5.38. HART also outperforms state-of-the-art diffusion models in both FID and CLIP score, with 4.5-7.7x higher throughput and 6.9-13.4x lower MACs. Our code is open sourced at https://github.com/mit-han-lab/hart.
comment: Demo: https://hart.mit.edu. The first two authors contributed equally to this work
☆ Deep Linear Probe Generators for Weight Space Learning
Weight space learning aims to extract information about a neural network, such as its training dataset or generalization error. Recent approaches learn directly from model weights, but this presents many challenges as weights are high-dimensional and include permutation symmetries between neurons. An alternative approach, Probing, represents a model by passing a set of learned inputs (probes) through the model, and training a predictor on top of the corresponding outputs. Although probing is typically not used as a stand alone approach, our preliminary experiment found that a vanilla probing baseline worked surprisingly well. However, we discover that current probe learning strategies are ineffective. We therefore propose Deep Linear Probe Generators (ProbeGen), a simple and effective modification to probing approaches. ProbeGen adds a shared generator module with a deep linear architecture, providing an inductive bias towards structured probes thus reducing overfitting. While simple, ProbeGen performs significantly better than the state-of-the-art and is very efficient, requiring between 30 to 1000 times fewer FLOPs than other top approaches.
☆ TrajDiffuse: A Conditional Diffusion Model for Environment-Aware Trajectory Prediction ICPR
Accurate prediction of human or vehicle trajectories with good diversity that captures their stochastic nature is an essential task for many applications. However, many trajectory prediction models produce unreasonable trajectory samples that focus on improving diversity or accuracy while neglecting other key requirements, such as collision avoidance with the surrounding environment. In this work, we propose TrajDiffuse, a planning-based trajectory prediction method using a novel guided conditional diffusion model. We form the trajectory prediction problem as a denoising impaint task and design a map-based guidance term for the diffusion process. TrajDiffuse is able to generate trajectory predictions that match or exceed the accuracy and diversity of the SOTA, while adhering almost perfectly to environmental constraints. We demonstrate the utility of our model through experiments on the nuScenes and PFSD datasets and provide an extensive benchmark analysis against the SOTA methods.
comment: Accepted to be published as inpreceedings of the 2024 International Conference on Pattern Recognition (ICPR)
☆ Generalizable Humanoid Manipulation with Improved 3D Diffusion Policies
Humanoid robots capable of autonomous operation in diverse environments have long been a goal for roboticists. However, autonomous manipulation by humanoid robots has largely been restricted to one specific scene, primarily due to the difficulty of acquiring generalizable skills. Recent advances in 3D visuomotor policies, such as the 3D Diffusion Policy (DP3), have shown promise in extending these capabilities to wilder environments. However, 3D visuomotor policies often rely on camera calibration and point-cloud segmentation, which present challenges for deployment on mobile robots like humanoids. In this work, we introduce the Improved 3D Diffusion Policy (iDP3), a novel 3D visuomotor policy that eliminates these constraints by leveraging egocentric 3D visual representations. We demonstrate that iDP3 enables a full-sized humanoid robot to autonomously perform skills in diverse real-world scenarios, using only data collected in the lab. Videos are available at: https://humanoid-manipulation.github.io
comment: Project website: https://humanoid-manipulation.github.io
☆ Boosting Camera Motion Control for Video Diffusion Transformers
Recent advancements in diffusion models have significantly enhanced the quality of video generation. However, fine-grained control over camera pose remains a challenge. While U-Net-based models have shown promising results for camera control, transformer-based diffusion models (DiT)-the preferred architecture for large-scale video generation - suffer from severe degradation in camera motion accuracy. In this paper, we investigate the underlying causes of this issue and propose solutions tailored to DiT architectures. Our study reveals that camera control performance depends heavily on the choice of conditioning methods rather than camera pose representations that is commonly believed. To address the persistent motion degradation in DiT, we introduce Camera Motion Guidance (CMG), based on classifier-free guidance, which boosts camera control by over 400%. Additionally, we present a sparse camera control pipeline, significantly simplifying the process of specifying camera poses for long videos. Our method universally applies to both U-Net and DiT models, offering improved camera control for video generation tasks.
☆ Towards Foundation Models for 3D Vision: How Close Are We?
Building a foundation model for 3D vision is a complex challenge that remains unsolved. Towards that goal, it is important to understand the 3D reasoning capabilities of current models as well as identify the gaps between these models and humans. Therefore, we construct a new 3D visual understanding benchmark that covers fundamental 3D vision tasks in the Visual Question Answering (VQA) format. We evaluate state-of-the-art Vision-Language Models (VLMs), specialized models, and human subjects on it. Our results show that VLMs generally perform poorly, while the specialized models are accurate but not robust, failing under geometric perturbations. In contrast, human vision continues to be the most reliable 3D visual system. We further demonstrate that neural networks align more closely with human 3D vision mechanisms compared to classical computer vision methods, and Transformer-based networks such as ViT align more closely with human 3D vision mechanisms than CNNs. We hope our study will benefit the future development of foundation models for 3D vision.
☆ MMAR: Towards Lossless Multi-Modal Auto-Regressive Prababilistic Modeling
Recent advancements in multi-modal large language models have propelled the development of joint probabilistic models capable of both image understanding and generation. However, we have identifed that recent methods inevitably suffer from loss of image information during understanding task, due to either image discretization or diffusion denoising steps. To address this issue, we propose a novel Multi-Modal Auto-Regressive (MMAR) probabilistic modeling framework. Unlike discretization line of method, MMAR takes in continuous-valued image tokens to avoid information loss. Differing from diffusion-based approaches, we disentangle the diffusion process from auto-regressive backbone model by employing a light-weight diffusion head on top each auto-regressed image patch embedding. In this way, when the model transits from image generation to understanding through text generation, the backbone model's hidden representation of the image is not limited to the last denoising step. To successfully train our method, we also propose a theoretically proven technique that addresses the numerical stability issue and a training strategy that balances the generation and understanding task goals. Through extensive evaluations on 18 image understanding benchmarks, MMAR demonstrates much more superior performance than other joint multi-modal models, matching the method that employs pretrained CLIP vision encoder, meanwhile being able to generate high quality images at the same time. We also showed that our method is scalable with larger data and model size.
☆ Semantic Image Inversion and Editing using Rectified Stochastic Differential Equations
Generative models transform random noise into images; their inversion aims to transform images back to structured noise for recovery and editing. This paper addresses two key tasks: (i) inversion and (ii) editing of a real image using stochastic equivalents of rectified flow models (such as Flux). Although Diffusion Models (DMs) have recently dominated the field of generative modeling for images, their inversion presents faithfulness and editability challenges due to nonlinearities in drift and diffusion. Existing state-of-the-art DM inversion approaches rely on training of additional parameters or test-time optimization of latent variables; both are expensive in practice. Rectified Flows (RFs) offer a promising alternative to diffusion models, yet their inversion has been underexplored. We propose RF inversion using dynamic optimal control derived via a linear quadratic regulator. We prove that the resulting vector field is equivalent to a rectified stochastic differential equation. Additionally, we extend our framework to design a stochastic sampler for Flux. Our inversion method allows for state-of-the-art performance in zero-shot inversion and editing, outperforming prior works in stroke-to-image synthesis and semantic image editing, with large-scale human evaluations confirming user preference.
comment: Preprint
☆ Condition-Aware Multimodal Fusion for Robust Semantic Perception of Driving Scenes
Leveraging multiple sensors is crucial for robust semantic perception in autonomous driving, as each sensor type has complementary strengths and weaknesses. However, existing sensor fusion methods often treat sensors uniformly across all conditions, leading to suboptimal performance. By contrast, we propose a novel, condition-aware multimodal fusion approach for robust semantic perception of driving scenes. Our method, CAFuser uses an RGB camera input to classify environmental conditions and generate a Condition Token that guides the fusion of multiple sensor modalities. We further newly introduce modality-specific feature adapters to align diverse sensor inputs into a shared latent space, enabling efficient integration with a single and shared pre-trained backbone. By dynamically adapting sensor fusion based on the actual condition, our model significantly improves robustness and accuracy, especially in adverse-condition scenarios. We set the new state of the art with CAFuser on the MUSES dataset with 59.7 PQ for multimodal panoptic segmentation and 78.2 mIoU for semantic segmentation, ranking first on the public benchmarks.
☆ Sitcom-Crafter: A Plot-Driven Human Motion Generation System in 3D Scenes
Recent advancements in human motion synthesis have focused on specific types of motions, such as human-scene interaction, locomotion or human-human interaction, however, there is a lack of a unified system capable of generating a diverse combination of motion types. In response, we introduce Sitcom-Crafter, a comprehensive and extendable system for human motion generation in 3D space, which can be guided by extensive plot contexts to enhance workflow efficiency for anime and game designers. The system is comprised of eight modules, three of which are dedicated to motion generation, while the remaining five are augmentation modules that ensure consistent fusion of motion sequences and system functionality. Central to the generation modules is our novel 3D scene-aware human-human interaction module, which addresses collision issues by synthesizing implicit 3D Signed Distance Function (SDF) points around motion spaces, thereby minimizing human-scene collisions without additional data collection costs. Complementing this, our locomotion and human-scene interaction modules leverage existing methods to enrich the system's motion generation capabilities. Augmentation modules encompass plot comprehension for command generation, motion synchronization for seamless integration of different motion types, hand pose retrieval to enhance motion realism, motion collision revision to prevent human collisions, and 3D retargeting to ensure visual fidelity. Experimental evaluations validate the system's ability to generate high-quality, diverse, and physically realistic motions, underscoring its potential for advancing creative workflows.
comment: Code Page: https://github.com/WindVChen/Sitcom-Crafter
☆ LiveXiv -- A Multi-Modal Live Benchmark Based on Arxiv Papers Content
The large-scale training of multi-modal models on data scraped from the web has shown outstanding utility in infusing these models with the required world knowledge to perform effectively on multiple downstream tasks. However, one downside of scraping data from the web can be the potential sacrifice of the benchmarks on which the abilities of these models are often evaluated. To safeguard against test data contamination and to truly test the abilities of these foundation models we propose LiveXiv: A scalable evolving live benchmark based on scientific ArXiv papers. LiveXiv accesses domain-specific manuscripts at any given timestamp and proposes to automatically generate visual question-answer pairs (VQA). This is done without any human-in-the-loop, using the multi-modal content in the manuscripts, like graphs, charts, and tables. Moreover, we introduce an efficient evaluation approach that estimates the performance of all models on the evolving benchmark using evaluations of only a subset of models. This significantly reduces the overall evaluation cost. We benchmark multiple open and proprietary Large Multi-modal Models (LMMs) on the first version of our benchmark, showing its challenging nature and exposing the models true abilities, avoiding contamination. Lastly, in our commitment to high quality, we have collected and evaluated a manually verified subset. By comparing its overall results to our automatic annotations, we have found that the performance variance is indeed minimal (<2.5%). Our dataset is available online on HuggingFace, and our code will be available here.
☆ 3DArticCyclists: Generating Simulated Dynamic 3D Cyclists for Human-Object Interaction (HOI) and Autonomous Driving Applications
Human-object interaction (HOI) and human-scene interaction (HSI) are crucial for human-centric scene understanding applications in Embodied Artificial Intelligence (EAI), robotics, and augmented reality (AR). A common limitation faced in these research areas is the data scarcity problem: insufficient labeled human-scene object pairs on the input images, and limited interaction complexity and granularity between them. Recent HOI and HSI methods have addressed this issue by generating dynamic interactions with rigid objects. But more complex dynamic interactions such as a human rider pedaling an articulated bicycle have been unexplored. To address this limitation, and to enable research on complex dynamic human-articulated object interactions, in this paper we propose a method to generate simulated 3D dynamic cyclist assets and interactions. We designed a methodology for creating a new part-based multi-view articulated synthetic 3D bicycle dataset that we call 3DArticBikes that can be used to train NeRF and 3DGS-based 3D reconstruction methods. We then propose a 3DGS-based parametric bicycle composition model to assemble 8-DoF pose-controllable 3D bicycles. Finally, using dynamic information from cyclist videos, we build a complete synthetic dynamic 3D cyclist (rider pedaling a bicycle) by re-posing a selectable synthetic 3D person while automatically placing the rider onto one of our new articulated 3D bicycles using a proposed 3D Keypoint optimization-based Inverse Kinematics pose refinement. We present both, qualitative and quantitative results where we compare our generated cyclists against those from a recent stable diffusion-based method.
☆ ControlMM: Controllable Masked Motion Generation
Recent advances in motion diffusion models have enabled spatially controllable text-to-motion generation. However, despite achieving acceptable control precision, these models suffer from generation speed and fidelity limitations. To address these challenges, we propose ControlMM, a novel approach incorporating spatial control signals into the generative masked motion model. ControlMM achieves real-time, high-fidelity, and high-precision controllable motion generation simultaneously. Our approach introduces two key innovations. First, we propose masked consistency modeling, which ensures high-fidelity motion generation via random masking and reconstruction, while minimizing the inconsistency between the input control signals and the extracted control signals from the generated motion. To further enhance control precision, we introduce inference-time logit editing, which manipulates the predicted conditional motion distribution so that the generated motion, sampled from the adjusted distribution, closely adheres to the input control signals. During inference, ControlMM enables parallel and iterative decoding of multiple motion tokens, allowing for high-speed motion generation. Extensive experiments show that, compared to the state of the art, ControlMM delivers superior results in motion quality, with better FID scores (0.061 vs 0.271), and higher control precision (average error 0.0091 vs 0.0108). ControlMM generates motions 20 times faster than diffusion-based methods. Additionally, ControlMM unlocks diverse applications such as any joint any frame control, body part timeline control, and obstacle avoidance. Video visualization can be found at https://exitudio.github.io/ControlMM-page
comment: project page https://exitudio.github.io/ControlMM-page
☆ UniMatch V2: Pushing the Limit of Semi-Supervised Semantic Segmentation
Semi-supervised semantic segmentation (SSS) aims at learning rich visual knowledge from cheap unlabeled images to enhance semantic segmentation capability. Among recent works, UniMatch improves its precedents tremendously by amplifying the practice of weak-to-strong consistency regularization. Subsequent works typically follow similar pipelines and propose various delicate designs. Despite the achieved progress, strangely, even in this flourishing era of numerous powerful vision models, almost all SSS works are still sticking to 1) using outdated ResNet encoders with small-scale ImageNet-1K pre-training, and 2) evaluation on simple Pascal and Cityscapes datasets. In this work, we argue that, it is necessary to switch the baseline of SSS from ResNet-based encoders to more capable ViT-based encoders (e.g., DINOv2) that are pre-trained on massive data. A simple update on the encoder (even using 2x fewer parameters) can bring more significant improvement than careful method designs. Built on this competitive baseline, we present our upgraded and simplified UniMatch V2, inheriting the core spirit of weak-to-strong consistency from V1, but requiring less training cost and providing consistently better results. Additionally, witnessing the gradually saturated performance on Pascal and Cityscapes, we appeal that we should focus on more challenging benchmarks with complex taxonomy, such as ADE20K and COCO datasets. Code, models, and logs of all reported values, are available at https://github.com/LiheYoung/UniMatch-V2.
comment: 18 pages, 18 tables, 10 figures
☆ Cavia: Camera-controllable Multi-view Video Diffusion with View-Integrated Attention
In recent years there have been remarkable breakthroughs in image-to-video generation. However, the 3D consistency and camera controllability of generated frames have remained unsolved. Recent studies have attempted to incorporate camera control into the generation process, but their results are often limited to simple trajectories or lack the ability to generate consistent videos from multiple distinct camera paths for the same scene. To address these limitations, we introduce Cavia, a novel framework for camera-controllable, multi-view video generation, capable of converting an input image into multiple spatiotemporally consistent videos. Our framework extends the spatial and temporal attention modules into view-integrated attention modules, improving both viewpoint and temporal consistency. This flexible design allows for joint training with diverse curated data sources, including scene-level static videos, object-level synthetic multi-view dynamic videos, and real-world monocular dynamic videos. To our best knowledge, Cavia is the first of its kind that allows the user to precisely specify camera motion while obtaining object motion. Extensive experiments demonstrate that Cavia surpasses state-of-the-art methods in terms of geometric consistency and perceptual quality. Project Page: https://ir1d.github.io/Cavia/
comment: Project Page: https://ir1d.github.io/Cavia/
☆ Enhancing JEPAs with Spatial Conditioning: Robust and Efficient Representation Learning NeurIPS 2024
Image-based Joint-Embedding Predictive Architecture (IJEPA) offers an attractive alternative to Masked Autoencoder (MAE) for representation learning using the Masked Image Modeling framework. IJEPA drives representations to capture useful semantic information by predicting in latent rather than input space. However, IJEPA relies on carefully designed context and target windows to avoid representational collapse. The encoder modules in IJEPA cannot adaptively modulate the type of predicted and/or target features based on the feasibility of the masked prediction task as they are not given sufficient information of both context and targets. Based on the intuition that in natural images, information has a strong spatial bias with spatially local regions being highly predictive of one another compared to distant ones. We condition the target encoder and context encoder modules in IJEPA with positions of context and target windows respectively. Our "conditional" encoders show performance gains on several image classification benchmark datasets, improved robustness to context window size and sample-efficiency during pretraining.
comment: NeurIPS 2024 Workshop on Self-Supervised Learning - Theory and Practice. Comments welcome!
☆ Adaptive Diffusion Terrain Generator for Autonomous Uneven Terrain Navigation
Model-free reinforcement learning has emerged as a powerful method for developing robust robot control policies capable of navigating through complex and unstructured terrains. The effectiveness of these methods hinges on two essential elements: (1) the use of massively parallel physics simulations to expedite policy training, and (2) an environment generator tasked with crafting sufficiently challenging yet attainable terrains to facilitate continuous policy improvement. Existing methods of environment generation often rely on heuristics constrained by a set of parameters, limiting the diversity and realism. In this work, we introduce the Adaptive Diffusion Terrain Generator (ADTG), a novel method that leverages Denoising Diffusion Probabilistic Models to dynamically expand existing training environments by adding more diverse and complex terrains adaptive to the current policy. ADTG guides the diffusion model's generation process through initial noise optimization, blending noise-corrupted terrains from existing training environments weighted by the policy's performance in each corresponding environment. By manipulating the noise corruption level, ADTG seamlessly transitions between generating similar terrains for policy fine-tuning and novel ones to expand training diversity. Our experiments show that the policy trained by ADTG outperforms both procedural generated and natural environments, along with popular navigation methods.
☆ DragEntity: Trajectory Guided Video Generation using Entity and Positional Relationships ACM MM2024
In recent years, diffusion models have achieved tremendous success in the field of video generation, with controllable video generation receiving significant attention. However, existing control methods still face two limitations: Firstly, control conditions (such as depth maps, 3D Mesh) are difficult for ordinary users to obtain directly. Secondly, it's challenging to drive multiple objects through complex motions with multiple trajectories simultaneously. In this paper, we introduce DragEntity, a video generation model that utilizes entity representation for controlling the motion of multiple objects. Compared to previous methods, DragEntity offers two main advantages: 1) Our method is more user-friendly for interaction because it allows users to drag entities within the image rather than individual pixels. 2) We use entity representation to represent any object in the image, and multiple objects can maintain relative spatial relationships. Therefore, we allow multiple trajectories to control multiple objects in the image with different levels of complexity simultaneously. Our experiments validate the effectiveness of DragEntity, demonstrating its excellent performance in fine-grained control in video generation.
comment: ACM MM2024 Oral
☆ FlexGen: Flexible Multi-View Generation from Text and Image Inputs
In this work, we introduce FlexGen, a flexible framework designed to generate controllable and consistent multi-view images, conditioned on a single-view image, or a text prompt, or both. FlexGen tackles the challenges of controllable multi-view synthesis through additional conditioning on 3D-aware text annotations. We utilize the strong reasoning capabilities of GPT-4V to generate 3D-aware text annotations. By analyzing four orthogonal views of an object arranged as tiled multi-view images, GPT-4V can produce text annotations that include 3D-aware information with spatial relationship. By integrating the control signal with proposed adaptive dual-control module, our model can generate multi-view images that correspond to the specified text. FlexGen supports multiple controllable capabilities, allowing users to modify text prompts to generate reasonable and corresponding unseen parts. Additionally, users can influence attributes such as appearance and material properties, including metallic and roughness. Extensive experiments demonstrate that our approach offers enhanced multiple controllability, marking a significant advancement over existing multi-view diffusion models. This work has substantial implications for fields requiring rapid and flexible 3D content creation, including game development, animation, and virtual reality. Project page: https://xxu068.github.io/flexgen.github.io/.
comment: 16 pages, 13 figures
☆ Adversarially Robust Out-of-Distribution Detection Using Lyapunov-Stabilized Embeddings
Despite significant advancements in out-of-distribution (OOD) detection, existing methods still struggle to maintain robustness against adversarial attacks, compromising their reliability in critical real-world applications. Previous studies have attempted to address this challenge by exposing detectors to auxiliary OOD datasets alongside adversarial training. However, the increased data complexity inherent in adversarial training, and the myriad of ways that OOD samples can arise during testing, often prevent these approaches from establishing robust decision boundaries. To address these limitations, we propose AROS, a novel approach leveraging neural ordinary differential equations (NODEs) with Lyapunov stability theorem in order to obtain robust embeddings for OOD detection. By incorporating a tailored loss function, we apply Lyapunov stability theory to ensure that both in-distribution (ID) and OOD data converge to stable equilibrium points within the dynamical system. This approach encourages any perturbed input to return to its stable equilibrium, thereby enhancing the model's robustness against adversarial perturbations. To not use additional data, we generate fake OOD embeddings by sampling from low-likelihood regions of the ID data feature space, approximating the boundaries where OOD data are likely to reside. To then further enhance robustness, we propose the use of an orthogonal binary layer following the stable feature space, which maximizes the separation between the equilibrium points of ID and OOD samples. We validate our method through extensive experiments across several benchmarks, demonstrating superior performance, particularly under adversarial attacks. Notably, our approach improves robust detection performance from 37.8% to 80.1% on CIFAR-10 vs. CIFAR-100 and from 29.0% to 67.0% on CIFAR-100 vs. CIFAR-10.
comment: Code and pre-trained models are available at https://github.com/AdaptiveMotorControlLab/AROS
☆ DrivingDojo Dataset: Advancing Interactive and Knowledge-Enriched Driving World Model NeurIPS 2024
Driving world models have gained increasing attention due to their ability to model complex physical dynamics. However, their superb modeling capability is yet to be fully unleashed due to the limited video diversity in current driving datasets. We introduce DrivingDojo, the first dataset tailor-made for training interactive world models with complex driving dynamics. Our dataset features video clips with a complete set of driving maneuvers, diverse multi-agent interplay, and rich open-world driving knowledge, laying a stepping stone for future world model development. We further define an action instruction following (AIF) benchmark for world models and demonstrate the superiority of the proposed dataset for generating action-controlled future predictions.
comment: Accepted to NeurIPS 2024. Project page: https://drivingdojo.github.io/
☆ Deep Compression Autoencoder for Efficient High-Resolution Diffusion Models
We present Deep Compression Autoencoder (DC-AE), a new family of autoencoder models for accelerating high-resolution diffusion models. Existing autoencoder models have demonstrated impressive results at a moderate spatial compression ratio (e.g., 8x), but fail to maintain satisfactory reconstruction accuracy for high spatial compression ratios (e.g., 64x). We address this challenge by introducing two key techniques: (1) Residual Autoencoding, where we design our models to learn residuals based on the space-to-channel transformed features to alleviate the optimization difficulty of high spatial-compression autoencoders; (2) Decoupled High-Resolution Adaptation, an efficient decoupled three-phases training strategy for mitigating the generalization penalty of high spatial-compression autoencoders. With these designs, we improve the autoencoder's spatial compression ratio up to 128 while maintaining the reconstruction quality. Applying our DC-AE to latent diffusion models, we achieve significant speedup without accuracy drop. For example, on ImageNet 512x512, our DC-AE provides 19.1x inference speedup and 17.9x training speedup on H100 GPU for UViT-H while achieving a better FID, compared with the widely used SD-VAE-f8 autoencoder. Our code is available at https://github.com/mit-han-lab/efficientvit.
comment: Preprint. First two authors contributed equally to this work
☆ A Counterexample in Image Registration
Image registration is a widespread problem which applies models about image transformation or image similarity to align discrete images of the same scene. Nevertheless, the theoretical limits on its accuracy are not understood even in the case of one-dimensional data. Just as Nyquist's sampling theorem states conditions for the perfect reconstruction of signals from samples, there are bounds to the quality of reproductions of quantized functions from sets of ideal, noiseless samples in the absence of additional assumptions. In this work we estimate spatially-limited piecewise constant signals from two or more sets of noiseless sampling patterns. We mainly focus on the energy of the error function and find that the uncertainties of the positions of the discontinuity points of the function depend on the discontinuity point selected as the reference point of the signal. As a consequence, the accuracy of the estimate of the signal depends on the reference point of that signal.
☆ 4-LEGS: 4D Language Embedded Gaussian Splatting
The emergence of neural representations has revolutionized our means for digitally viewing a wide range of 3D scenes, enabling the synthesis of photorealistic images rendered from novel views. Recently, several techniques have been proposed for connecting these low-level representations with the high-level semantics understanding embodied within the scene. These methods elevate the rich semantic understanding from 2D imagery to 3D representations, distilling high-dimensional spatial features onto 3D space. In our work, we are interested in connecting language with a dynamic modeling of the world. We show how to lift spatio-temporal features to a 4D representation based on 3D Gaussian Splatting. %, \gal{while introducing a feature-proximity attention mechanism that allows for neighboring features in 3D space to interact}. This enables an interactive interface where the user can spatiotemporally localize events in the video from text prompts. We demonstrate our system on public 3D video datasets of people and animals performing various actions.
comment: Project webpage: https://tau-vailab.github.io/4-LEGS/
☆ Benefiting from Quantum? A Comparative Study of Q-Seg, Quantum-Inspired Techniques, and U-Net for Crack Segmentation
Exploring the potential of quantum hardware for enhancing classical and real-world applications is an ongoing challenge. This study evaluates the performance of quantum and quantum-inspired methods compared to classical models for crack segmentation. Using annotated gray-scale image patches of concrete samples, we benchmark a classical mean Gaussian mixture technique, a quantum-inspired fermion-based method, Q-Seg a quantum annealing-based method, and a U-Net deep learning architecture. Our results indicate that quantum-inspired and quantum methods offer a promising alternative for image segmentation, particularly for complex crack patterns, and could be applied in near-future applications.
☆ Ensemble of ConvNeXt V2 and MaxViT for Long-Tailed CXR Classification with View-Based Aggregation MICCAI
In this work, we present our solution for the MICCAI 2024 CXR-LT challenge, achieving 4th place in Subtask 2 and 5th in Subtask 1. We leveraged an ensemble of ConvNeXt V2 and MaxViT models, pretrained on an external chest X-ray dataset, to address the long-tailed distribution of chest findings. The proposed method combines state-of-the-art image classification techniques, asymmetric loss for handling class imbalance, and view-based prediction aggregation to enhance classification performance. Through experiments, we demonstrate the advantages of our approach in improving both detection accuracy and the handling of the long-tailed distribution in CXR findings. The code is available at \url{https://github.com/yamagishi0824/cxrlt24-multiview-pp}.
comment: Solution paper for MICCAI CXR-LT 2024 challenge. 4th place in Subtask 2, 5th in Subtask 1
☆ Early Diagnoses of Acute Lymphoblastic Leukemia Using YOLOv8 and YOLOv11 Deep Learning Models
Thousands of individuals succumb annually to leukemia alone. This study explores the application of image processing and deep learning techniques for detecting Acute Lymphoblastic Leukemia (ALL), a severe form of blood cancer responsible for numerous annual fatalities. As artificial intelligence technologies advance, the research investigates the reliability of these methods in real-world scenarios. The study focuses on recent developments in ALL detection, particularly using the latest YOLO series models, to distinguish between malignant and benign white blood cells and to identify different stages of ALL, including early stages. Additionally, the models are capable of detecting hematogones, which are often misclassified as ALL. By utilizing advanced deep learning models like YOLOv8 and YOLOv11, the study achieves high accuracy rates reaching 98.8%, demonstrating the effectiveness of these algorithms across multiple datasets and various real-world situations.
comment: 4 pages, 6 figures, 3 tables
☆ TALK-Act: Enhance Textural-Awareness for 2D Speaking Avatar Reenactment with Diffusion Model SIGGRAPH
Recently, 2D speaking avatars have increasingly participated in everyday scenarios due to the fast development of facial animation techniques. However, most existing works neglect the explicit control of human bodies. In this paper, we propose to drive not only the faces but also the torso and gesture movements of a speaking figure. Inspired by recent advances in diffusion models, we propose the Motion-Enhanced Textural-Aware ModeLing for SpeaKing Avatar Reenactment (TALK-Act) framework, which enables high-fidelity avatar reenactment from only short footage of monocular video. Our key idea is to enhance the textural awareness with explicit motion guidance in diffusion modeling. Specifically, we carefully construct 2D and 3D structural information as intermediate guidance. While recent diffusion models adopt a side network for control information injection, they fail to synthesize temporally stable results even with person-specific fine-tuning. We propose a Motion-Enhanced Textural Alignment module to enhance the bond between driving and target signals. Moreover, we build a Memory-based Hand-Recovering module to help with the difficulties in hand-shape preserving. After pre-training, our model can achieve high-fidelity 2D avatar reenactment with only 30 seconds of person-specific data. Extensive experiments demonstrate the effectiveness and superiority of our proposed framework. Resources can be found at https://guanjz20.github.io/projects/TALK-Act.
comment: Accepted to SIGGRAPH Asia 2024 (conference track). Project page: https://guanjz20.github.io/projects/TALK-Act
☆ Both Ears Wide Open: Towards Language-Driven Spatial Audio Generation
Recently, diffusion models have achieved great success in mono-channel audio generation. However, when it comes to stereo audio generation, the soundscapes often have a complex scene of multiple objects and directions. Controlling stereo audio with spatial contexts remains challenging due to high data costs and unstable generative models. To the best of our knowledge, this work represents the first attempt to address these issues. We first construct a large-scale, simulation-based, and GPT-assisted dataset, BEWO-1M, with abundant soundscapes and descriptions even including moving and multiple sources. Beyond text modality, we have also acquired a set of images and rationally paired stereo audios through retrieval to advance multimodal generation. Existing audio generation models tend to generate rather random and indistinct spatial audio. To provide accurate guidance for latent diffusion models, we introduce the SpatialSonic model utilizing spatial-aware encoders and azimuth state matrices to reveal reasonable spatial guidance. By leveraging spatial guidance, our unified model not only achieves the objective of generating immersive and controllable spatial audio from text and image but also enables interactive audio generation during inference. Finally, under fair settings, we conduct subjective and objective evaluations on simulated and real-world data to compare our approach with prevailing methods. The results demonstrate the effectiveness of our method, highlighting its capability to generate spatial audio that adheres to physical rules.
☆ Cross-Modal Few-Shot Learning: a Generative Transfer Learning Framework
Most existing studies on few-shot learning focus on unimodal settings, where models are trained to generalize on unseen data using only a small number of labeled examples from the same modality. However, real-world data are inherently multi-modal, and unimodal approaches limit the practical applications of few-shot learning. To address this gap, this paper introduces the Cross-modal Few-Shot Learning (CFSL) task, which aims to recognize instances from multiple modalities when only a few labeled examples are available. This task presents additional challenges compared to classical few-shot learning due to the distinct visual characteristics and structural properties unique to each modality. To tackle these challenges, we propose a Generative Transfer Learning (GTL) framework consisting of two stages: the first stage involves training on abundant unimodal data, and the second stage focuses on transfer learning to adapt to novel data. Our GTL framework jointly estimates the latent shared concept across modalities and in-modality disturbance in both stages, while freezing the generative module during the transfer phase to maintain the stability of the learned representations and prevent overfitting to the limited multi-modal samples. Our finds demonstrate that GTL has superior performance compared to state-of-the-art methods across four distinct multi-modal datasets: Sketchy, TU-Berlin, Mask1K, and SKSF-A. Additionally, the results suggest that the model can estimate latent concepts from vast unimodal data and generalize these concepts to unseen modalities using only a limited number of available samples, much like human cognitive processes.
comment: 19 pages, 7 figures
☆ Transforming Game Play: A Comparative Study of DCQN and DTQN Architectures in Reinforcement Learning
In this study, we investigate the performance of Deep Q-Networks utilizing Convolutional Neural Networks (CNNs) and Transformer architectures across three different Atari games. The advent of DQNs has significantly advanced Reinforcement Learning, enabling agents to directly learn optimal policies from high-dimensional sensory inputs from pixel or RAM data. While CNN-based DQNs have been extensively studied and deployed in various domains, Transformer-based DQNs are relatively unexplored. Our research aims to fill this gap by benchmarking the performance of both DCQNs and DTQNs across the Atari games Asteroids, Space Invaders, and Centipede. We find that in the 35-40 million parameter range, the DCQN outperforms the DTQN in speed across both ViT and Projection Architectures. We also find the DCQN outperforms the DTQN in all games except for Centipede.
comment: KSU C-Day Spring 2024
☆ PCF-Lift: Panoptic Lifting by Probabilistic Contrastive Fusion ECCV 2024
Panoptic lifting is an effective technique to address the 3D panoptic segmentation task by unprojecting 2D panoptic segmentations from multi-views to 3D scene. However, the quality of its results largely depends on the 2D segmentations, which could be noisy and error-prone, so its performance often drops significantly for complex scenes. In this work, we design a new pipeline coined PCF-Lift based on our Probabilis-tic Contrastive Fusion (PCF) to learn and embed probabilistic features throughout our pipeline to actively consider inaccurate segmentations and inconsistent instance IDs. Technical-wise, we first model the probabilistic feature embeddings through multivariate Gaussian distributions. To fuse the probabilistic features, we incorporate the probability product kernel into the contrastive loss formulation and design a cross-view constraint to enhance the feature consistency across different views. For the inference, we introduce a new probabilistic clustering method to effectively associate prototype features with the underlying 3D object instances for the generation of consistent panoptic segmentation results. Further, we provide a theoretical analysis to justify the superiority of the proposed probabilistic solution. By conducting extensive experiments, our PCF-lift not only significantly outperforms the state-of-the-art methods on widely used benchmarks including the ScanNet dataset and the challenging Messy Room dataset (4.4% improvement of scene-level PQ), but also demonstrates strong robustness when incorporating various 2D segmentation models or different levels of hand-crafted noise.
comment: ECCV 2024. The code is publicly available at https://github.com/Runsong123/PCF-Lift
☆ SANA: Efficient High-Resolution Image Synthesis with Linear Diffusion Transformers
We introduce \model, a text-to-image framework that can efficiently generate images up to 4096$\times$4096 resolution. \model can synthesize high-resolution, high-quality images with strong text-image alignment at a remarkably fast speed, deployable on laptop GPU. Core designs include: (1) Deep compression autoencoder: unlike traditional AEs, which compress images only 8$\times$, we trained an AE that can compress images 32$\times$, effectively reducing the number of latent tokens. (2) Linear DiT: we replace all vanilla attention in DiT with linear attention, which is more efficient at high resolutions without sacrificing quality. (3) Decoder-only text encoder: we replaced T5 with modern decoder-only small LLM as the text encoder and designed complex human instruction with in-context learning to enhance the image-text alignment. (4) Efficient training and sampling: we propose Flow-DPM-Solver to reduce sampling steps, with efficient caption labeling and selection to accelerate convergence. As a result, \model-0.6B is very competitive with modern giant diffusion model (e.g. Flux-12B), being 20 times smaller and 100+ times faster in measured throughput. Moreover, \model-0.6B can be deployed on a 16GB laptop GPU, taking less than 1 second to generate a 1024$\times$1024 resolution image. Sana enables content creation at low cost. Code and model will be publicly released.
comment: Technical Report
☆ BrainMVP: Multi-modal Vision Pre-training for Brain Image Analysis using Multi-parametric MRI
Accurate diagnosis of brain abnormalities is greatly enhanced by the inclusion of complementary multi-parametric MRI imaging data. There is significant potential to develop a universal pre-training model that can be quickly adapted for image modalities and various clinical scenarios. However, current models often rely on uni-modal image data, neglecting the cross-modal correlations among different image modalities or struggling to scale up pre-training in the presence of missing modality data. In this paper, we propose BrainMVP, a multi-modal vision pre-training framework for brain image analysis using multi-parametric MRI scans. First, we collect 16,022 brain MRI scans (over 2.4 million images), encompassing eight MRI modalities sourced from a diverse range of centers and devices. Then, a novel pre-training paradigm is proposed for the multi-modal MRI data, addressing the issue of missing modalities and achieving multi-modal information fusion. Cross-modal reconstruction is explored to learn distinctive brain image embeddings and efficient modality fusion capabilities. A modality-wise data distillation module is proposed to extract the essence representation of each MR image modality for both the pre-training and downstream application purposes. Furthermore, we introduce a modality-aware contrastive learning module to enhance the cross-modality association within a study. Extensive experiments on downstream tasks demonstrate superior performance compared to state-of-the-art pre-training methods in the medical domain, with Dice Score improvement of 0.28%-14.47% across six segmentation benchmarks and a consistent accuracy improvement of 0.65%-18.07% in four individual classification tasks.
☆ VisRAG: Vision-based Retrieval-augmented Generation on Multi-modality Documents
Retrieval-augmented generation (RAG) is an effective technique that enables large language models (LLMs) to utilize external knowledge sources for generation. However, current RAG systems are solely based on text, rendering it impossible to utilize vision information like layout and images that play crucial roles in real-world multi-modality documents. In this paper, we introduce VisRAG, which tackles this issue by establishing a vision-language model (VLM)-based RAG pipeline. In this pipeline, instead of first parsing the document to obtain text, the document is directly embedded using a VLM as an image and then retrieved to enhance the generation of a VLM. Compared to traditional text-based RAG, VisRAG maximizes the retention and utilization of the data information in the original documents, eliminating the information loss introduced during the parsing process. We collect both open-source and synthetic data to train the retriever in VisRAG and explore a variety of generation methods. Experiments demonstrate that VisRAG outperforms traditional RAG in both the retrieval and generation stages, achieving a 25--39\% end-to-end performance gain over traditional text-based RAG pipeline. Further analysis reveals that VisRAG is effective in utilizing training data and demonstrates strong generalization capability, positioning it as a promising solution for RAG on multi-modality documents. Our code and data are available at https://github.com/openbmb/visrag .
☆ MoTE: Reconciling Generalization with Specialization for Visual-Language to Video Knowledge Transfer NeurIPS 2024
Transferring visual-language knowledge from large-scale foundation models for video recognition has proved to be effective. To bridge the domain gap, additional parametric modules are added to capture the temporal information. However, zero-shot generalization diminishes with the increase in the number of specialized parameters, making existing works a trade-off between zero-shot and close-set performance. In this paper, we present MoTE, a novel framework that enables generalization and specialization to be balanced in one unified model. Our approach tunes a mixture of temporal experts to learn multiple task views with various degrees of data fitting. To maximally preserve the knowledge of each expert, we propose \emph{Weight Merging Regularization}, which regularizes the merging process of experts in weight space. Additionally with temporal feature modulation to regularize the contribution of temporal feature during test. We achieve a sound balance between zero-shot and close-set video recognition tasks and obtain state-of-the-art or competitive results on various datasets, including Kinetics-400 \& 600, UCF, and HMDB. Code is available at \url{https://github.com/ZMHH-H/MoTE}.
comment: NeurIPS 2024 Camera Ready
☆ TopoFR: A Closer Look at Topology Alignment on Face Recognition NeurIPS 2024
The field of face recognition (FR) has undergone significant advancements with the rise of deep learning. Recently, the success of unsupervised learning and graph neural networks has demonstrated the effectiveness of data structure information. Considering that the FR task can leverage large-scale training data, which intrinsically contains significant structure information, we aim to investigate how to encode such critical structure information into the latent space. As revealed from our observations, directly aligning the structure information between the input and latent spaces inevitably suffers from an overfitting problem, leading to a structure collapse phenomenon in the latent space. To address this problem, we propose TopoFR, a novel FR model that leverages a topological structure alignment strategy called PTSA and a hard sample mining strategy named SDE. Concretely, PTSA uses persistent homology to align the topological structures of the input and latent spaces, effectively preserving the structure information and improving the generalization performance of FR model. To mitigate the impact of hard samples on the latent space structure, SDE accurately identifies hard samples by automatically computing structure damage score (SDS) for each sample, and directs the model to prioritize optimizing these samples. Experimental results on popular face benchmarks demonstrate the superiority of our TopoFR over the state-of-the-art methods. Code and models are available at: https://github.com/modelscope/facechain/tree/main/face_module/TopoFR.
comment: Accepted by NeurIPS 2024
☆ Queryable Prototype Multiple Instance Learning with Vision-Language Models for Incremental Whole Slide Image Classification
Whole Slide Image (WSI) classification has very significant applications in clinical pathology, e.g., tumor identification and cancer diagnosis. Currently, most research attention is focused on Multiple Instance Learning (MIL) using static datasets. One of the most obvious weaknesses of these methods is that they cannot efficiently preserve and utilize previously learned knowledge. With any new data arriving, classification models are required to be re-trained on both previous and current new data. To overcome this shortcoming and break through traditional vision modality, this paper proposes the first Vision-Language-based framework with Queryable Prototype Multiple Instance Learning (QPMIL-VL) specially designed for incremental WSI classification. This framework mainly consists of two information processing branches. One is for generating the bag-level feature by prototype-guided aggregating on the instance features. While the other is for enhancing the class feature through class ensemble, tunable vector and class similarity loss. The experiments on four TCGA datasets demonstrate that our QPMIL-VL framework is effective for incremental WSI classification and often significantly outperforms other compared methods, achieving state-of-the-art (SOTA) performance.
comment: 16 pages, 10 tables, 11 figures
☆ MEGA-Bench: Scaling Multimodal Evaluation to over 500 Real-World Tasks
We present MEGA-Bench, an evaluation suite that scales multimodal evaluation to over 500 real-world tasks, to address the highly heterogeneous daily use cases of end users. Our objective is to optimize for a set of high-quality data samples that cover a highly diverse and rich set of multimodal tasks, while enabling cost-effective and accurate model evaluation. In particular, we collected 505 realistic tasks encompassing over 8,000 samples from 16 expert annotators to extensively cover the multimodal task space. Instead of unifying these problems into standard multi-choice questions (like MMMU, MMBench, and MMT-Bench), we embrace a wide range of output formats like numbers, phrases, code, \LaTeX, coordinates, JSON, free-form, etc. To accommodate these formats, we developed over 40 metrics to evaluate these tasks. Unlike existing benchmarks, MEGA-Bench offers a fine-grained capability report across multiple dimensions (e.g., application, input type, output format, skill), allowing users to interact with and visualize model capabilities in depth. We evaluate a wide variety of frontier vision-language models on MEGA-Bench to understand their capabilities across these dimensions.
comment: Technical report. Project page: https://tiger-ai-lab.github.io/MEGA-Bench/
☆ ROSAR: An Adversarial Re-Training Framework for Robust Side-Scan Sonar Object Detection
This paper introduces ROSAR, a novel framework enhancing the robustness of deep learning object detection models tailored for side-scan sonar (SSS) images, generated by autonomous underwater vehicles using sonar sensors. By extending our prior work on knowledge distillation (KD), this framework integrates KD with adversarial retraining to address the dual challenges of model efficiency and robustness against SSS noises. We introduce three novel, publicly available SSS datasets, capturing different sonar setups and noise conditions. We propose and formalize two SSS safety properties and utilize them to generate adversarial datasets for retraining. Through a comparative analysis of projected gradient descent (PGD) and patch-based adversarial attacks, ROSAR demonstrates significant improvements in model robustness and detection accuracy under SSS-specific conditions, enhancing the model's robustness by up to 1.85%. ROSAR is available at https://github.com/remaro-network/ROSAR-framework.
☆ Preserving Cardiac Integrity: A Topology-Infused Approach to Whole Heart Segmentation
Whole heart segmentation (WHS) supports cardiovascular disease (CVD) diagnosis, disease monitoring, treatment planning, and prognosis. Deep learning has become the most widely used method for WHS applications in recent years. However, segmentation of whole-heart structures faces numerous challenges including heart shape variability during the cardiac cycle, clinical artifacts like motion and poor contrast-to-noise ratio, domain shifts in multi-center data, and the distinct modalities of CT and MRI. To address these limitations and improve segmentation quality, this paper introduces a new topology-preserving module that is integrated into deep neural networks. The implementation achieves anatomically plausible segmentation by using learned topology-preserving fields, which are based entirely on 3D convolution and are therefore very effective for 3D voxel data. We incorporate natural constraints between structures into the end-to-end training and enrich the feature representation of the neural network. The effectiveness of the proposed method is validated on an open-source medical heart dataset, specifically using the WHS++ data. The results demonstrate that the architecture performs exceptionally well, achieving a Dice coefficient of 0.939 during testing. This indicates full topology preservation for individual structures and significantly outperforms other baselines in preserving the overall scene topology.
☆ RICASSO: Reinforced Imbalance Learning with Class-Aware Self-Supervised Outliers Exposure
In real-world scenarios, deep learning models often face challenges from both imbalanced (long-tailed) and out-of-distribution (OOD) data. However, existing joint methods rely on real OOD data, which leads to unnecessary trade-offs. In contrast, our research shows that data mixing, a potent augmentation technique for long-tailed recognition, can generate pseudo-OOD data that exhibit the features of both in-distribution (ID) data and OOD data. Therefore, by using mixed data instead of real OOD data, we can address long-tailed recognition and OOD detection holistically. We propose a unified framework called Reinforced Imbalance Learning with Class-Aware Self-Supervised Outliers Exposure (RICASSO), where "self-supervised" denotes that we only use ID data for outlier exposure. RICASSO includes three main strategies: Norm-Odd-Duality-Based Outlier Exposure: Uses mixed data as pseudo-OOD data, enabling simultaneous ID data rebalancing and outlier exposure through a single loss function. Ambiguity-Aware Logits Adjustment: Utilizes the ambiguity of ID data to adaptively recalibrate logits. Contrastive Boundary-Center Learning: Combines Virtual Boundary Learning and Dual-Entropy Center Learning to use mixed data for better feature separation and clustering, with Representation Consistency Learning for robustness. Extensive experiments demonstrate that RICASSO achieves state-of-the-art performance in long-tailed recognition and significantly improves OOD detection compared to our baseline (27% improvement in AUROC and 61% reduction in FPR on the iNaturalist2018 dataset). On iNaturalist2018, we even outperforms methods using real OOD data. The code will be made public soon.
comment: 14 pages, 2 figures
☆ Hybrid Transformer for Early Alzheimer's Detection: Integration of Handwriting-Based 2D Images and 1D Signal Features
Alzheimer's Disease (AD) is a prevalent neurodegenerative condition where early detection is vital. Handwriting, often affected early in AD, offers a non-invasive and cost-effective way to capture subtle motor changes. State-of-the-art research on handwriting, mostly online, based AD detection has predominantly relied on manually extracted features, fed as input to shallow machine learning models. Some recent works have proposed deep learning (DL)-based models, either 1D-CNN or 2D-CNN architectures, with performance comparing favorably to handcrafted schemes. These approaches, however, overlook the intrinsic relationship between the 2D spatial patterns of handwriting strokes and their 1D dynamic characteristics, thus limiting their capacity to capture the multimodal nature of handwriting data. Moreover, the application of Transformer models remains basically unexplored. To address these limitations, we propose a novel approach for AD detection, consisting of a learnable multimodal hybrid attention model that integrates simultaneously 2D handwriting images with 1D dynamic handwriting signals. Our model leverages a gated mechanism to combine similarity and difference attention, blending the two modalities and learning robust features by incorporating information at different scales. Our model achieved state-of-the-art performance on the DARWIN dataset, with an F1-score of 90.32\% and accuracy of 90.91\% in Task 8 ('L' writing), surpassing the previous best by 4.61% and 6.06% respectively.
☆ Motion-guided small MAV detection in complex and non-planar scenes
In recent years, there has been a growing interest in the visual detection of micro aerial vehicles (MAVs) due to its importance in numerous applications. However, the existing methods based on either appearance or motion features encounter difficulties when the background is complex or the MAV is too small. In this paper, we propose a novel motion-guided MAV detector that can accurately identify small MAVs in complex and non-planar scenes. This detector first exploits a motion feature enhancement module to capture the motion features of small MAVs. Then it uses multi-object tracking and trajectory filtering to eliminate false positives caused by motion parallax. Finally, an appearance-based classifier and an appearance-based detector that operates on the cropped regions are used to achieve precise detection results. Our proposed method can effectively and efficiently detect extremely small MAVs from dynamic and complex backgrounds because it aggregates pixel-level motion features and eliminates false positives based on the motion and appearance features of MAVs. Experiments on the ARD-MAV dataset demonstrate that the proposed method could achieve high performance in small MAV detection under challenging conditions and outperform other state-of-the-art methods across various metrics
comment: 8 pages, 6 figures
☆ Customize Your Visual Autoregressive Recipe with Set Autoregressive Modeling
We introduce a new paradigm for AutoRegressive (AR) image generation, termed Set AutoRegressive Modeling (SAR). SAR generalizes the conventional AR to the next-set setting, i.e., splitting the sequence into arbitrary sets containing multiple tokens, rather than outputting each token in a fixed raster order. To accommodate SAR, we develop a straightforward architecture termed Fully Masked Transformer. We reveal that existing AR variants correspond to specific design choices of sequence order and output intervals within the SAR framework, with AR and Masked AR (MAR) as two extreme instances. Notably, SAR facilitates a seamless transition from AR to MAR, where intermediate states allow for training a causal model that benefits from both few-step inference and KV cache acceleration, thus leveraging the advantages of both AR and MAR. On the ImageNet benchmark, we carefully explore the properties of SAR by analyzing the impact of sequence order and output intervals on performance, as well as the generalization ability regarding inference order and steps. We further validate the potential of SAR by training a 900M text-to-image model capable of synthesizing photo-realistic images with any resolution. We hope our work may inspire more exploration and application of AR-based modeling across diverse modalities.
comment: 19 pages, 17 figures, 8 tables, github repo: https://github.com/poppuppy/SAR
☆ Exploiting Local Features and Range Images for Small Data Real-Time Point Cloud Semantic Segmentation IROS
Semantic segmentation of point clouds is an essential task for understanding the environment in autonomous driving and robotics. Recent range-based works achieve real-time efficiency, while point- and voxel-based methods produce better results but are affected by high computational complexity. Moreover, highly complex deep learning models are often not suited to efficiently learn from small datasets. Their generalization capabilities can easily be driven by the abundance of data rather than the architecture design. In this paper, we harness the information from the three-dimensional representation to proficiently capture local features, while introducing the range image representation to incorporate additional information and facilitate fast computation. A GPU-based KDTree allows for rapid building, querying, and enhancing projection with straightforward operations. Extensive experiments on SemanticKITTI and nuScenes datasets demonstrate the benefits of our modification in a ``small data'' setup, in which only one sequence of the dataset is used to train the models, but also in the conventional setup, where all sequences except one are used for training. We show that a reduced version of our model not only demonstrates strong competitiveness against full-scale state-of-the-art models but also operates in real-time, making it a viable choice for real-world case applications. The code of our method is available at https://github.com/Bender97/WaffleAndRange.
comment: This paper has been accepted for publication at the 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
☆ Artificial Intelligence-Based Triaging of Cutaneous Melanocytic Lesions
Pathologists are facing an increasing workload due to a growing volume of cases and the need for more comprehensive diagnoses. Aiming to facilitate workload reduction and faster turnaround times, we developed an artificial intelligence (AI) model for triaging cutaneous melanocytic lesions based on whole slide images. The AI model was developed and validated using a retrospective cohort from the UMC Utrecht. The dataset consisted of 52,202 whole slide images from 27,167 unique specimens, acquired from 20,707 patients. Specimens with only common nevi were assigned to the low complexity category (86.6%). In contrast, specimens with any other melanocytic lesion subtype, including non-common nevi, melanocytomas, and melanomas, were assigned to the high complexity category (13.4%). The dataset was split on patient level into a development set (80%) and test sets (20%) for independent evaluation. Predictive performance was primarily measured using the area under the receiver operating characteristic curve (AUROC) and the area under the precision-recall curve (AUPRC). A simulation experiment was performed to study the effect of implementing AI-based triaging in the clinic. The AI model reached an AUROC of 0.966 (95% CI, 0.960-0.972) and an AUPRC of 0.857 (95% CI, 0.836-0.877) on the in-distribution test set, and an AUROC of 0.899 (95% CI, 0.860-0.934) and an AUPRC of 0.498 (95% CI, 0.360-0.639) on the out-of-distribution test set. In the simulation experiment, using random case assignment as baseline, AI-based triaging prevented an average of 43.9 (95% CI, 36-55) initial examinations of high complexity cases by general pathologists for every 500 cases. In conclusion, the AI model achieved a strong predictive performance in differentiating between cutaneous melanocytic lesions of high and low complexity. The improvement in workflow efficiency due to AI-based triaging could be substantial.
comment: 14 pages, 6 figures
☆ Continual Learning Improves Zero-Shot Action Recognition ACCV 2024
Zero-shot action recognition requires a strong ability to generalize from pre-training and seen classes to novel unseen classes. Similarly, continual learning aims to develop models that can generalize effectively and learn new tasks without forgetting the ones previously learned. The generalization goals of zero-shot and continual learning are closely aligned, however techniques from continual learning have not been applied to zero-shot action recognition. In this paper, we propose a novel method based on continual learning to address zero-shot action recognition. This model, which we call {\em Generative Iterative Learning} (GIL) uses a memory of synthesized features of past classes, and combines these synthetic features with real ones from novel classes. The memory is used to train a classification model, ensuring a balanced exposure to both old and new classes. Experiments demonstrate that {\em GIL} improves generalization in unseen classes, achieving a new state-of-the-art in zero-shot recognition across multiple benchmarks. Importantly, {\em GIL} also boosts performance in the more challenging generalized zero-shot setting, where models need to retain knowledge about classes seen before fine-tuning.
comment: Accepted in ACCV 2024
☆ Vision-guided and Mask-enhanced Adaptive Denoising for Prompt-based Image Editing
Text-to-image diffusion models have demonstrated remarkable progress in synthesizing high-quality images from text prompts, which boosts researches on prompt-based image editing that edits a source image according to a target prompt. Despite their advances, existing methods still encounter three key issues: 1) limited capacity of the text prompt in guiding target image generation, 2) insufficient mining of word-to-patch and patch-to-patch relationships for grounding editing areas, and 3) unified editing strength for all regions during each denoising step. To address these issues, we present a Vision-guided and Mask-enhanced Adaptive Editing (ViMAEdit) method with three key novel designs. First, we propose to leverage image embeddings as explicit guidance to enhance the conventional textual prompt-based denoising process, where a CLIP-based target image embedding estimation strategy is introduced. Second, we devise a self-attention-guided iterative editing area grounding strategy, which iteratively exploits patch-to-patch relationships conveyed by self-attention maps to refine those word-to-patch relationships contained in cross-attention maps. Last, we present a spatially adaptive variance-guided sampling, which highlights sampling variances for critical image regions to promote the editing capability. Experimental results demonstrate the superior editing capacity of ViMAEdit over all existing methods.
☆ Learning to Ground VLMs without Forgetting
Spatial awareness is key to enable embodied multimodal AI systems. Yet, without vast amounts of spatial supervision, current Visual Language Models (VLMs) struggle at this task. In this paper, we introduce LynX, a framework that equips pretrained VLMs with visual grounding ability without forgetting their existing image and language understanding skills. To this end, we propose a Dual Mixture of Experts module that modifies only the decoder layer of the language model, using one frozen Mixture of Experts (MoE) pre-trained on image and language understanding and another learnable MoE for new grounding capabilities. This allows the VLM to retain previously learned knowledge and skills, while acquiring what is missing. To train the model effectively, we generate a high-quality synthetic dataset we call SCouT, which mimics human reasoning in visual grounding. This dataset provides rich supervision signals, describing a step-by-step multimodal reasoning process, thereby simplifying the task of visual grounding. We evaluate LynX on several object detection and visual grounding datasets, demonstrating strong performance in object detection, zero-shot localization and grounded reasoning while maintaining its original image and language understanding capabilities on seven standard benchmark datasets.
☆ A Novel No-Reference Image Quality Metric For Assessing Sharpness In Satellite Imagery
This study introduces a novel no-reference image quality metric aimed at assessing image sharpness. Designed to be robust against variations in noise, exposure, contrast, and image content, it measures the normalized decay rate of gradients along pronounced edges, offering an objective method for sharpness evaluation without reference images. Primarily developed for satellite imagery to align with human visual perception of sharpness, this metric supports monitoring and quality characterization of satellite fleets. It demonstrates significant utility and superior performance in consistency with human perception across various image types and operational conditions. Unlike conventional metrics, this heuristic approach provides a way to score images from lower to higher sharpness, making it a reliable and versatile tool for enhancing quality assessment processes without the need for pristine or ground truth comparison. Additionally, this metric is computationally efficient compared to deep learning analysis, ensuring faster and more resource-effective sharpness evaluations.
comment: 10 pages, 6 figures
☆ Advancing Newborn Care: Precise Birth Time Detection Using AI-Driven Thermal Imaging with Adaptive Normalization
Around 5-10\% of newborns need assistance to start breathing. Currently, there is a lack of evidence-based research, objective data collection, and opportunities for learning from real newborn resuscitation emergency events. Generating and evaluating automated newborn resuscitation algorithm activity timelines relative to the Time of Birth (ToB) offers a promising opportunity to enhance newborn care practices. Given the importance of prompt resuscitation interventions within the "golden minute" after birth, having an accurate ToB with second precision is essential for effective subsequent analysis of newborn resuscitation episodes. Instead, ToB is generally registered manually, often with minute precision, making the process inefficient and susceptible to error and imprecision. In this work, we explore the fusion of Artificial Intelligence (AI) and thermal imaging to develop the first AI-driven ToB detector. The use of temperature information offers a promising alternative to detect the newborn while respecting the privacy of healthcare providers and mothers. However, the frequent inconsistencies in thermal measurements, especially in a multi-camera setup, make normalization strategies critical. Our methodology involves a three-step process: first, we propose an adaptive normalization method based on Gaussian mixture models (GMM) to mitigate issues related to temperature variations; second, we implement and deploy an AI model to detect the presence of the newborn within the thermal video frames; and third, we evaluate and post-process the model's predictions to estimate the ToB. A precision of 88.1\% and a recall of 89.3\% are reported in the detection of the newborn within thermal frames during performance evaluation. Our approach achieves an absolute median deviation of 2.7 seconds in estimating the ToB relative to the manual annotations.
comment: Paper submitted to Computer in Biology and Medicine, ELSEVIER
☆ ReLayout: Towards Real-World Document Understanding via Layout-enhanced Pre-training
Recent approaches for visually-rich document understanding (VrDU) uses manually annotated semantic groups, where a semantic group encompasses all semantically relevant but not obviously grouped words. As OCR tools are unable to automatically identify such grouping, we argue that current VrDU approaches are unrealistic. We thus introduce a new variant of the VrDU task, real-world visually-rich document understanding (ReVrDU), that does not allow for using manually annotated semantic groups. We also propose a new method, ReLayout, compliant with the ReVrDU scenario, which learns to capture semantic grouping through arranging words and bringing the representations of words that belong to the potential same semantic group closer together. Our experimental results demonstrate the performance of existing methods is deteriorated with the ReVrDU task, while ReLayout shows superiour performance.
☆ Improve Meta-learning for Few-Shot Text Classification with All You Can Acquire from the Tasks EMNLP 2024
Meta-learning has emerged as a prominent technology for few-shot text classification and has achieved promising performance. However, existing methods often encounter difficulties in drawing accurate class prototypes from support set samples, primarily due to probable large intra-class differences and small inter-class differences within the task. Recent approaches attempt to incorporate external knowledge or pre-trained language models to augment data, but this requires additional resources and thus does not suit many few-shot scenarios. In this paper, we propose a novel solution to address this issue by adequately leveraging the information within the task itself. Specifically, we utilize label information to construct a task-adaptive metric space, thereby adaptively reducing the intra-class differences and magnifying the inter-class differences. We further employ the optimal transport technique to estimate class prototypes with query set samples together, mitigating the problem of inaccurate and ambiguous support set samples caused by large intra-class differences. We conduct extensive experiments on eight benchmark datasets, and our approach shows obvious advantages over state-of-the-art models across all the tasks on all the datasets. For reproducibility, all the datasets and codes are available at https://github.com/YvoGao/LAQDA.
comment: Accepted by EMNLP 2024 Findings
☆ Self-Assessed Generation: Trustworthy Label Generation for Optical Flow and Stereo Matching in Real-world
A significant challenge facing current optical flow and stereo methods is the difficulty in generalizing them well to the real world. This is mainly due to the high costs required to produce datasets, and the limitations of existing self-supervised methods on fuzzy results and complex model training problems. To address the above challenges, we propose a unified self-supervised generalization framework for optical flow and stereo tasks: Self-Assessed Generation (SAG). Unlike previous self-supervised methods, SAG is data-driven, using advanced reconstruction techniques to construct a reconstruction field from RGB images and generate datasets based on it. Afterward, we quantified the confidence level of the generated results from multiple perspectives, such as reconstruction field distribution, geometric consistency, and structural similarity, to eliminate inevitable defects in the generation process. We also designed a 3D flight foreground automatic rendering pipeline in SAG to encourage the network to learn occlusion and motion foreground. Experimentally, because SAG does not involve changes to methods or loss functions, it can directly self-supervised train the state-of-the-art deep networks, greatly improving the generalization performance of self-supervised methods on current mainstream optical flow and stereo-matching datasets. Compared to previous training modes, SAG is more generalized, cost-effective, and accurate.
☆ Domain-Conditioned Transformer for Fully Test-time Adaptation
Fully test-time adaptation aims to adapt a network model online based on sequential analysis of input samples during the inference stage. We observe that, when applying a transformer network model into a new domain, the self-attention profiles of image samples in the target domain deviate significantly from those in the source domain, which results in large performance degradation during domain changes. To address this important issue, we propose a new structure for the self-attention modules in the transformer. Specifically, we incorporate three domain-conditioning vectors, called domain conditioners, into the query, key, and value components of the self-attention module. We learn a network to generate these three domain conditioners from the class token at each transformer network layer. We find that, during fully online test-time adaptation, these domain conditioners at each transform network layer are able to gradually remove the impact of domain shift and largely recover the original self-attention profile. Our extensive experimental results demonstrate that the proposed domain-conditioned transformer significantly improves the online fully test-time domain adaptation performance and outperforms existing state-of-the-art methods by large margins.
☆ Free Video-LLM: Prompt-guided Visual Perception for Efficient Training-free Video LLMs
Vision-language large models have achieved remarkable success in various multi-modal tasks, yet applying them to video understanding remains challenging due to the inherent complexity and computational demands of video data. While training-based video-LLMs deliver high performance, they often require substantial resources for training and inference. Conversely, training-free approaches offer a more efficient alternative by adapting pre-trained image-LLMs models for video tasks without additional training, but they face inference efficiency bottlenecks due to the large number of visual tokens generated from video frames. In this work, we present a novel prompt-guided visual perception framework (abbreviated as \emph{Free Video-LLM}) for efficient inference of training-free video LLMs. The proposed framework decouples spatial-temporal dimension and performs temporal frame sampling and spatial RoI cropping respectively based on task-specific prompts. Our method effectively reduces the number of visual tokens while maintaining high performance across multiple video question-answering benchmarks. Extensive experiments demonstrate that our approach achieves competitive results with significantly fewer tokens, offering an optimal trade-off between accuracy and computational efficiency compared to state-of-the-art video LLMs. The code will be available at \url{https://github.com/contrastive/FreeVideoLLM}.
comment: Tech report
☆ Towards Reliable Verification of Unauthorized Data Usage in Personalized Text-to-Image Diffusion Models
Text-to-image diffusion models are pushing the boundaries of what generative AI can achieve in our lives. Beyond their ability to generate general images, new personalization techniques have been proposed to customize the pre-trained base models for crafting images with specific themes or styles. Such a lightweight solution, enabling AI practitioners and developers to easily build their own personalized models, also poses a new concern regarding whether the personalized models are trained from unauthorized data. A promising solution is to proactively enable data traceability in generative models, where data owners embed external coatings (e.g., image watermarks or backdoor triggers) onto the datasets before releasing. Later the models trained over such datasets will also learn the coatings and unconsciously reproduce them in the generated mimicries, which can be extracted and used as the data usage evidence. However, we identify the existing coatings cannot be effectively learned in personalization tasks, making the corresponding verification less reliable. In this paper, we introduce SIREN, a novel methodology to proactively trace unauthorized data usage in black-box personalized text-to-image diffusion models. Our approach optimizes the coating in a delicate way to be recognized by the model as a feature relevant to the personalization task, thus significantly improving its learnability. We also utilize a human perceptual-aware constraint, a hypersphere classification technique, and a hypothesis-testing-guided verification method to enhance the stealthiness and detection accuracy of the coating. The effectiveness of SIREN is verified through extensive experiments on a diverse set of benchmark datasets, models, and learning algorithms. SIREN is also effective in various real-world scenarios and evaluated against potential countermeasures. Our code is publicly available.
comment: To appear in the IEEE Symposium on Security & Privacy, May 2025
☆ LKASeg:Remote-Sensing Image Semantic Segmentation with Large Kernel Attention and Full-Scale Skip Connections ICASSP 2025
Semantic segmentation of remote sensing images is a fundamental task in geospatial research. However, widely used Convolutional Neural Networks (CNNs) and Transformers have notable drawbacks: CNNs may be limited by insufficient remote sensing modeling capability, while Transformers face challenges due to computational complexity. In this paper, we propose a remote-sensing image semantic segmentation network named LKASeg, which combines Large Kernel Attention(LSKA) and Full-Scale Skip Connections(FSC). Specifically, we propose a decoder based on Large Kernel Attention (LKA), which extract global features while avoiding the computational overhead of self-attention and providing channel adaptability. To achieve full-scale feature learning and fusion, we apply Full-Scale Skip Connections (FSC) between the encoder and decoder. We conducted experiments by combining the LKA-based decoder with FSC. On the ISPRS Vaihingen dataset, the mF1 and mIoU scores achieved 90.33% and 82.77%.
comment: The paper is under consideration at 2025 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2025)
☆ DOME: Taming Diffusion Model into High-Fidelity Controllable Occupancy World Model
We propose DOME, a diffusion-based world model that predicts future occupancy frames based on past occupancy observations. The ability of this world model to capture the evolution of the environment is crucial for planning in autonomous driving. Compared to 2D video-based world models, the occupancy world model utilizes a native 3D representation, which features easily obtainable annotations and is modality-agnostic. This flexibility has the potential to facilitate the development of more advanced world models. Existing occupancy world models either suffer from detail loss due to discrete tokenization or rely on simplistic diffusion architectures, leading to inefficiencies and difficulties in predicting future occupancy with controllability. Our DOME exhibits two key features:(1) High-Fidelity and Long-Duration Generation. We adopt a spatial-temporal diffusion transformer to predict future occupancy frames based on historical context. This architecture efficiently captures spatial-temporal information, enabling high-fidelity details and the ability to generate predictions over long durations. (2)Fine-grained Controllability. We address the challenge of controllability in predictions by introducing a trajectory resampling method, which significantly enhances the model's ability to generate controlled predictions. Extensive experiments on the widely used nuScenes dataset demonstrate that our method surpasses existing baselines in both qualitative and quantitative evaluations, establishing a new state-of-the-art performance on nuScenes. Specifically, our approach surpasses the baseline by 10.5% in mIoU and 21.2% in IoU for occupancy reconstruction and by 36.0% in mIoU and 24.6% in IoU for 4D occupancy forecasting.
comment: Please visit our project page at https://gusongen.github.io/DOME
☆ 4DStyleGaussian: Zero-shot 4D Style Transfer with Gaussian Splatting
3D neural style transfer has gained significant attention for its potential to provide user-friendly stylization with spatial consistency. However, existing 3D style transfer methods often fall short in terms of inference efficiency, generalization ability, and struggle to handle dynamic scenes with temporal consistency. In this paper, we introduce 4DStyleGaussian, a novel 4D style transfer framework designed to achieve real-time stylization of arbitrary style references while maintaining reasonable content affinity, multi-view consistency, and temporal coherence. Our approach leverages an embedded 4D Gaussian Splatting technique, which is trained using a reversible neural network for reducing content loss in the feature distillation process. Utilizing the 4D embedded Gaussians, we predict a 4D style transformation matrix that facilitates spatially and temporally consistent style transfer with Gaussian Splatting. Experiments demonstrate that our method can achieve high-quality and zero-shot stylization for 4D scenarios with enhanced efficiency and spatial-temporal consistency.
☆ Parameterize Structure with Differentiable Template for 3D Shape Generation
Structural representation is crucial for reconstructing and generating editable 3D shapes with part semantics. Recent 3D shape generation works employ complicated networks and structure definitions relying on hierarchical annotations and pay less attention to the details inside parts. In this paper, we propose the method that parameterizes the shared structure in the same category using a differentiable template and corresponding fixed-length parameters. Specific parameters are fed into the template to calculate cuboids that indicate a concrete shape. We utilize the boundaries of three-view drawings of each cuboid to further describe the inside details. Shapes are represented with the parameters and three-view details inside cuboids, from which the SDF can be calculated to recover the object. Benefiting from our fixed-length parameters and three-view details, our networks for reconstruction and generation are simple and effective to learn the latent space. Our method can reconstruct or generate diverse shapes with complicated details, and interpolate them smoothly. Extensive evaluations demonstrate the superiority of our method on reconstruction from point cloud, generation, and interpolation.
☆ PIVOT-R: Primitive-Driven Waypoint-Aware World Model for Robotic Manipulation NeurIPS 2024
Language-guided robotic manipulation is a challenging task that requires an embodied agent to follow abstract user instructions to accomplish various complex manipulation tasks. Previous work trivially fitting the data without revealing the relation between instruction and low-level executable actions, these models are prone to memorizing the surficial pattern of the data instead of acquiring the transferable knowledge, and thus are fragile to dynamic environment changes. To address this issue, we propose a PrIrmitive-driVen waypOinT-aware world model for Robotic manipulation (PIVOT-R) that focuses solely on the prediction of task-relevant waypoints. Specifically, PIVOT-R consists of a Waypoint-aware World Model (WAWM) and a lightweight action prediction module. The former performs primitive action parsing and primitive-driven waypoint prediction, while the latter focuses on decoding low-level actions. Additionally, we also design an asynchronous hierarchical executor (AHE), which can use different execution frequencies for different modules of the model, thereby helping the model reduce computational redundancy and improve model execution efficiency. Our PIVOT-R outperforms state-of-the-art (SoTA) open-source models on the SeaWave benchmark, achieving an average relative improvement of 19.45% across four levels of instruction tasks. Moreover, compared to the synchronously executed PIVOT-R, the execution efficiency of PIVOT-R with AHE is increased by 28-fold, with only a 2.9% drop in performance. These results provide compelling evidence that our PIVOT-R can significantly improve both the performance and efficiency of robotic manipulation.
comment: Accepted to NeurIPS 2024
☆ Reverse Refinement Network for Narrow Rural Road Detection in High-Resolution Satellite Imagery
The automated extraction of rural roads is pivotal for rural development and transportation planning, serving as a cornerstone for socio-economic progress. Current research primarily focuses on road extraction in urban areas. However, rural roads present unique challenges due to their narrow and irregular nature, posing significant difficulties for road extraction. In this article, a reverse refinement network (R2-Net) is proposed to extract narrow rural roads, enhancing their connectivity and distinctiveness from the background. Specifically, to preserve the fine details of roads within high-resolution feature maps, R2-Net utilizes an axis context aware module (ACAM) to capture the long-distance spatial context information in various layers. Subsequently, the multi-level features are aggregated through a global aggregation module (GAM). Moreover, in the decoder stage, R2-Net employs a reverse-aware module (RAM) to direct the attention of the network to the complex background, thus amplifying its separability. In experiments, we compare R2-Net with several state-of-the-art methods using the DeepGlobe road extraction dataset and the WHU-RuR+ global large-scale rural road dataset. R2-Net achieved superior performance and especially excelled in accurately detecting narrow roads. Furthermore, we explored the applicability of R2-Net for large-scale rural road mapping. The results show that the proposed R2-Net has significant performance advantages for large-scale rural road mapping applications.
☆ V2M: Visual 2-Dimensional Mamba for Image Representation Learning
Mamba has garnered widespread attention due to its flexible design and efficient hardware performance to process 1D sequences based on the state space model (SSM). Recent studies have attempted to apply Mamba to the visual domain by flattening 2D images into patches and then regarding them as a 1D sequence. To compensate for the 2D structure information loss (e.g., local similarity) of the original image, most existing methods focus on designing different orders to sequentially process the tokens, which could only alleviate this issue to some extent. In this paper, we propose a Visual 2-Dimensional Mamba (V2M) model as a complete solution, which directly processes image tokens in the 2D space. We first generalize SSM to the 2-dimensional space which generates the next state considering two adjacent states on both dimensions (e.g., columns and rows). We then construct our V2M based on the 2-dimensional SSM formulation and incorporate Mamba to achieve hardware-efficient parallel processing. The proposed V2M effectively incorporates the 2D locality prior yet inherits the efficiency and input-dependent scalability of Mamba. Extensive experimental results on ImageNet classification and downstream visual tasks including object detection and instance segmentation on COCO and semantic segmentation on ADE20K demonstrate the effectiveness of our V2M compared with other visual backbones.
☆ Class Balancing Diversity Multimodal Ensemble for Alzheimer's Disease Diagnosis and Early Detection
Alzheimer's disease (AD) poses significant global health challenges due to its increasing prevalence and associated societal costs. Early detection and diagnosis of AD are critical for delaying progression and improving patient outcomes. Traditional diagnostic methods and single-modality data often fall short in identifying early-stage AD and distinguishing it from Mild Cognitive Impairment (MCI). This study addresses these challenges by introducing a novel approach: multImodal enseMble via class BALancing diversity for iMbalancEd Data (IMBALMED). IMBALMED integrates multimodal data from the Alzheimer's Disease Neuroimaging Initiative database, including clinical assessments, neuroimaging phenotypes, biospecimen and subject characteristics data. It employs an ensemble of model classifiers, each trained with different class balancing techniques, to overcome class imbalance and enhance model accuracy. We evaluate IMBALMED on two diagnostic tasks (binary and ternary classification) and four binary early detection tasks (at 12, 24, 36, and 48 months), comparing its performance with state-of-the-art algorithms and an unbalanced dataset method. IMBALMED demonstrates superior diagnostic accuracy and predictive performance in both binary and ternary classification tasks, significantly improving early detection of MCI at 48-month time point. The method shows improved classification performance and robustness, offering a promising solution for early detection and management of AD.
☆ Affinity-Graph-Guided Contractive Learning for Pretext-Free Medical Image Segmentation with Minimal Annotation
The combination of semi-supervised learning (SemiSL) and contrastive learning (CL) has been successful in medical image segmentation with limited annotations. However, these works often rely on pretext tasks that lack the specificity required for pixel-level segmentation, and still face overfitting issues due to insufficient supervision signals resulting from too few annotations. Therefore, this paper proposes an affinity-graph-guided semi-supervised contrastive learning framework (Semi-AGCL) by establishing additional affinity-graph-based supervision signals between the student and teacher network, to achieve medical image segmentation with minimal annotations without pretext. The framework first designs an average-patch-entropy-driven inter-patch sampling method, which can provide a robust initial feature space without relying on pretext tasks. Furthermore, the framework designs an affinity-graph-guided loss function, which can improve the quality of the learned representation and the model generalization ability by exploiting the inherent structure of the data, thus mitigating overfitting. Our experiments indicate that with merely 10% of the complete annotation set, our model approaches the accuracy of the fully annotated baseline, manifesting a marginal deviation of only 2.52%. Under the stringent conditions where only 5% of the annotations are employed, our model exhibits a significant enhancement in performance surpassing the second best baseline by 23.09% on the dice metric and achieving an improvement of 26.57% on the notably arduous CRAG and ACDC datasets.
comment: BIBM 2024
☆ FasterDiT: Towards Faster Diffusion Transformers Training without Architecture Modification NeurIPS 2024
Diffusion Transformers (DiT) have attracted significant attention in research. However, they suffer from a slow convergence rate. In this paper, we aim to accelerate DiT training without any architectural modification. We identify the following issues in the training process: firstly, certain training strategies do not consistently perform well across different data. Secondly, the effectiveness of supervision at specific timesteps is limited. In response, we propose the following contributions: (1) We introduce a new perspective for interpreting the failure of the strategies. Specifically, we slightly extend the definition of Signal-to-Noise Ratio (SNR) and suggest observing the Probability Density Function (PDF) of SNR to understand the essence of the data robustness of the strategy. (2) We conduct numerous experiments and report over one hundred experimental results to empirically summarize a unified accelerating strategy from the perspective of PDF. (3) We develop a new supervision method that further accelerates the training process of DiT. Based on them, we propose FasterDiT, an exceedingly simple and practicable design strategy. With few lines of code modifications, it achieves 2.30 FID on ImageNet 256 resolution at 1000k iterations, which is comparable to DiT (2.27 FID) but 7 times faster in training.
comment: NeurIPS 2024 (poster)
☆ Pubic Symphysis-Fetal Head Segmentation Network Using BiFormer Attention Mechanism and Multipath Dilated Convolution
Pubic symphysis-fetal head segmentation in transperineal ultrasound images plays a critical role for the assessment of fetal head descent and progression. Existing transformer \iffalse-based\fi segmentation methods based on sparse attention mechanism use handcrafted static patterns, which leads to great differences \iffalse in \fi in terms of segmentation performance on specific datasets. To address this issue, we introduce a dynamic, query-aware sparse attention mechanism for ultrasound image segmentation. Specifically, we propose a novel method, named BRAU-Net to solve the pubic symphysis-fetal head segmentation task in this paper. The method adopts a U-Net-like encoder-decoder architecture with bi-level routing attention and skip connections, which effectively learns local-global semantic information. In addition, we propose an inverted bottleneck patch expanding (IBPE) module to reduce information loss while performing up-sampling operations. The proposed BRAU-Net is evaluated on FH-PS-AoP and HC18 datasets. The results demonstrate that our method could achieve excellent segmentation results. The code is available on GitHub.
comment: MMM2025;Camera-ready Version;The code is available at https://github.com/Caipengzhou/BRAU-Net
☆ On Representation of 3D Rotation in the Context of Deep Learning ICCV
This paper investigates various methods of representing 3D rotations and their impact on the learning process of deep neural networks. We evaluated the performance of ResNet18 networks for 3D rotation estimation using several rotation representations and loss functions on both synthetic and real data. The real datasets contained 3D scans of industrial bins, while the synthetic datasets included views of a simple asymmetric object rendered under different rotations. On synthetic data, we also assessed the effects of different rotation distributions within the training and test sets, as well as the impact of the object's texture. In line with previous research, we found that networks using the continuous 5D and 6D representations performed better than the discontinuous ones.
comment: Accepted at International Conference on Computer Vision and Graphics ICCVG 2024. The proceedings of the conference will be published in Lecture Notes in Networks and Systems (LNNS), Springer
☆ Anatomical feature-prioritized loss for enhanced MR to CT translation
In medical image synthesis, the precision of localized structural details is crucial, particularly when addressing specific clinical requirements such as the identification and measurement of fine structures. Traditional methods for image translation and synthesis are generally optimized for global image reconstruction but often fall short in providing the finesse required for detailed local analysis. This study represents a step toward addressing this challenge by introducing a novel anatomical feature-prioritized (AFP) loss function into the synthesis process. This method enhances reconstruction by focusing on clinically significant structures, utilizing features from a pre-trained model designed for a specific downstream task, such as the segmentation of particular anatomical regions. The AFP loss function can replace or complement global reconstruction methods, ensuring a balanced emphasis on both global image fidelity and local structural details. Various implementations of this loss function are explored, including its integration into different synthesis networks such as GAN-based and CNN-based models. Our approach is applied and evaluated in two contexts: lung MR to CT translation, focusing on high-quality reconstruction of bronchial structures, using a private dataset; and pelvis MR to CT synthesis, targeting the accurate representation of organs and muscles, utilizing a public dataset from the Synthrad2023 challenge. We leverage embeddings from pre-trained segmentation models specific to these anatomical regions to demonstrate the capability of the AFP loss to prioritize and accurately reconstruct essential features. This tailored approach shows promising potential for enhancing the specificity and practicality of medical image synthesis in clinical applications.
☆ Spatial-Aware Efficient Projector for MLLMs via Multi-Layer Feature Aggregation
The projector plays a crucial role in multi-modal language models (MLLMs). The number of visual tokens it outputs affects the efficiency of the MLLM, while the quality of the visual tokens influences the visual understanding capabilities of the MLLM. Current explorations on the projector focus on reducing the number of visual tokens to improve efficiency, often overlooking the inherent spatial discrepancy between the serialized 2-dimensional visual token sequences and natural language token sequences. A Spatial-Aware Efficient Projector (SAEP) is proposed to address this issue. In detail, our SAEP method employs an modified separable depthwise convolution module on multi-layer visual features to enhance the spatial information of visual tokens. As a result, our SAEP method can not only largely reduce the number of visual tokens by 75\%, but also significantly improve the multimodal spatial understanding capability of MLLMs. Moreover, compared to existing projectors, our SAEP gets best performances on massive multimodal evaluation benchmarks, which denotes its effectiveness on bridging the modality gap.
comment: 10 pages, 3 figures
☆ QIANets: Quantum-Integrated Adaptive Networks for Reduced Latency and Improved Inference Times in CNN Models NeurIPS 2024
Convolutional neural networks (CNNs) have made significant advances in computer vision tasks, yet their high inference times and latency often limit real-world applicability. While model compression techniques have gained popularity as solutions, they often overlook the critical balance between low latency and uncompromised accuracy. By harnessing quantum-inspired pruning, tensor decomposition, and annealing-based matrix factorization - three quantum-inspired concepts - we introduce QIANets: a novel approach of redesigning the traditional GoogLeNet, DenseNet, and ResNet-18 model architectures to process more parameters and computations whilst maintaining low inference times. Despite experimental limitations, the method was tested and evaluated, demonstrating reductions in inference times, along with effective accuracy preservations.
comment: Accepted to NeurIPS 2024 workshop on Neural Compression
☆ GlobalMamba: Global Image Serialization for Vision Mamba
Vision mambas have demonstrated strong performance with linear complexity to the number of vision tokens. Their efficiency results from processing image tokens sequentially. However, most existing methods employ patch-based image tokenization and then flatten them into 1D sequences for causal processing, which ignore the intrinsic 2D structural correlations of images. It is also difficult to extract global information by sequential processing of local patches. In this paper, we propose a global image serialization method to transform the image into a sequence of causal tokens, which contain global information of the 2D image. We first convert the image from the spatial domain to the frequency domain using Discrete Cosine Transform (DCT) and then arrange the pixels with corresponding frequency ranges. We further transform each set within the same frequency band back to the spatial domain to obtain a series of images before tokenization. We construct a vision mamba model, GlobalMamba, with a causal input format based on the proposed global image serialization, which can better exploit the causal relations among image sequences. Extensive experiments demonstrate the effectiveness of our GlobalMamba, including image classification on ImageNet-1K, object detection on COCO, and semantic segmentation on ADE20K.
☆ LG-CAV: Train Any Concept Activation Vector with Language Guidance
Concept activation vector (CAV) has attracted broad research interest in explainable AI, by elegantly attributing model predictions to specific concepts. However, the training of CAV often necessitates a large number of high-quality images, which are expensive to curate and thus limited to a predefined set of concepts. To address this issue, we propose Language-Guided CAV (LG-CAV) to harness the abundant concept knowledge within the certain pre-trained vision-language models (e.g., CLIP). This method allows training any CAV without labeled data, by utilizing the corresponding concept descriptions as guidance. To bridge the gap between vision-language model and the target model, we calculate the activation values of concept descriptions on a common pool of images (probe images) with vision-language model and utilize them as language guidance to train the LG-CAV. Furthermore, after training high-quality LG-CAVs related to all the predicted classes in the target model, we propose the activation sample reweighting (ASR), serving as a model correction technique, to improve the performance of the target model in return. Experiments on four datasets across nine architectures demonstrate that LG-CAV achieves significantly superior quality to previous CAV methods given any concept, and our model correction method achieves state-of-the-art performance compared to existing concept-based methods. Our code is available at https://github.com/hqhQAQ/LG-CAV.
☆ Animate-X: Universal Character Image Animation with Enhanced Motion Representation
Character image animation, which generates high-quality videos from a reference image and target pose sequence, has seen significant progress in recent years. However, most existing methods only apply to human figures, which usually do not generalize well on anthropomorphic characters commonly used in industries like gaming and entertainment. Our in-depth analysis suggests to attribute this limitation to their insufficient modeling of motion, which is unable to comprehend the movement pattern of the driving video, thus imposing a pose sequence rigidly onto the target character. To this end, this paper proposes Animate-X, a universal animation framework based on LDM for various character types (collectively named X), including anthropomorphic characters. To enhance motion representation, we introduce the Pose Indicator, which captures comprehensive motion pattern from the driving video through both implicit and explicit manner. The former leverages CLIP visual features of a driving video to extract its gist of motion, like the overall movement pattern and temporal relations among motions, while the latter strengthens the generalization of LDM by simulating possible inputs in advance that may arise during inference. Moreover, we introduce a new Animated Anthropomorphic Benchmark (A^2Bench) to evaluate the performance of Animate-X on universal and widely applicable animation images. Extensive experiments demonstrate the superiority and effectiveness of Animate-X compared to state-of-the-art methods.
comment: 25 pages, 15 figures, conference
☆ ROA-BEV: 2D Region-Oriented Attention for BEV-based 3D Object
Vision-based BEV (Bird-Eye-View) 3D object detection has recently become popular in autonomous driving. However, objects with a high similarity to the background from a camera perspective cannot be detected well by existing methods. In this paper, we propose 2D Region-oriented Attention for a BEV-based 3D Object Detection Network (ROA-BEV), which can make the backbone focus more on feature learning in areas where objects may exist. Moreover, our method increases the information content of ROA through a multi-scale structure. In addition, every block of ROA utilizes a large kernel to ensure that the receptive field is large enough to catch large objects' information. Experiments on nuScenes show that ROA-BEV improves the performance based on BEVDet and BEVDepth. The code will be released soon.
☆ A Consistency-Aware Spot-Guided Transformer for Versatile and Hierarchical Point Cloud Registration NeurIPS 2024
Deep learning-based feature matching has shown great superiority for point cloud registration in the absence of pose priors. Although coarse-to-fine matching approaches are prevalent, the coarse matching of existing methods is typically sparse and loose without consideration of geometric consistency, which makes the subsequent fine matching rely on ineffective optimal transport and hypothesis-and-selection methods for consistency. Therefore, these methods are neither efficient nor scalable for real-time applications such as odometry in robotics. To address these issues, we design a consistency-aware spot-guided Transformer (CAST), which incorporates a spot-guided cross-attention module to avoid interfering with irrelevant areas, and a consistency-aware self-attention module to enhance matching capabilities with geometrically consistent correspondences. Furthermore, a lightweight fine matching module for both sparse keypoints and dense features can estimate the transformation accurately. Extensive experiments on both outdoor LiDAR point cloud datasets and indoor RGBD point cloud datasets demonstrate that our method achieves state-of-the-art accuracy, efficiency, and robustness.
comment: Accepted by NeurIPS 2024 as poster
☆ Fine-grained Abnormality Prompt Learning for Zero-shot Anomaly Detection
Current zero-shot anomaly detection (ZSAD) methods show remarkable success in prompting large pre-trained vision-language models to detect anomalies in a target dataset without using any dataset-specific training or demonstration. However, these methods are often focused on crafting/learning prompts that capture only coarse-grained semantics of abnormality, e.g., high-level semantics like "damaged", "imperfect", or "defective" on carpet. They therefore have limited capability in recognizing diverse abnormality details with distinctive visual appearance, e.g., specific defect types like color stains, cuts, holes, and threads on carpet. To address this limitation, we propose FAPrompt, a novel framework designed to learn Fine-grained Abnormality Prompts for more accurate ZSAD. To this end, we introduce a novel compound abnormality prompting module in FAPrompt to learn a set of complementary, decomposed abnormality prompts, where each abnormality prompt is formed by a compound of shared normal tokens and a few learnable abnormal tokens. On the other hand, the fine-grained abnormality patterns can be very different from one dataset to another. To enhance their cross-dataset generalization, we further introduce a data-dependent abnormality prior module that learns to derive abnormality features from each query/test image as a sample-wise abnormality prior to ground the abnormality prompts in a given target dataset. Comprehensive experiments conducted across 19 real-world datasets, covering both industrial defects and medical anomalies, demonstrate that FAPrompt substantially outperforms state-of-the-art methods by at least 3%-5% AUC/AP in both image- and pixel-level ZSAD tasks. Code is available at https://github.com/mala-lab/FAPrompt.
comment: 27 pages, 19 figures
☆ Manifold-Aware Local Feature Modeling for Semi-Supervised Medical Image Segmentation
Achieving precise medical image segmentation is vital for effective treatment planning and accurate disease diagnosis. Traditional fully-supervised deep learning methods, though highly precise, are heavily reliant on large volumes of labeled data, which are often difficult to obtain due to the expertise required for medical annotations. This has led to the rise of semi-supervised learning approaches that utilize both labeled and unlabeled data to mitigate the label scarcity issue. In this paper, we introduce the Manifold-Aware Local Feature Modeling Network (MANet), which enhances the U-Net architecture by incorporating manifold supervision signals. This approach focuses on improving boundary accuracy, which is crucial for reliable medical diagnosis. To further extend the versatility of our method, we propose two variants: MA-Sobel and MA-Canny. The MA-Sobel variant employs the Sobel operator, which is effective for both 2D and 3D data, while the MA-Canny variant utilizes the Canny operator, specifically designed for 2D images, to refine boundary detection. These variants allow our method to adapt to various medical image modalities and dimensionalities, ensuring broader applicability. Our extensive experiments on datasets such as ACDC, LA, and Pancreas-NIH demonstrate that MANet consistently surpasses state-of-the-art methods in performance metrics like Dice and Jaccard scores. The proposed method also shows improved generalization across various semi-supervised segmentation networks, highlighting its robustness and effectiveness. Visual analysis of segmentation results confirms that MANet offers clearer and more accurate class boundaries, underscoring the value of manifold information in medical image segmentation.
comment: 11 pages
☆ Exploring Semi-Supervised Learning for Online Mapping
Online mapping is important for scaling autonomous driving beyond well-defined areas. Training a model to produce a local map, including lane markers, road edges, and pedestrian crossings using only onboard sensory information, traditionally requires extensive labelled data, which is difficult and costly to obtain. This paper draws inspiration from semi-supervised learning techniques in other domains, demonstrating their applicability to online mapping. Additionally, we propose a simple yet effective method to exploit inherent attributes of online mapping to further enhance performance by fusing the teacher's pseudo-labels from multiple samples. The performance gap to using all labels is reduced from 29.6 to 3.4 mIoU on Argoverse, and from 12 to 3.4 mIoU on NuScenes utilising only 10% of the labelled data. We also demonstrate strong performance in extrapolating to new cities outside those in the training data. Specifically, for challenging nuScenes, adapting from Boston to Singapore, performance increases by 6.6 mIoU when unlabelled data from Singapore is included in training.
☆ Two-Stage Approach for Brain MR Image Synthesis: 2D Image Synthesis and 3D Refinement MICCAI 2024
Despite significant advancements in automatic brain tumor segmentation methods, their performance is not guaranteed when certain MR sequences are missing. Addressing this issue, it is crucial to synthesize the missing MR images that reflect the unique characteristics of the absent modality with precise tumor representation. Typically, MRI synthesis methods generate partial images rather than full-sized volumes due to computational constraints. This limitation can lead to a lack of comprehensive 3D volumetric information and result in image artifacts during the merging process. In this paper, we propose a two-stage approach that first synthesizes MR images from 2D slices using a novel intensity encoding method and then refines the synthesized MRI. The proposed intensity encoding reduces artifacts when synthesizing MRI on a 2D slice basis. Then, the \textit{Refiner}, which leverages complete 3D volume information, further improves the quality of the synthesized images and enhances their applicability to segmentation methods. Experimental results demonstrate that the intensity encoding effectively minimizes artifacts in the synthesized MRI and improves perceptual quality. Furthermore, using the \textit{Refiner} on synthesized MRI significantly improves brain tumor segmentation results, highlighting the potential of our approach in practical applications.
comment: MICCAI 2024 BraSyn Challenge 1st place
☆ big.LITTLE Vision Transformer for Efficient Visual Recognition
In this paper, we introduce the big.LITTLE Vision Transformer, an innovative architecture aimed at achieving efficient visual recognition. This dual-transformer system is composed of two distinct blocks: the big performance block, characterized by its high capacity and substantial computational demands, and the LITTLE efficiency block, designed for speed with lower capacity. The key innovation of our approach lies in its dynamic inference mechanism. When processing an image, our system determines the importance of each token and allocates them accordingly: essential tokens are processed by the high-performance big model, while less critical tokens are handled by the more efficient little model. This selective processing significantly reduces computational load without sacrificing the overall performance of the model, as it ensures that detailed analysis is reserved for the most important information. To validate the effectiveness of our big.LITTLE Vision Transformer, we conducted comprehensive experiments on image classification and segment anything task. Our results demonstrate that the big.LITTLE architecture not only maintains high accuracy but also achieves substantial computational savings. Specifically, our approach enables the efficient handling of large-scale visual recognition tasks by dynamically balancing the trade-offs between performance and efficiency. The success of our method underscores the potential of hybrid models in optimizing both computation and performance in visual recognition tasks, paving the way for more practical and scalable deployment of advanced neural networks in real-world applications.
☆ Slide-based Graph Collaborative Training for Histopathology Whole Slide Image Analysis
The development of computational pathology lies in the consensus that pathological characteristics of tumors are significant guidance for cancer diagnostics. Most existing research focuses on the inner-contextual information within each WSI yet ignores the possible inter-correlations between slides. As the development of tumors is a continuous process involving a series of histological, morphological, and genetic changes that accumulate over time, the similarities and differences between WSIs across various stages, grades, locations and patients should potentially contribute to the representation of WSIs and deserve to be taken into account in WSI modeling. To verify the advancement of introducing the slide inter-correlations into the representation learning of WSIs, we proposed a generic WSI analysis pipeline SlideGCD that can be adapted to any existing Multiple Instance Learning (MIL) frameworks and improve their performance. With the new paradigm, the prior knowledge of cancer development can participate in the end-to-end workflow, which concurrently initializes and refines the slide representation, as a guide for message passing in the slide-based graph. Extensive comparisons and experiments are conducted to validate the effectiveness and robustness of the proposed pipeline across 4 different tasks, including cancer subtyping, cancer staging, survival prediction, and gene mutation prediction, with 7 representative SOTA WSI analysis frameworks as backbones.
☆ Saliency Guided Optimization of Diffusion Latents
With the rapid advances in diffusion models, generating decent images from text prompts is no longer challenging. The key to text-to-image generation is how to optimize the results of a text-to-image generation model so that they can be better aligned with human intentions or prompts. Existing optimization methods commonly treat the entire image uniformly and conduct global optimization. These methods overlook the fact that when viewing an image, the human visual system naturally prioritizes attention toward salient areas, often neglecting less or non-salient regions. That is, humans are likely to neglect optimizations in non-salient areas. Consequently, although model retaining is conducted under the guidance of additional large and multimodality models, existing methods, which perform uniform optimizations, yield sub-optimal results. To address this alignment challenge effectively and efficiently, we propose Saliency Guided Optimization Of Diffusion Latents (SGOOL). We first employ a saliency detector to mimic the human visual attention system and mark out the salient regions. To avoid retraining an additional model, our method directly optimizes the diffusion latents. Besides, SGOOL utilizes an invertible diffusion process and endows it with the merits of constant memory implementation. Hence, our method becomes a parameter-efficient and plug-and-play fine-tuning method. Extensive experiments have been done with several metrics and human evaluation. Experimental results demonstrate the superiority of SGOOL in image quality and prompt alignment.
☆ Automated extraction of 4D aircraft trajectories from video recordings
The Bureau d'Enqu{\^e}tes et d'Analyses pour la S{\'e}curit{\'e} de l'Aviation Civile (BEA) has to analyze accident videos from on-board or ground cameras involving all types of aircraft. Until now, this analysis has been manual and time-consuming. The aim of this study is to identify the applications of photogrammetry and to automate the extraction of 4D trajectories from these videos. Taking into account all potential flight configurations, photogrammetric algorithms are being developed on the basis of IGN's MicMac software and tested in the field. The results of these automated processes are intended to replace flight data from recorders such as FDRs or CVRs, which are sometimes missing. The information of interest to the BEA includes: three-dimensional position with the associated time component, the orientations of the aircraft's three axes (pitch, roll and yaw navigation angles) and average speeds (including rate of climb).
comment: in French language, CFPT-RFIAP 2018, SFPT (Soci\'et\'e Fran\c{c}aise de Photogramm\'etrie et de T\'el\'ed\'etection); RFIAP (Reconnaissance des Formes, Image, Apprentissage et Perception), Jun 2018, Champs sur Marne - Marne la Vall\'ee, France
☆ LOBG:Less Overfitting for Better Generalization in Vision-Language Model
Existing prompt learning methods in Vision-Language Models (VLM) have effectively enhanced the transfer capability of VLM to downstream tasks, but they suffer from a significant decline in generalization due to severe overfitting. To address this issue, we propose a framework named LOBG for vision-language models. Specifically, we use CLIP to filter out fine-grained foreground information that might cause overfitting, thereby guiding prompts with basic visual concepts. To further mitigate overfitting, we devel oped a structural topology preservation (STP) loss at the feature level, which endows the feature space with overall plasticity, allowing effective reshaping of the feature space during optimization. Additionally, we employed hierarchical logit distilation (HLD) at the output level to constrain outputs, complementing STP at the output end. Extensive experimental results demonstrate that our method significantly improves generalization capability and alleviates overfitting compared to state-of-the-art approaches.
☆ Capture Artifacts via Progressive Disentangling and Purifying Blended Identities for Deepfake Detection
The Deepfake technology has raised serious concerns regarding privacy breaches and trust issues. To tackle these challenges, Deepfake detection technology has emerged. Current methods over-rely on the global feature space, which contains redundant information independent of the artifacts. As a result, existing Deepfake detection techniques suffer performance degradation when encountering unknown datasets. To reduce information redundancy, the current methods use disentanglement techniques to roughly separate the fake faces into artifacts and content information. However, these methods lack a solid disentanglement foundation and cannot guarantee the reliability of their disentangling process. To address these issues, a Deepfake detection method based on progressive disentangling and purifying blended identities is innovatively proposed in this paper. Based on the artifact generation mechanism, the coarse- and fine-grained strategies are combined to ensure the reliability of the disentanglement method. Our method aims to more accurately capture and separate artifact features in fake faces. Specifically, we first perform the coarse-grained disentangling on fake faces to obtain a pair of blended identities that require no additional annotation to distinguish between source face and target face. Then, the artifact features from each identity are separated to achieve fine-grained disentanglement. To obtain pure identity information and artifacts, an Identity-Artifact Correlation Compression module (IACC) is designed based on the information bottleneck theory, effectively reducing the potential correlation between identity information and artifacts. Additionally, an Identity-Artifact Separation Contrast Loss is designed to enhance the independence of artifact features post-disentangling. Finally, the classifier only focuses on pure artifact features to achieve a generalized Deepfake detector.
☆ ForgeryGPT: Multimodal Large Language Model For Explainable Image Forgery Detection and Localization
Multimodal Large Language Models (MLLMs), such as GPT4o, have shown strong capabilities in visual reasoning and explanation generation. However, despite these strengths, they face significant challenges in the increasingly critical task of Image Forgery Detection and Localization (IFDL). Moreover, existing IFDL methods are typically limited to the learning of low-level semantic-agnostic clues and merely provide a single outcome judgment. To tackle these issues, we propose ForgeryGPT, a novel framework that advances the IFDL task by capturing high-order forensics knowledge correlations of forged images from diverse linguistic feature spaces, while enabling explainable generation and interactive dialogue through a newly customized Large Language Model (LLM) architecture. Specifically, ForgeryGPT enhances traditional LLMs by integrating the Mask-Aware Forgery Extractor, which enables the excavating of precise forgery mask information from input images and facilitating pixel-level understanding of tampering artifacts. The Mask-Aware Forgery Extractor consists of a Forgery Localization Expert (FL-Expert) and a Mask Encoder, where the FL-Expert is augmented with an Object-agnostic Forgery Prompt and a Vocabulary-enhanced Vision Encoder, allowing for effectively capturing of multi-scale fine-grained forgery details. To enhance its performance, we implement a three-stage training strategy, supported by our designed Mask-Text Alignment and IFDL Task-Specific Instruction Tuning datasets, which align vision-language modalities and improve forgery detection and instruction-following capabilities. Extensive experiments demonstrate the effectiveness of the proposed method.
comment: 16 pages, 14 figures
☆ LADMIM: Logical Anomaly Detection with Masked Image Modeling in Discrete Latent Space
Detecting anomalies such as incorrect combinations of objects or deviations in their positions is a challenging problem in industrial anomaly detection. Traditional methods mainly focus on local features of normal images, such as scratches and dirt, making detecting anomalies in the relationships between features difficult. Masked image modeling(MIM) is a self-supervised learning technique that predicts the feature representation of masked regions in an image. To reconstruct the masked regions, it is necessary to understand how the image is composed, allowing the learning of relationships between features within the image. We propose a novel approach that leverages the characteristics of MIM to detect logical anomalies effectively. To address blurriness in the reconstructed image, we replace pixel prediction with predicting the probability distribution of discrete latent variables of the masked regions using a tokenizer. We evaluated the proposed method on the MVTecLOCO dataset, achieving an average AUC of 0.867, surpassing traditional reconstruction-based and distillation-based methods.
comment: Under Review
☆ KNN Transformer with Pyramid Prompts for Few-Shot Learning
Few-Shot Learning (FSL) aims to recognize new classes with limited labeled data. Recent studies have attempted to address the challenge of rare samples with textual prompts to modulate visual features. However, they usually struggle to capture complex semantic relationships between textual and visual features. Moreover, vanilla self-attention is heavily affected by useless information in images, severely constraining the potential of semantic priors in FSL due to the confusion of numerous irrelevant tokens during interaction. To address these aforementioned issues, a K-NN Transformer with Pyramid Prompts (KTPP) is proposed to select discriminative information with K-NN Context Attention (KCA) and adaptively modulate visual features with Pyramid Cross-modal Prompts (PCP). First, for each token, the KCA only selects the K most relevant tokens to compute the self-attention matrix and incorporates the mean of all tokens as the context prompt to provide the global context in three cascaded stages. As a result, irrelevant tokens can be progressively suppressed. Secondly, pyramid prompts are introduced in the PCP to emphasize visual features via interactions between text-based class-aware prompts and multi-scale visual features. This allows the ViT to dynamically adjust the importance weights of visual features based on rich semantic information at different scales, making models robust to spatial variations. Finally, augmented visual features and class-aware prompts are interacted via the KCA to extract class-specific features. Consequently, our model further enhances noise-free visual representations via deep cross-modal interactions, extracting generalized visual representation in scenarios with few labeled samples. Extensive experiments on four benchmark datasets demonstrate the effectiveness of our method.
comment: 10 pages, 5 figures, accepted by ACM Multimedia 2024
☆ Detecting Unforeseen Data Properties with Diffusion Autoencoder Embeddings using Spine MRI data MICCAI 2024
Deep learning has made significant strides in medical imaging, leveraging the use of large datasets to improve diagnostics and prognostics. However, large datasets often come with inherent errors through subject selection and acquisition. In this paper, we investigate the use of Diffusion Autoencoder (DAE) embeddings for uncovering and understanding data characteristics and biases, including biases for protected variables like sex and data abnormalities indicative of unwanted protocol variations. We use sagittal T2-weighted magnetic resonance (MR) images of the neck, chest, and lumbar region from 11186 German National Cohort (NAKO) participants. We compare DAE embeddings with existing generative models like StyleGAN and Variational Autoencoder. Evaluations on a large-scale dataset consisting of sagittal T2-weighted MR images of three spine regions show that DAE embeddings effectively separate protected variables such as sex and age. Furthermore, we used t-SNE visualization to identify unwanted variations in imaging protocols, revealing differences in head positioning. Our embedding can identify samples where a sex predictor will have issues learning the correct sex. Our findings highlight the potential of using advanced embedding techniques like DAEs to detect data quality issues and biases in medical imaging datasets. Identifying such hidden relations can enhance the reliability and fairness of deep learning models in healthcare applications, ultimately improving patient care and outcomes.
comment: This paper was accepted in the "Workshop on Interpretability of Machine Intelligence in Medical Image Computing" (iMIMIC) at MICCAI 2024
☆ MagicEraser: Erasing Any Objects via Semantics-Aware Control ECCV 2024
The traditional image inpainting task aims to restore corrupted regions by referencing surrounding background and foreground. However, the object erasure task, which is in increasing demand, aims to erase objects and generate harmonious background. Previous GAN-based inpainting methods struggle with intricate texture generation. Emerging diffusion model-based algorithms, such as Stable Diffusion Inpainting, exhibit the capability to generate novel content, but they often produce incongruent results at the locations of the erased objects and require high-quality text prompt inputs. To address these challenges, we introduce MagicEraser, a diffusion model-based framework tailored for the object erasure task. It consists of two phases: content initialization and controllable generation. In the latter phase, we develop two plug-and-play modules called prompt tuning and semantics-aware attention refocus. Additionally, we propose a data construction strategy that generates training data specially suitable for this task. MagicEraser achieves fine and effective control of content generation while mitigating undesired artifacts. Experimental results highlight a valuable advancement of our approach in the object erasure task.
comment: Accepted by ECCV 2024
☆ Eliminating the Language Bias for Visual Question Answering with fine-grained Causal Intervention
Despite the remarkable advancements in Visual Question Answering (VQA), the challenge of mitigating the language bias introduced by textual information remains unresolved. Previous approaches capture language bias from a coarse-grained perspective. However, the finer-grained information within a sentence, such as context and keywords, can result in different biases. Due to the ignorance of fine-grained information, most existing methods fail to sufficiently capture language bias. In this paper, we propose a novel causal intervention training scheme named CIBi to eliminate language bias from a finer-grained perspective. Specifically, we divide the language bias into context bias and keyword bias. We employ causal intervention and contrastive learning to eliminate context bias and improve the multi-modal representation. Additionally, we design a new question-only branch based on counterfactual generation to distill and eliminate keyword bias. Experimental results illustrate that CIBi is applicable to various VQA models, yielding competitive performance.
☆ Identity-Focused Inference and Extraction Attacks on Diffusion Models
The increasing reliance on diffusion models for generating synthetic images has amplified concerns about the unauthorized use of personal data, particularly facial images, in model training. In this paper, we introduce a novel identity inference framework to hold model owners accountable for including individuals' identities in their training data. Our approach moves beyond traditional membership inference attacks by focusing on identity-level inference, providing a new perspective on data privacy violations. Through comprehensive evaluations on two facial image datasets, Labeled Faces in the Wild (LFW) and CelebA, our experiments demonstrate that the proposed membership inference attack surpasses baseline methods, achieving an attack success rate of up to 89% and an AUC-ROC of 0.91, while the identity inference attack attains 92% on LDM models trained on LFW, and the data extraction attack achieves 91.6% accuracy on DDPMs, validating the effectiveness of our approach across diffusion models.
comment: 5 figures, 3 tables,12 pages main body content
☆ Generative Human Video Compression with Multi-granularity Temporal Trajectory Factorization
In this paper, we propose a novel Multi-granularity Temporal Trajectory Factorization framework for generative human video compression, which holds great potential for bandwidth-constrained human-centric video communication. In particular, the proposed motion factorization strategy can facilitate to implicitly characterize the high-dimensional visual signal into compact motion vectors for representation compactness and further transform these vectors into a fine-grained field for motion expressibility. As such, the coded bit-stream can be entailed with enough visual motion information at the lowest representation cost. Meanwhile, a resolution-expandable generative module is developed with enhanced background stability, such that the proposed framework can be optimized towards higher reconstruction robustness and more flexible resolution adaptation. Experimental results show that proposed method outperforms latest generative models and the state-of-the-art video coding standard Versatile Video Coding (VVC) on both talking-face videos and moving-body videos in terms of both objective and subjective quality. The project page can be found at https://github.com/xyzysz/Extreme-Human-Video-Compression-with-MTTF.
comment: Submitted to TCSVT
☆ First Creating Backgrounds Then Rendering Texts: A New Paradigm for Visual Text Blending ECAI2024
Diffusion models, known for their impressive image generation abilities, have played a pivotal role in the rise of visual text generation. Nevertheless, existing visual text generation methods often focus on generating entire images with text prompts, leading to imprecise control and limited practicality. A more promising direction is visual text blending, which focuses on seamlessly merging texts onto text-free backgrounds. However, existing visual text blending methods often struggle to generate high-fidelity and diverse images due to a shortage of backgrounds for synthesis and limited generalization capabilities. To overcome these challenges, we propose a new visual text blending paradigm including both creating backgrounds and rendering texts. Specifically, a background generator is developed to produce high-fidelity and text-free natural images. Moreover, a text renderer named GlyphOnly is designed for achieving visually plausible text-background integration. GlyphOnly, built on a Stable Diffusion framework, utilizes glyphs and backgrounds as conditions for accurate rendering and consistency control, as well as equipped with an adaptive text block exploration strategy for small-scale text rendering. We also explore several downstream applications based on our method, including scene text dataset synthesis for boosting scene text detectors, as well as text image customization and editing. Code and model will be available at \url{https://github.com/Zhenhang-Li/GlyphOnly}.
comment: Accepted to ECAI2024
☆ X-Fi: A Modality-Invariant Foundation Model for Multimodal Human Sensing
Human sensing, which employs various sensors and advanced deep learning technologies to accurately capture and interpret human body information, has significantly impacted fields like public security and robotics. However, current human sensing primarily depends on modalities such as cameras and LiDAR, each of which has its own strengths and limitations. Furthermore, existing multi-modal fusion solutions are typically designed for fixed modality combinations, requiring extensive retraining when modalities are added or removed for diverse scenarios. In this paper, we propose a modality-invariant foundation model for all modalities, X-Fi, to address this issue. X-Fi enables the independent or combinatory use of sensor modalities without additional training by utilizing a transformer structure to accommodate variable input sizes and incorporating a novel "X-fusion" mechanism to preserve modality-specific features during multimodal integration. This approach not only enhances adaptability but also facilitates the learning of complementary features across modalities. Extensive experiments conducted on the MM-Fi and XRF55 datasets, employing six distinct modalities, demonstrate that X-Fi achieves state-of-the-art performance in human pose estimation (HPE) and human activity recognition (HAR) tasks. The findings indicate that our proposed model can efficiently support a wide range of human sensing applications, ultimately contributing to the evolution of scalable, multimodal sensing technologies.
☆ Will the Inclusion of Generated Data Amplify Bias Across Generations in Future Image Classification Models?
As the demand for high-quality training data escalates, researchers have increasingly turned to generative models to create synthetic data, addressing data scarcity and enabling continuous model improvement. However, reliance on self-generated data introduces a critical question: Will this practice amplify bias in future models? While most research has focused on overall performance, the impact on model bias, particularly subgroup bias, remains underexplored. In this work, we investigate the effects of the generated data on image classification tasks, with a specific focus on bias. We develop a practical simulation environment that integrates a self-consuming loop, where the generative model and classification model are trained synergistically. Hundreds of experiments are conducted on Colorized MNIST, CIFAR-20/100, and Hard ImageNet datasets to reveal changes in fairness metrics across generations. In addition, we provide a conjecture to explain the bias dynamics when training models on continuously augmented datasets across generations. Our findings contribute to the ongoing debate on the implications of synthetic data for fairness in real-world applications.
comment: 15 pages, 7 figures
☆ Fast and Accurate Neural Rendering Using Semi-Gradients
We propose a simple yet effective neural network-based framework for global illumination rendering. Recently, rendering techniques that learn neural radiance caches by minimizing the difference (i.e., residual) between the left and right sides of the rendering equation have been suggested. Due to their ease of implementation and the advantage of excluding path integral calculations, these techniques have been applied to various fields, such as free-viewpoint rendering, differentiable rendering, and real-time rendering. However, issues of slow training and occasionally darkened renders have been noted. We identify the cause of these issues as the bias and high variance present in the gradient estimates of the existing residual-based objective function. To address this, we introduce a new objective function that maintains the same global optimum as before but allows for unbiased and low-variance gradient estimates, enabling faster and more accurate training of neural networks. In conclusion, this method is simply implemented by ignoring the partial derivatives of the right-hand side, and theoretical and experimental analyses demonstrate the effectiveness of the proposed loss.
☆ Performance Evaluation of Deep Learning and Transformer Models Using Multimodal Data for Breast Cancer Classification MICCAI 2024
Rising breast cancer (BC) occurrence and mortality are major global concerns for women. Deep learning (DL) has demonstrated superior diagnostic performance in BC classification compared to human expert readers. However, the predominant use of unimodal (digital mammography) features may limit the current performance of diagnostic models. To address this, we collected a novel multimodal dataset comprising both imaging and textual data. This study proposes a multimodal DL architecture for BC classification, utilising images (mammograms; four views) and textual data (radiological reports) from our new in-house dataset. Various augmentation techniques were applied to enhance the training data size for both imaging and textual data. We explored the performance of eleven SOTA DL architectures (VGG16, VGG19, ResNet34, ResNet50, MobileNet-v3, EffNet-b0, EffNet-b1, EffNet-b2, EffNet-b3, EffNet-b7, and Vision Transformer (ViT)) as imaging feature extractors. For textual feature extraction, we utilised either artificial neural networks (ANNs) or long short-term memory (LSTM) networks. The combined imaging and textual features were then inputted into an ANN classifier for BC classification, using the late fusion technique. We evaluated different feature extractor and classifier arrangements. The VGG19 and ANN combinations achieved the highest accuracy of 0.951. For precision, the VGG19 and ANN combination again surpassed other CNN and LSTM, ANN based architectures by achieving a score of 0.95. The best sensitivity score of 0.903 was achieved by the VGG16+LSTM. The highest F1 score of 0.931 was achieved by VGG19+LSTM. Only the VGG16+LSTM achieved the best area under the curve (AUC) of 0.937, with VGG16+LSTM closely following with a 0.929 AUC score.
comment: The paper was accepted and presented in 3rd Workshop on Cancer Prevention, detection, and intervenTion (CaPTion @ MICCAI 2024)
☆ Hi-Mamba: Hierarchical Mamba for Efficient Image Super-Resolution
State Space Models (SSM), such as Mamba, have shown strong representation ability in modeling long-range dependency with linear complexity, achieving successful applications from high-level to low-level vision tasks. However, SSM's sequential nature necessitates multiple scans in different directions to compensate for the loss of spatial dependency when unfolding the image into a 1D sequence. This multi-direction scanning strategy significantly increases the computation overhead and is unbearable for high-resolution image processing. To address this problem, we propose a novel Hierarchical Mamba network, namely, Hi-Mamba, for image super-resolution (SR). Hi-Mamba consists of two key designs: (1) The Hierarchical Mamba Block (HMB) assembled by a Local SSM (L-SSM) and a Region SSM (R-SSM) both with the single-direction scanning, aggregates multi-scale representations to enhance the context modeling ability. (2) The Direction Alternation Hierarchical Mamba Group (DA-HMG) allocates the isomeric single-direction scanning into cascading HMBs to enrich the spatial relationship modeling. Extensive experiments demonstrate the superiority of Hi-Mamba across five benchmark datasets for efficient SR. For example, Hi-Mamba achieves a significant PSNR improvement of 0.29 dB on Manga109 for $\times3$ SR, compared to the strong lightweight MambaIR.
☆ MMIE: Massive Multimodal Interleaved Comprehension Benchmark for Large Vision-Language Models
Interleaved multimodal comprehension and generation, enabling models to produce and interpret both images and text in arbitrary sequences, have become a pivotal area in multimodal learning. Despite significant advancements, the evaluation of this capability remains insufficient. Existing benchmarks suffer from limitations in data scale, scope, and evaluation depth, while current evaluation metrics are often costly or biased, lacking in reliability for practical applications. To address these challenges, we introduce MMIE, a large-scale knowledge-intensive benchmark for evaluating interleaved multimodal comprehension and generation in Large Vision-Language Models (LVLMs). MMIE comprises 20K meticulously curated multimodal queries, spanning 3 categories, 12 fields, and 102 subfields, including mathematics, coding, physics, literature, health, and arts. It supports both interleaved inputs and outputs, offering a mix of multiple-choice and open-ended question formats to evaluate diverse competencies. Moreover, we propose a reliable automated evaluation metric, leveraging a scoring model fine-tuned with human-annotated data and systematic evaluation criteria, aimed at reducing bias and improving evaluation accuracy. Extensive experiments demonstrate the effectiveness of our benchmark and metrics in providing a comprehensive evaluation of interleaved LVLMs. Specifically, we evaluate eight LVLMs, revealing that even the best models show significant room for improvement, with most achieving only moderate results. We believe MMIE will drive further advancements in the development of interleaved LVLMs. We publicly release our benchmark and code in https://mmie-bench.github.io/.
☆ TextCtrl: Diffusion-based Scene Text Editing with Prior Guidance Control
Centred on content modification and style preservation, Scene Text Editing (STE) remains a challenging task despite considerable progress in text-to-image synthesis and text-driven image manipulation recently. GAN-based STE methods generally encounter a common issue of model generalization, while Diffusion-based STE methods suffer from undesired style deviations. To address these problems, we propose TextCtrl, a diffusion-based method that edits text with prior guidance control. Our method consists of two key components: (i) By constructing fine-grained text style disentanglement and robust text glyph structure representation, TextCtrl explicitly incorporates Style-Structure guidance into model design and network training, significantly improving text style consistency and rendering accuracy. (ii) To further leverage the style prior, a Glyph-adaptive Mutual Self-attention mechanism is proposed which deconstructs the implicit fine-grained features of the source image to enhance style consistency and vision quality during inference. Furthermore, to fill the vacancy of the real-world STE evaluation benchmark, we create the first real-world image-pair dataset termed ScenePair for fair comparisons. Experiments demonstrate the effectiveness of TextCtrl compared with previous methods concerning both style fidelity and text accuracy.
☆ MuseTalk: Real-Time High Quality Lip Synchronization with Latent Space Inpainting
Achieving high-resolution, identity consistency, and accurate lip-speech synchronization in face visual dubbing presents significant challenges, particularly for real-time applications like live video streaming. We propose MuseTalk, which generates lip-sync targets in a latent space encoded by a Variational Autoencoder, enabling high-fidelity talking face video generation with efficient inference. Specifically, we project the occluded lower half of the face image and itself as an reference into a low-dimensional latent space and use a multi-scale U-Net to fuse audio and visual features at various levels. We further propose a novel sampling strategy during training, which selects reference images with head poses closely matching the target, allowing the model to focus on precise lip movement by filtering out redundant information. Additionally, we analyze the mechanism of lip-sync loss and reveal its relationship with input information volume. Extensive experiments show that MuseTalk consistently outperforms recent state-of-the-art methods in visual fidelity and achieves comparable lip-sync accuracy. As MuseTalk supports the online generation of face at 256x256 at more than 30 FPS with negligible starting latency, it paves the way for real-time applications.
comment: 15 pages, 4 figures
☆ Interaction-Guided Two-Branch Image Dehazing Network ACCV 2024
Image dehazing aims to restore clean images from hazy ones. Convolutional Neural Networks (CNNs) and Transformers have demonstrated exceptional performance in local and global feature extraction, respectively, and currently represent the two mainstream frameworks in image dehazing. In this paper, we propose a novel dual-branch image dehazing framework that guides CNN and Transformer components interactively. We reconsider the complementary characteristics of CNNs and Transformers by leveraging the differential relationships between global and local features for interactive guidance. This approach enables the capture of local feature positions through global attention maps, allowing the CNN to focus solely on feature information at effective positions. The single-branch Transformer design ensures the network's global information recovery capability. Extensive experiments demonstrate that our proposed method yields competitive qualitative and quantitative evaluation performance on both synthetic and real public datasets. Codes are available at https://github.com/Feecuin/Two-Branch-Dehazing
comment: Accepted by ACCV 2024
☆ StegaINR4MIH: steganography by implicit neural representation for multi-image hiding
Multi-image hiding, which embeds multiple secret images into a cover image and is able to recover these images with high quality, has gradually become a research hotspot in the field of image steganography. However, due to the need to embed a large amount of data in a limited cover image space, issues such as contour shadowing or color distortion often arise, posing significant challenges for multi-image hiding. In this paper, we propose StegaINR4MIH, a novel implicit neural representation steganography framework that enables the hiding of multiple images within a single implicit representation function. In contrast to traditional methods that use multiple encoders to achieve multi-image embedding, our approach leverages the redundancy of implicit representation function parameters and employs magnitude-based weight selection and secret weight substitution on pre-trained cover image functions to effectively hide and independently extract multiple secret images. We conduct experiments on images with a resolution of from three different datasets: CelebA-HQ, COCO, and DIV2K. When hiding two secret images, the PSNR values of both the secret images and the stego images exceed 42. When hiding five secret images, the PSNR values of both the secret images and the stego images exceed 39. Extensive experiments demonstrate the superior performance of the proposed method in terms of visual quality and undetectability.
comment: 46pages,14figures
☆ Mixture of Experts Made Personalized: Federated Prompt Learning for Vision-Language Models
Prompt learning for pre-trained Vision-Language Models (VLMs) like CLIP has demonstrated potent applicability across diverse downstream tasks. This lightweight approach has quickly gained traction from federated learning (FL) researchers who seek to efficiently adapt VLMs to heterogeneous scenarios. However, current federated prompt learning methods are habitually restricted to the traditional FL paradigm, where the participating clients are generally only allowed to download a single globally aggregated model from the server. While justifiable for training full-sized models under federated settings, in this work, we argue that this paradigm is ill-suited for lightweight prompts. By facilitating the clients to download multiple pre-aggregated prompts as fixed non-local experts, we propose Personalized Federated Mixture of Adaptive Prompts (pFedMoAP), a novel FL framework that personalizes the prompt learning process through the lens of Mixture of Experts (MoE). pFedMoAP implements a local attention-based gating network that learns to generate enhanced text features for better alignment with local image data on the client, benefiting from both local and downloaded non-local adaptive prompt experts. The non-local experts are sparsely selected from a server-maintained pool, fostering collaborative learning across clients. To evaluate the proposed algorithm, we conduct extensive experiments across 9 datasets under various heterogeneous federated settings. The results show that pFedMoAP consistently outperforms the state-of-the-art alternatives, underscoring its efficacy in personalizing prompt learning for CLIP within the federated learning paradigm.
comment: 16 pages, 4 figures
☆ Can We Predict Performance of Large Models across Vision-Language Tasks?
Evaluating large vision-language models (LVLMs) is very expensive, due to the high computational costs and the wide variety of tasks. The good news is that if we already have some observed performance scores, we may be able to infer unknown ones. In this study, we propose a new framework for predicting unknown performance scores based on observed ones from other LVLMs or tasks. We first formulate the performance prediction as a matrix completion task. Specifically, we construct a sparse performance matrix $\boldsymbol{R}$, where each entry $R_{mn}$ represents the performance score of the $m$-th model on the $n$-th dataset. By applying probabilistic matrix factorization (PMF) with Markov chain Monte Carlo (MCMC), we can complete the performance matrix, that is, predict unknown scores. Additionally, we estimate the uncertainty of performance prediction based on MCMC. Practitioners can evaluate their models on untested tasks with higher uncertainty first, quickly reducing errors in performance prediction. We further introduce several improvements to enhance PMF for scenarios with sparse observed performance scores. In experiments, we systematically evaluate 108 LVLMs on 176 datasets from 36 benchmarks, constructing training and testing sets for validating our framework. Our experiments demonstrate the accuracy of PMF in predicting unknown scores, the reliability of uncertainty estimates in ordering evaluations, and the effectiveness of our enhancements for handling sparse data.
comment: Under Review. Project page: https://github.com/Qinyu-Allen-Zhao/CrossPred-LVLM
☆ High-Precision Dichotomous Image Segmentation via Probing Diffusion Capacity
In the realm of high-resolution (HR), fine-grained image segmentation, the primary challenge is balancing broad contextual awareness with the precision required for detailed object delineation, capturing intricate details and the finest edges of objects. Diffusion models, trained on vast datasets comprising billions of image-text pairs, such as SD V2.1, have revolutionized text-to-image synthesis by delivering exceptional quality, fine detail resolution, and strong contextual awareness, making them an attractive solution for high-resolution image segmentation. To this end, we propose DiffDIS, a diffusion-driven segmentation model that taps into the potential of the pre-trained U-Net within diffusion models, specifically designed for high-resolution, fine-grained object segmentation. By leveraging the robust generalization capabilities and rich, versatile image representation prior of the SD models, coupled with a task-specific stable one-step denoising approach, we significantly reduce the inference time while preserving high-fidelity, detailed generation. Additionally, we introduce an auxiliary edge generation task to not only enhance the preservation of fine details of the object boundaries, but reconcile the probabilistic nature of diffusion with the deterministic demands of segmentation. With these refined strategies in place, DiffDIS serves as a rapid object mask generation model, specifically optimized for generating detailed binary maps at high resolutions, while demonstrating impressive accuracy and swift processing. Experiments on the DIS5K dataset demonstrate the superiority of DiffDIS, achieving state-of-the-art results through a streamlined inference process. Our code will be made publicly available.
comment: 13 pages
☆ REHRSeg: Unleashing the Power of Self-Supervised Super-Resolution for Resource-Efficient 3D MRI Segmentation
High-resolution (HR) 3D magnetic resonance imaging (MRI) can provide detailed anatomical structural information, enabling precise segmentation of regions of interest for various medical image analysis tasks. Due to the high demands of acquisition device, collection of HR images with their annotations is always impractical in clinical scenarios. Consequently, segmentation results based on low-resolution (LR) images with large slice thickness are often unsatisfactory for subsequent tasks. In this paper, we propose a novel Resource-Efficient High-Resolution Segmentation framework (REHRSeg) to address the above-mentioned challenges in real-world applications, which can achieve HR segmentation while only employing the LR images as input. REHRSeg is designed to leverage self-supervised super-resolution (self-SR) to provide pseudo supervision, therefore the relatively easier-to-acquire LR annotated images generated by 2D scanning protocols can be directly used for model training. The main contribution to ensure the effectiveness in self-SR for enhancing segmentation is three-fold: (1) We mitigate the data scarcity problem in the medical field by using pseudo-data for training the segmentation model. (2) We design an uncertainty-aware super-resolution (UASR) head in self-SR to raise the awareness of segmentation uncertainty as commonly appeared on the ROI boundaries. (3) We align the spatial features for self-SR and segmentation through structural knowledge distillation to enable a better capture of region correlations. Experimental results demonstrate that REHRSeg achieves high-quality HR segmentation without intensive supervision, while also significantly improving the baseline performance for LR segmentation.
☆ Innovative Deep Learning Techniques for Obstacle Recognition: A Comparative Study of Modern Detection Algorithms
This study explores a comprehensive approach to obstacle detection using advanced YOLO models, specifically YOLOv8, YOLOv7, YOLOv6, and YOLOv5. Leveraging deep learning techniques, the research focuses on the performance comparison of these models in real-time detection scenarios. The findings demonstrate that YOLOv8 achieves the highest accuracy with improved precision-recall metrics. Detailed training processes, algorithmic principles, and a range of experimental results are presented to validate the model's effectiveness.
☆ Out-of-Bounding-Box Triggers: A Stealthy Approach to Cheat Object Detectors ECCV 2024
In recent years, the study of adversarial robustness in object detection systems, particularly those based on deep neural networks (DNNs), has become a pivotal area of research. Traditional physical attacks targeting object detectors, such as adversarial patches and texture manipulations, directly manipulate the surface of the object. While these methods are effective, their overt manipulation of objects may draw attention in real-world applications. To address this, this paper introduces a more subtle approach: an inconspicuous adversarial trigger that operates outside the bounding boxes, rendering the object undetectable to the model. We further enhance this approach by proposing the Feature Guidance (FG) technique and the Universal Auto-PGD (UAPGD) optimization strategy for crafting high-quality triggers. The effectiveness of our method is validated through extensive empirical testing, demonstrating its high performance in both digital and physical environments. The code and video will be available at: https://github.com/linToTao/Out-of-bbox-attack.
comment: ECCV 2024
☆ The Ingredients for Robotic Diffusion Transformers
In recent years roboticists have achieved remarkable progress in solving increasingly general tasks on dexterous robotic hardware by leveraging high capacity Transformer network architectures and generative diffusion models. Unfortunately, combining these two orthogonal improvements has proven surprisingly difficult, since there is no clear and well-understood process for making important design choices. In this paper, we identify, study and improve key architectural design decisions for high-capacity diffusion transformer policies. The resulting models can efficiently solve diverse tasks on multiple robot embodiments, without the excruciating pain of per-setup hyper-parameter tuning. By combining the results of our investigation with our improved model components, we are able to present a novel architecture, named \method, that significantly outperforms the state of the art in solving long-horizon ($1500+$ time-steps) dexterous tasks on a bi-manual ALOHA robot. In addition, we find that our policies show improved scaling performance when trained on 10 hours of highly multi-modal, language annotated ALOHA demonstration data. We hope this work will open the door for future robot learning techniques that leverage the efficiency of generative diffusion modeling with the scalability of large scale transformer architectures. Code, robot dataset, and videos are available at: https://dit-policy.github.io
☆ PointNet with KAN versus PointNet with MLP for 3D Classification and Segmentation of Point Sets
We introduce PointNet-KAN, a neural network for 3D point cloud classification and segmentation tasks, built upon two key components. First, it employs Kolmogorov-Arnold Networks (KANs) instead of traditional Multilayer Perceptrons (MLPs). Second, it retains the core principle of PointNet by using shared KAN layers and applying symmetric functions for global feature extraction, ensuring permutation invariance with respect to the input features. In traditional MLPs, the goal is to train the weights and biases with fixed activation functions; however, in KANs, the goal is to train the activation functions themselves. We use Jacobi polynomials to construct the KAN layers. We extensively evaluate PointNet-KAN across various polynomial degrees and special types such as the Lagrange, Chebyshev, and Gegenbauer polynomials. Our results show that PointNet-KAN achieves competitive performance compared to PointNet with MLPs on benchmark datasets for 3D object classification and segmentation, despite employing a shallower and simpler network architecture. We hope this work serves as a foundation and provides guidance for integrating KANs, as an alternative to MLPs, into more advanced point cloud processing architectures.
☆ Learning to Customize Text-to-Image Diffusion In Diverse Context
Most text-to-image customization techniques fine-tune models on a small set of \emph{personal concept} images captured in minimal contexts. This often results in the model becoming overfitted to these training images and unable to generalize to new contexts in future text prompts. Existing customization methods are built on the success of effectively representing personal concepts as textual embeddings. Thus, in this work, we resort to diversifying the context of these personal concepts \emph{solely} within the textual space by simply creating a contextually rich set of text prompts, together with a widely used self-supervised learning objective. Surprisingly, this straightforward and cost-effective method significantly improves semantic alignment in the textual space, and this effect further extends to the image space, resulting in higher prompt fidelity for generated images. Additionally, our approach does not require any architectural modifications, making it highly compatible with existing text-to-image customization methods. We demonstrate the broad applicability of our approach by combining it with four different baseline methods, achieving notable CLIP score improvements.
☆ DINTR: Tracking via Diffusion-based Interpolation NeurIPS 2024
Object tracking is a fundamental task in computer vision, requiring the localization of objects of interest across video frames. Diffusion models have shown remarkable capabilities in visual generation, making them well-suited for addressing several requirements of the tracking problem. This work proposes a novel diffusion-based methodology to formulate the tracking task. Firstly, their conditional process allows for injecting indications of the target object into the generation process. Secondly, diffusion mechanics can be developed to inherently model temporal correspondences, enabling the reconstruction of actual frames in video. However, existing diffusion models rely on extensive and unnecessary mapping to a Gaussian noise domain, which can be replaced by a more efficient and stable interpolation process. Our proposed interpolation mechanism draws inspiration from classic image-processing techniques, offering a more interpretable, stable, and faster approach tailored specifically for the object tracking task. By leveraging the strengths of diffusion models while circumventing their limitations, our Diffusion-based INterpolation TrackeR (DINTR) presents a promising new paradigm and achieves a superior multiplicity on seven benchmarks across five indicator representations.
comment: Accepted at NeurIPS 2024
☆ UAV3D: A Large-scale 3D Perception Benchmark for Unmanned Aerial Vehicles NeurIPS 2024
Unmanned Aerial Vehicles (UAVs), equipped with cameras, are employed in numerous applications, including aerial photography, surveillance, and agriculture. In these applications, robust object detection and tracking are essential for the effective deployment of UAVs. However, existing benchmarks for UAV applications are mainly designed for traditional 2D perception tasks, restricting the development of real-world applications that require a 3D understanding of the environment. Furthermore, despite recent advancements in single-UAV perception, limited views of a single UAV platform significantly constrain its perception capabilities over long distances or in occluded areas. To address these challenges, we introduce UAV3D, a benchmark designed to advance research in both 3D and collaborative 3D perception tasks with UAVs. UAV3D comprises 1,000 scenes, each of which has 20 frames with fully annotated 3D bounding boxes on vehicles. We provide the benchmark for four 3D perception tasks: single-UAV 3D object detection, single-UAV object tracking, collaborative-UAV 3D object detection, and collaborative-UAV object tracking. Our dataset and code are available at https://huiyegit.github.io/UAV3D_Benchmark/.
comment: Accepted at NeurIPS 2024
☆ Real-Time Localization and Bimodal Point Pattern Analysis of Palms Using UAV Imagery
Understanding the spatial distribution of palms within tropical forests is essential for effective ecological monitoring, conservation strategies, and the sustainable integration of natural forest products into local and global supply chains. However, the analysis of remotely sensed data in these environments faces significant challenges, such as overlapping palm and tree crowns, uneven shading across the canopy surface, and the heterogeneous nature of the forest landscapes, which often affect the performance of palm detection and segmentation algorithms. To overcome these issues, we introduce PalmDSNet, a deep learning framework for real-time detection, segmentation, and counting of canopy palms. Additionally, we employ a bimodal reproduction algorithm that simulates palm spatial propagation to further enhance the understanding of these point patterns using PalmDSNet's results. We used UAV-captured imagery to create orthomosaics from 21 sites across western Ecuadorian tropical forests, covering a gradient from the everwet Choc\'o forests near Colombia to the drier forests of southwestern Ecuador. Expert annotations were used to create a comprehensive dataset, including 7,356 bounding boxes on image patches and 7,603 palm centers across five orthomosaics, encompassing a total area of 449 hectares. By combining PalmDSNet with the bimodal reproduction algorithm, which optimizes parameters for both local and global spatial variability, we effectively simulate the spatial distribution of palms in diverse and dense tropical environments, validating its utility for advanced applications in tropical forest monitoring and remote sensing analysis.
comment: 25 pages, 8 figures, 5 tables
☆ MoonMetaSync: Lunar Image Registration Analysis
This paper compares scale-invariant (SIFT) and scale-variant (ORB) feature detection methods, alongside our novel feature detector, IntFeat, specifically applied to lunar imagery. We evaluate these methods using low (128x128) and high-resolution (1024x1024) lunar image patches, providing insights into their performance across scales in challenging extraterrestrial environments. IntFeat combines high-level features from SIFT and low-level features from ORB into a single vector space for robust lunar image registration. We introduce SyncVision, a Python package that compares lunar images using various registration methods, including SIFT, ORB, and IntFeat. Our analysis includes upscaling low-resolution lunar images using bi-linear and bi-cubic interpolation, offering a unique perspective on registration effectiveness across scales and feature detectors in lunar landscapes. This research contributes to computer vision and planetary science by comparing feature detection methods for lunar imagery and introducing a versatile tool for lunar image registration and evaluation, with implications for multi-resolution image analysis in space exploration applications.
☆ Classifying Healthy and Defective Fruits with a Multi-Input Architecture and CNN Models
This study presents an investigation into the utilization of a Multi-Input architecture for the classification of fruits (apples and mangoes) into healthy and defective states, employing both RGB and silhouette images. The primary aim is to enhance the accuracy of CNN models. The methodology encompasses image acquisition, preprocessing of datasets, training, and evaluation of two CNN models: MobileNetV2 and VGG16. Results reveal that the inclusion of silhouette images alongside the Multi-Input architecture yields models with superior performance compared to using only RGB images for fruit classification, whether healthy or defective. Specifically, optimal results were achieved using the MobileNetV2 model, achieving 100\% accuracy. This finding suggests the efficacy of this combined methodology in improving the precise classification of healthy or defective fruits, which could have significant implications for applications related to external quality inspection of fruits.
comment: 7 pages, 11 figures
☆ EchoApex: A General-Purpose Vision Foundation Model for Echocardiography
Quantitative evaluation of echocardiography is essential for precise assessment of cardiac condition, monitoring disease progression, and guiding treatment decisions. The diverse nature of echo images, including variations in probe types, manufacturers, and pathologies, poses challenges for developing artificial intelligent models that can generalize across different clinical practice. We introduce EchoApex, the first general-purpose vision foundation model echocardiography with applications on a variety of clinical practice. Leveraging self-supervised learning, EchoApex is pretrained on over 20 million echo images from 11 clinical centres. By incorporating task-specific decoders and adapter modules, we demonstrate the effectiveness of EchoApex on 4 different kind of clinical applications with 28 sub-tasks, including view classification, interactive structure segmentation, left ventricle hypertrophy detection and automated ejection fraction estimation from view sequences. Compared to state-of-the-art task-specific models, EchoApex attains improved performance with a unified image encoding architecture, demonstrating the benefits of model pretraining at scale with in-domain data. Furthermore, EchoApex illustrates the potential for developing a general-purpose vision foundation model tailored specifically for echocardiography, capable of addressing a diverse range of clinical applications with high efficiency and efficacy.
☆ Locality Alignment Improves Vision-Language Models
Vision language models (VLMs) have seen growing adoption in recent years, but many still struggle with basic spatial reasoning errors. We hypothesize that this is due to VLMs adopting pre-trained vision backbones, specifically vision transformers (ViTs) trained with image-level supervision and minimal inductive biases. Such models may fail to encode the class contents at each position in the image, and our goal is to resolve this by ensuring that the vision backbone effectively captures both local and global image semantics. Our main insight is that we do not require new supervision to learn this capability -- pre-trained models contain significant knowledge of local semantics that we can extract and use for scalable self-supervision. We propose a new efficient post-training stage for ViTs called locality alignment and a novel fine-tuning procedure called MaskEmbed that uses a masked reconstruction loss to learn semantic contributions for each image patch. We first evaluate locality alignment with a vision-only benchmark, finding that it improves a model's performance at a patch-level semantic segmentation task, especially for strong backbones trained with image-caption pairs (e.g., CLIP and SigLIP). We then train a series of VLMs with and without locality alignment, and show that locality-aligned backbones improve performance across a range of benchmarks, particularly ones that involve spatial understanding (e.g., RefCOCO, OCID-Ref, TallyQA, VSR, AI2D). Overall, we demonstrate that we can efficiently learn local semantic extraction via a locality alignment stage, and that this procedure complements existing VLM training recipes that use off-the-shelf vision backbones.
☆ Few-shot Novel View Synthesis using Depth Aware 3D Gaussian Splatting ECCV 2024
3D Gaussian splatting has surpassed neural radiance field methods in novel view synthesis by achieving lower computational costs and real-time high-quality rendering. Although it produces a high-quality rendering with a lot of input views, its performance drops significantly when only a few views are available. In this work, we address this by proposing a depth-aware Gaussian splatting method for few-shot novel view synthesis. We use monocular depth prediction as a prior, along with a scale-invariant depth loss, to constrain the 3D shape under just a few input views. We also model color using lower-order spherical harmonics to avoid overfitting. Further, we observe that removing splats with lower opacity periodically, as performed in the original work, leads to a very sparse point cloud and, hence, a lower-quality rendering. To mitigate this, we retain all the splats, leading to a better reconstruction in a few view settings. Experimental results show that our method outperforms the traditional 3D Gaussian splatting methods by achieving improvements of 10.5% in peak signal-to-noise ratio, 6% in structural similarity index, and 14.1% in perceptual similarity, thereby validating the effectiveness of our approach. The code will be made available at: https://github.com/raja-kumar/depth-aware-3DGS
comment: Presented in ECCV 2024 workshop S3DSGR
☆ Character-aware audio-visual subtitling in context ACCV 2024
This paper presents an improved framework for character-aware audio-visual subtitling in TV shows. Our approach integrates speech recognition, speaker diarisation, and character recognition, utilising both audio and visual cues. This holistic solution addresses what is said, when it's said, and who is speaking, providing a more comprehensive and accurate character-aware subtitling for TV shows. Our approach brings improvements on two fronts: first, we show that audio-visual synchronisation can be used to pick out the talking face amongst others present in a video clip, and assign an identity to the corresponding speech segment. This audio-visual approach improves recognition accuracy and yield over current methods. Second, we show that the speaker of short segments can be determined by using the temporal context of the dialogue within a scene. We propose an approach using local voice embeddings of the audio, and large language model reasoning on the text transcription. This overcomes a limitation of existing methods that they are unable to accurately assign speakers to short temporal segments. We validate the method on a dataset with 12 TV shows, demonstrating superior performance in speaker diarisation and character recognition accuracy compared to existing approaches. Project page : https://www.robots.ox.ac.uk/~vgg/research/llr-context/
comment: ACCV 2024
☆ CleanUMamba: A Compact Mamba Network for Speech Denoising using Channel Pruning
This paper presents CleanUMamba, a time-domain neural network architecture designed for real-time causal audio denoising directly applied to raw waveforms. CleanUMamba leverages a U-Net encoder-decoder structure, incorporating the Mamba state-space model in the bottleneck layer. By replacing conventional self-attention and LSTM mechanisms with Mamba, our architecture offers superior denoising performance while maintaining a constant memory footprint, enabling streaming operation. To enhance efficiency, we applied structured channel pruning, achieving an 8X reduction in model size without compromising audio quality. Our model demonstrates strong results in the Interspeech 2020 Deep Noise Suppression challenge. Specifically, CleanUMamba achieves a PESQ score of 2.42 and STOI of 95.1% with only 442K parameters and 468M MACs, matching or outperforming larger models in real-time performance. Code will be available at: https://github.com/lab-emi/CleanUMamba
comment: 5 pages, 4 figures
☆ Beyond Fixed Topologies: Unregistered Training and Comprehensive Evaluation Metrics for 3D Talking Heads
Generating speech-driven 3D talking heads presents numerous challenges; among those is dealing with varying mesh topologies. Existing methods require a registered setting, where all meshes share a common topology: a point-wise correspondence across all meshes the model can animate. While simplifying the problem, it limits applicability as unseen meshes must adhere to the training topology. This work presents a framework capable of animating 3D faces in arbitrary topologies, including real scanned data. Our approach relies on a model leveraging heat diffusion over meshes to overcome the fixed topology constraint. We explore two training settings: a supervised one, in which training sequences share a fixed topology within a sequence but any mesh can be animated at test time, and an unsupervised one, which allows effective training with varying mesh structures. Additionally, we highlight the limitations of current evaluation metrics and propose new metrics for better lip-syncing evaluation between speech and facial movements. Our extensive evaluation shows our approach performs favorably compared to fixed topology techniques, setting a new benchmark by offering a versatile and high-fidelity solution for 3D talking head generation.
☆ ET-Former: Efficient Triplane Deformable Attention for 3D Semantic Scene Completion From Monocular Camera
We introduce ET-Former, a novel end-to-end algorithm for semantic scene completion using a single monocular camera. Our approach generates a semantic occupancy map from single RGB observation while simultaneously providing uncertainty estimates for semantic predictions. By designing a triplane-based deformable attention mechanism, our approach improves geometric understanding of the scene than other SOTA approaches and reduces noise in semantic predictions. Additionally, through the use of a Conditional Variational AutoEncoder (CVAE), we estimate the uncertainties of these predictions. The generated semantic and uncertainty maps will aid in the formulation of navigation strategies that facilitate safe and permissible decision-making in the future. Evaluated on the Semantic-KITTI dataset, ET-Former achieves the highest IoU and mIoU, surpassing other methods by 15.16% in IoU and 24.24% in mIoU, while reducing GPU memory usage of existing methods by 25%-50.5%.
♻ ☆ Revisiting Few-Shot Object Detection with Vision-Language Models NeurIPS
The era of vision-language models (VLMs) trained on web-scale datasets challenges conventional formulations of "open-world" perception. In this work, we revisit the task of few-shot object detection (FSOD) in the context of recent foundational VLMs. First, we point out that zero-shot predictions from VLMs such as GroundingDINO significantly outperform state-of-the-art few-shot detectors (48 vs. 33 AP) on COCO. Despite their strong zero-shot performance, such foundation models may still be sub-optimal. For example, trucks on the web may be defined differently from trucks for a target application such as autonomous vehicle perception. We argue that the task of few-shot recognition can be reformulated as aligning foundation models to target concepts using a few examples. Interestingly, such examples can be multi-modal, using both text and visual cues, mimicking instructions that are often given to human annotators when defining a target concept of interest. Concretely, we propose Foundational FSOD, a new benchmark protocol that evaluates detectors pre-trained on any external data and fine-tuned on multi-modal (text and visual) K-shot examples per target class. We repurpose nuImages for Foundational FSOD, benchmark several popular open-source VLMs, and provide an empirical analysis of state-of-the-art methods. Lastly, we discuss our recent CVPR 2024 Foundational FSOD competition and share insights from the community. Notably, the winning team significantly outperforms our baseline by 23.3 mAP! Our code and dataset splits are available at https://github.com/anishmadan23/foundational_fsod
comment: The first two authors contributed equally. This work has been accepted to the Neural Information Processing Systems (NeurIPS) 2024 Datasets & Benchmark Track
♻ ☆ Enhancing Performance of Point Cloud Completion Networks with Consistency Loss
Point cloud completion networks are conventionally trained to minimize the disparities between the completed point cloud and the ground-truth counterpart. However, an incomplete object-level point cloud can have multiple valid completion solutions when it is examined in isolation. This one-to-many mapping issue can cause contradictory supervision signals to the network because the loss function may produce different values for identical input-output pairs of the network. In many cases, this issue could adversely affect the network optimization process. In this work, we propose to enhance the conventional learning objective using a novel completion consistency loss to mitigate the one-to-many mapping problem. Specifically, the proposed consistency loss ensure that a point cloud completion network generates a coherent completion solution for incomplete objects originating from the same source point cloud. Experimental results across multiple well-established datasets and benchmarks demonstrated the proposed completion consistency loss have excellent capability to enhance the completion performance of various existing networks without any modification to the design of the networks. The proposed consistency loss enhances the performance of the point completion network without affecting the inference speed, thereby increasing the accuracy of point cloud completion. Notably, a state-of-the-art point completion network trained with the proposed consistency loss can achieve state-of-the-art accuracy on the challenging new MVP dataset. The code and result of experiment various point completion models using proposed consistency loss will be available at: https://github.com/kaist-avelab/ConsistencyLoss .
comment: First version of Paper "Enhancing Performance of Point Cloud Completion Networks with Consistency Loss" by Kevin Tirta Wijaya and Christofel Rio Goenawan. In process submission to Neurocomputing Journal 2024
♻ ☆ Sign Stitching: A Novel Approach to Sign Language Production BMVC 2024
Sign Language Production (SLP) is a challenging task, given the limited resources available and the inherent diversity within sign data. As a result, previous works have suffered from the problem of regression to the mean, leading to under-articulated and incomprehensible signing. In this paper, we propose using dictionary examples to create expressive sign language sequences. However, simply concatenating the signs would create robotic and unnatural sequences. Therefore, we present a 7-step approach to effectively stitch the signs together. First, by normalising each sign into a canonical pose, cropping and stitching we create a continuous sequence. Then by applying filtering in the frequency domain and resampling each sign we create cohesive natural sequences, that mimic the prosody found in the original data. We leverage the SignGAN model to map the output to a photo-realistic signer and present a complete Text-to-Sign (T2S) SLP pipeline. Our evaluation demonstrates the effectiveness of this approach, showcasing state-of-the-art performance across all datasets.
comment: 18 pages, 3 figures, 4 tables. Accepted to BMVC 2024
♻ ☆ Detecting Audio-Visual Deepfakes with Fine-Grained Inconsistencies BMVC 2024
Existing methods on audio-visual deepfake detection mainly focus on high-level features for modeling inconsistencies between audio and visual data. As a result, these approaches usually overlook finer audio-visual artifacts, which are inherent to deepfakes. Herein, we propose the introduction of fine-grained mechanisms for detecting subtle artifacts in both spatial and temporal domains. First, we introduce a local audio-visual model capable of capturing small spatial regions that are prone to inconsistencies with audio. For that purpose, a fine-grained mechanism based on a spatially-local distance coupled with an attention module is adopted. Second, we introduce a temporally-local pseudo-fake augmentation to include samples incorporating subtle temporal inconsistencies in our training set. Experiments on the DFDC and the FakeAVCeleb datasets demonstrate the superiority of the proposed method in terms of generalization as compared to the state-of-the-art under both in-dataset and cross-dataset settings.
comment: Accepted in BMVC 2024
♻ ☆ Extracting Training Data from Unconditional Diffusion Models
As diffusion probabilistic models (DPMs) are being employed as mainstream models for generative artificial intelligence (AI), the study of their memorization of the raw training data has attracted growing attention. Existing works in this direction aim to establish an understanding of whether or to what extent DPMs learn by memorization. Such an understanding is crucial for identifying potential risks of data leakage and copyright infringement in diffusion models and, more importantly, for more controllable generation and trustworthy application of Artificial Intelligence Generated Content (AIGC). While previous works have made important observations of when DPMs are prone to memorization, these findings are mostly empirical, and the developed data extraction methods only work for conditional diffusion models. In this work, we aim to establish a theoretical understanding of memorization in DPMs with 1) a memorization metric for theoretical analysis, 2) an analysis of conditional memorization with informative and random labels, and 3) two better evaluation metrics for measuring memorization. Based on the theoretical analysis, we further propose a novel data extraction method called \textbf{Surrogate condItional Data Extraction (SIDE)} that leverages a classifier trained on generated data as a surrogate condition to extract training data directly from unconditional diffusion models. Our empirical results demonstrate that SIDE can extract training data from diffusion models where previous methods fail, and it is on average over 50\% more effective across different scales of the CelebA dataset.
comment: This is an old version of arXiv:2410.02467. Please refer to the new version
♻ ☆ MCTrack: A Unified 3D Multi-Object Tracking Framework for Autonomous Driving
This paper introduces MCTrack, a new 3D multi-object tracking method that achieves state-of-the-art (SOTA) performance across KITTI, nuScenes, and Waymo datasets. Addressing the gap in existing tracking paradigms, which often perform well on specific datasets but lack generalizability, MCTrack offers a unified solution. Additionally, we have standardized the format of perceptual results across various datasets, termed BaseVersion, facilitating researchers in the field of multi-object tracking (MOT) to concentrate on the core algorithmic development without the undue burden of data preprocessing. Finally, recognizing the limitations of current evaluation metrics, we propose a novel set that assesses motion information output, such as velocity and acceleration, crucial for downstream tasks. The source codes of the proposed method are available at this link: https://github.com/megvii-research/MCTrack}{https://github.com/megvii-research/MCTrack
comment: 14 pages, 7 figures
♻ ☆ GAIA: Rethinking Action Quality Assessment for AI-Generated Videos NeurIPS2024
Assessing action quality is both imperative and challenging due to its significant impact on the quality of AI-generated videos, further complicated by the inherently ambiguous nature of actions within AI-generated video (AIGV). Current action quality assessment (AQA) algorithms predominantly focus on actions from real specific scenarios and are pre-trained with normative action features, thus rendering them inapplicable in AIGVs. To address these problems, we construct GAIA, a Generic AI-generated Action dataset, by conducting a large-scale subjective evaluation from a novel causal reasoning-based perspective, resulting in 971,244 ratings among 9,180 video-action pairs. Based on GAIA, we evaluate a suite of popular text-to-video (T2V) models on their ability to generate visually rational actions, revealing their pros and cons on different categories of actions. We also extend GAIA as a testbed to benchmark the AQA capacity of existing automatic evaluation methods. Results show that traditional AQA methods, action-related metrics in recent T2V benchmarks, and mainstream video quality methods perform poorly with an average SRCC of 0.454, 0.191, and 0.519, respectively, indicating a sizable gap between current models and human action perception patterns in AIGVs. Our findings underscore the significance of action quality as a unique perspective for studying AIGVs and can catalyze progress towards methods with enhanced capacities for AQA in AIGVs.
comment: Accepted by NeurIPS2024 Dataset and Benchmark Track as Spotlight. 33 pages, 15 figures
♻ ☆ AR-TTA: A Simple Method for Real-World Continual Test-Time Adaptation BMVC2024
Test-time adaptation is a promising research direction that allows the source model to adapt itself to changes in data distribution without any supervision. Yet, current methods are usually evaluated on benchmarks that are only a simplification of real-world scenarios. Hence, we propose to validate test-time adaptation methods using the recently introduced datasets for autonomous driving, namely CLAD-C and SHIFT. We observe that current test-time adaptation methods struggle to effectively handle varying degrees of domain shift, often resulting in degraded performance that falls below that of the source model. We noticed that the root of the problem lies in the inability to preserve the knowledge of the source model and adapt to dynamically changing, temporally correlated data streams. Therefore, we enhance the well-established self-training framework by incorporating a small memory buffer to increase model stability and at the same time perform dynamic adaptation based on the intensity of domain shift. The proposed method, named AR-TTA, outperforms existing approaches on both synthetic and more real-world benchmarks and shows robustness across a variety of TTA scenarios. The code is available at https://github.com/dmn-sjk/AR-TTA.
comment: Accepted for BMVC2024
♻ ☆ TV-SAM: Increasing Zero-Shot Segmentation Performance on Multimodal Medical Images Using GPT-4 Generated Descriptive Prompts Without Human Annotation
This study presents a novel multimodal medical image zero-shot segmentation algorithm named the text-visual-prompt segment anything model (TV-SAM) without any manual annotations. The TV-SAM incorporates and integrates the large language model GPT-4, the vision language model GLIP, and the SAM to autonomously generate descriptive text prompts and visual bounding box prompts from medical images, thereby enhancing the SAM's capability for zero-shot segmentation. Comprehensive evaluations are implemented on seven public datasets encompassing eight imaging modalities to demonstrate that TV-SAM can effectively segment unseen targets across various modalities without additional training. TV-SAM significantly outperforms SAM AUTO and GSAM, closely matching the performance of SAM BBOX with gold standard bounding box prompts and surpasses the state-of-the-art methods on specific datasets such as ISIC and WBC. The study indicates that TV-SAM serves as an effective multimodal medical image zero-shot segmentation algorithm, highlighting the significant contribution of GPT-4 to zero-shot segmentation. By integrating foundational models such as GPT-4, GLIP, and SAM, the ability to address complex problems in specialized domains can be enhanced.
comment: 13 pages, 5 figures, 4 tables, accepted by BDMA Journal
♻ ☆ Learning Temporally Equivariance for Degenerative Disease Progression in OCT by Predicting Future Representations MICCAI 2024
Contrastive pretraining provides robust representations by ensuring their invariance to different image transformations while simultaneously preventing representational collapse. Equivariant contrastive learning, on the other hand, provides representations sensitive to specific image transformations while remaining invariant to others. By introducing equivariance to time-induced transformations, such as disease-related anatomical changes in longitudinal imaging, the model can effectively capture such changes in the representation space. In this work, we propose a Time-equivariant Contrastive Learning (TC) method. First, an encoder embeds two unlabeled scans from different time points of the same patient into the representation space. Next, a temporal equivariance module is trained to predict the representation of a later visit based on the representation from one of the previous visits and the corresponding time interval with a novel regularization loss term while preserving the invariance property to irrelevant image transformations. On a large longitudinal dataset, our model clearly outperforms existing equivariant contrastive methods in predicting progression from intermediate age-related macular degeneration (AMD) to advanced wet-AMD within a specified time-window.
comment: Accepted at MICCAI 2024 (early accept, top 11%)
♻ ☆ Geometry-Informed Neural Networks
Geometry is a ubiquitous tool in computer graphics, design, and engineering. However, the lack of large shape datasets limits the application of state-of-the-art supervised learning methods and motivates the exploration of alternative learning strategies. To this end, we introduce geometry-informed neural networks (GINNs) -- a framework for training shape-generative neural fields without data by leveraging user-specified design requirements in the form of objectives and constraints. By adding diversity as an explicit constraint, GINNs avoid mode-collapse and can generate multiple diverse solutions, often required in geometry tasks. Experimentally, we apply GINNs to several validation problems and a realistic 3D engineering design problem, showing control over geometrical and topological properties, such as surface smoothness or the number of holes. These results demonstrate the potential of training shape-generative models without data, paving the way for new generative design approaches without large datasets.
♻ ☆ Uplifting Lower-Income Data: Strategies for Socioeconomic Perspective Shifts in Large Multi-modal Models
Recent work has demonstrated that the unequal representation of cultures and socioeconomic groups in training data leads to biased Large Multi-modal (LMM) models. To improve LMM model performance on underrepresented data, we propose and evaluate several prompting strategies using non-English, geographic, and socioeconomic attributes. We show that these geographic and socioeconomic integrated prompts favor retrieving topic appearances commonly found in data from low-income households across different countries leading to improved LMM model performance on lower-income data. Our analyses identify and highlight contexts where these strategies yield the most improvements.
♻ ☆ Hallo2: Long-Duration and High-Resolution Audio-Driven Portrait Image Animation
Recent advances in latent diffusion-based generative models for portrait image animation, such as Hallo, have achieved impressive results in short-duration video synthesis. In this paper, we present updates to Hallo, introducing several design enhancements to extend its capabilities. First, we extend the method to produce long-duration videos. To address substantial challenges such as appearance drift and temporal artifacts, we investigate augmentation strategies within the image space of conditional motion frames. Specifically, we introduce a patch-drop technique augmented with Gaussian noise to enhance visual consistency and temporal coherence over long duration. Second, we achieve 4K resolution portrait video generation. To accomplish this, we implement vector quantization of latent codes and apply temporal alignment techniques to maintain coherence across the temporal dimension. By integrating a high-quality decoder, we realize visual synthesis at 4K resolution. Third, we incorporate adjustable semantic textual labels for portrait expressions as conditional inputs. This extends beyond traditional audio cues to improve controllability and increase the diversity of the generated content. To the best of our knowledge, Hallo2, proposed in this paper, is the first method to achieve 4K resolution and generate hour-long, audio-driven portrait image animations enhanced with textual prompts. We have conducted extensive experiments to evaluate our method on publicly available datasets, including HDTF, CelebV, and our introduced "Wild" dataset. The experimental results demonstrate that our approach achieves state-of-the-art performance in long-duration portrait video animation, successfully generating rich and controllable content at 4K resolution for duration extending up to tens of minutes. Project page https://fudan-generative-vision.github.io/hallo2
♻ ☆ TotalVibeSegmentator: Full Torso Segmentation for the NAKO and UK Biobank in Volumetric Interpolated Breath-hold Examination Body Images
Objectives: To present a publicly available torso segmentation network for large epidemiology datasets on volumetric interpolated breath-hold examination (VIBE) images. Materials & Methods: We extracted preliminary segmentations from TotalSegmentator, spine, and body composition networks for VIBE images, then improved them iteratively and retrained a nnUNet network. Using subsets of NAKO (85 subjects) and UK Biobank (16 subjects), we evaluated with Dice-score on a holdout set (12 subjects) and existing organ segmentation approach (1000 subjects), generating 71 semantic segmentation types for VIBE images. We provide an additional network for the vertebra segments 22 individual vertebra types. Results: We achieved an average Dice score of 0.89 +- 0.07 overall 71 segmentation labels. We scored > 0.90 Dice-score on the abdominal organs except for the pancreas with a Dice of 0.70. Conclusion: Our work offers a detailed and refined publicly available full torso segmentation on VIBE images.
comment: https://github.com/robert-graf/TotalVibeSegmentator
♻ ☆ HAMMR: HierArchical MultiModal React agents for generic VQA
Combining Large Language Models (LLMs) with external specialized tools (LLMs+tools) is a recent paradigm to solve multimodal tasks such as Visual Question Answering (VQA). While this approach was demonstrated to work well when optimized and evaluated for each individual benchmark, in practice it is crucial for the next generation of real-world AI systems to handle a broad range of multimodal problems. Therefore we pose the VQA problem from a unified perspective and evaluate a single system on a varied suite of VQA tasks including counting, spatial reasoning, OCR-based reasoning, visual pointing, external knowledge, and more. In this setting, we demonstrate that naively applying the LLM+tools approach using the combined set of all tools leads to poor results. This motivates us to introduce HAMMR: HierArchical MultiModal React. We start from a multimodal ReAct-based system and make it hierarchical by enabling our HAMMR agents to call upon other specialized agents. This enhances the compositionality of the LLM+tools approach, which we show to be critical for obtaining high accuracy on generic VQA. Concretely, on our generic VQA suite, HAMMR outperforms the naive LLM+tools approach by 19.5%. Additionally, HAMMR achieves state-of-the-art results on this task, outperforming the generic standalone PaLI-X VQA model by 5.0%.
♻ ☆ Benchmarking Vision Language Models for Cultural Understanding EMNLP 2024
Foundation models and vision-language pre-training have notably advanced Vision Language Models (VLMs), enabling multimodal processing of visual and linguistic data. However, their performance has been typically assessed on general scene understanding - recognizing objects, attributes, and actions - rather than cultural comprehension. This study introduces CulturalVQA, a visual question-answering benchmark aimed at assessing VLM's geo-diverse cultural understanding. We curate a collection of 2,378 image-question pairs with 1-5 answers per question representing cultures from 11 countries across 5 continents. The questions probe understanding of various facets of culture such as clothing, food, drinks, rituals, and traditions. Benchmarking VLMs on CulturalVQA, including GPT-4V and Gemini, reveals disparity in their level of cultural understanding across regions, with strong cultural understanding capabilities for North America while significantly lower performance for Africa. We observe disparity in their performance across cultural facets too, with clothing, rituals, and traditions seeing higher performances than food and drink. These disparities help us identify areas where VLMs lack cultural understanding and demonstrate the potential of CulturalVQA as a comprehensive evaluation set for gauging VLM progress in understanding diverse cultures.
comment: Accepted to EMNLP 2024 Main Conference
♻ ☆ SpikeGS: Learning 3D Gaussian Fields from Continuous Spike Stream ACCV 2024
A spike camera is a specialized high-speed visual sensor that offers advantages such as high temporal resolution and high dynamic range compared to conventional frame cameras. These features provide the camera with significant advantages in many computer vision tasks. However, the tasks of novel view synthesis based on spike cameras remain underdeveloped. Although there are existing methods for learning neural radiance fields from spike stream, they either lack robustness in extremely noisy, low-quality lighting conditions or suffer from high computational complexity due to the deep fully connected neural networks and ray marching rendering strategies used in neural radiance fields, making it difficult to recover fine texture details. In contrast, the latest advancements in 3DGS have achieved high-quality real-time rendering by optimizing the point cloud representation into Gaussian ellipsoids. Building on this, we introduce SpikeGS, the method to learn 3D Gaussian fields solely from spike stream. We designed a differentiable spike stream rendering framework based on 3DGS, incorporating noise embedding and spiking neurons. By leveraging the multi-view consistency of 3DGS and the tile-based multi-threaded parallel rendering mechanism, we achieved high-quality real-time rendering results. Additionally, we introduced a spike rendering loss function that generalizes under varying illumination conditions. Our method can reconstruct view synthesis results with fine texture details from a continuous spike stream captured by a moving spike camera, while demonstrating high robustness in extremely noisy low-light scenarios. Experimental results on both real and synthetic datasets demonstrate that our method surpasses existing approaches in terms of rendering quality and speed. Our code will be available at https://github.com/520jz/SpikeGS.
comment: Accepted by ACCV 2024
♻ ☆ ScaleFlow++: Robust and Accurate Estimation of 3D Motion from Video
Perceiving and understanding 3D motion is a core technology in fields such as autonomous driving, robots, and motion prediction. This paper proposes a 3D motion perception method called ScaleFlow++ that is easy to generalize. With just a pair of RGB images, ScaleFlow++ can robustly estimate optical flow and motion-in-depth (MID). Most existing methods directly regress MID from two RGB frames or optical flow, resulting in inaccurate and unstable results. Our key insight is cross-scale matching, which extracts deep motion clues by matching objects in pairs of images at different scales. Unlike previous methods, ScaleFlow++ integrates optical flow and MID estimation into a unified architecture, estimating optical flow and MID end-to-end based on feature matching. Moreover, we also proposed modules such as global initialization network, global iterative optimizer, and hybrid training pipeline to integrate global motion information, reduce the number of iterations, and prevent overfitting during training. On KITTI, ScaleFlow++ achieved the best monocular scene flow estimation performance, reducing SF-all from 6.21 to 5.79. The evaluation of MID even surpasses RGBD-based methods. In addition, ScaleFlow++ has achieved stunning zero-shot generalization performance in both rigid and nonrigid scenes. Code is available at \url{https://github.com/HanLingsgjk/CSCV}.
comment: This is a product uploaded incorrectly. I originally intended to use it to replace ScaleRAFT (arXiv:2407.09797), but I made a mistake in the operation
♻ ☆ Self-Distilled Depth Refinement with Noisy Poisson Fusion NeurIPS 2024
Depth refinement aims to infer high-resolution depth with fine-grained edges and details, refining low-resolution results of depth estimation models. The prevailing methods adopt tile-based manners by merging numerous patches, which lacks efficiency and produces inconsistency. Besides, prior arts suffer from fuzzy depth boundaries and limited generalizability. Analyzing the fundamental reasons for these limitations, we model depth refinement as a noisy Poisson fusion problem with local inconsistency and edge deformation noises. We propose the Self-distilled Depth Refinement (SDDR) framework to enforce robustness against the noises, which mainly consists of depth edge representation and edge-based guidance. With noisy depth predictions as input, SDDR generates low-noise depth edge representations as pseudo-labels by coarse-to-fine self-distillation. Edge-based guidance with edge-guided gradient loss and edge-based fusion loss serves as the optimization objective equivalent to Poisson fusion. When depth maps are better refined, the labels also become more noise-free. Our model can acquire strong robustness to the noises, achieving significant improvements in accuracy, edge quality, efficiency, and generalizability on five different benchmarks. Moreover, directly training another model with edge labels produced by SDDR brings improvements, suggesting that our method could help with training robust refinement models in future works.
comment: Accepted by NeurIPS 2024
♻ ☆ Autoencoded Image Compression for Secure and Fast Transmission
With exponential growth in the use of digital image data, the need for efficient transmission methods has become imperative. Traditional image compression techniques often sacrifice image fidelity for reduced file sizes, challenging maintaining quality and efficiency. They also compromise security, leaving images vulnerable to threats such as man-in-the-middle attacks. This paper proposes an autoencoder architecture for image compression to not only help in dimensionality reduction but also inherently encrypt the images. The paper also introduces a composite loss function that combines reconstruction loss and residual loss for improved performance. The autoencoder architecture is designed to achieve optimal dimensionality reduction and regeneration accuracy while safeguarding the compressed data during transmission or storage. Images regenerated by the autoencoder are evaluated against three key metrics: reconstruction quality, compression ratio, and one-way delay during image transfer. The experiments reveal that the proposed architecture achieves an SSIM of 97.5% over the regenerated images and an average latency reduction of 87.5%, indicating its effectiveness as a secure and efficient solution for compressed image transfer.
comment: 6 pages, 7 figures
♻ ☆ Automatic Classification of White Blood Cell Images using Convolutional Neural Network
Human immune system contains white blood cells (WBC) that are good indicator of many diseases like bacterial infections, AIDS, cancer, spleen, etc. White blood cells have been sub classified into four types: monocytes, lymphocytes, eosinophils and neutrophils on the basis of their nucleus, shape and cytoplasm. Traditionally in laboratories, pathologists and hematologists analyze these blood cells through microscope and then classify them manually. This manual process takes more time and increases the chance of human error. Hence, there is a need to automate this process. In this paper, first we have used different CNN pre-train models such as ResNet-50, InceptionV3, VGG16 and MobileNetV2 to automatically classify the white blood cells. These pre-train models are applied on Kaggle dataset of microscopic images. Although we achieved reasonable accuracy ranging between 92 to 95%, still there is need to enhance the performance. Hence, inspired by these architectures, a framework has been proposed to automatically categorize the four kinds of white blood cells with increased accuracy. The aim is to develop a convolution neural network (CNN) based classification system with decent generalization ability. The proposed CNN model has been tested on white blood cells images from Kaggle and LISC datasets. Accuracy achieved is 99.57% and 98.67% for both datasets respectively. Our proposed convolutional neural network-based model provides competitive performance as compared to previous results reported in literature.
♻ ☆ Gait Sequence Upsampling using Diffusion Models for Single LiDAR Sensors
Recently, 3D LiDAR has emerged as a promising technique in the field of gait-based person identification, serving as an alternative to traditional RGB cameras, due to its robustness under varying lighting conditions and its ability to capture 3D geometric information. However, long capture distances or the use of low-cost LiDAR sensors often result in sparse human point clouds, leading to a decline in identification performance. To address these challenges, we propose a sparse-to-dense upsampling model for pedestrian point clouds in LiDAR-based gait recognition, named LidarGSU, which is designed to improve the generalization capability of existing identification models. Our method utilizes diffusion probabilistic models (DPMs), which have shown high fidelity in generative tasks such as image completion. In this work, we leverage DPMs on sparse sequential pedestrian point clouds as conditional masks in a video-to-video translation approach, applied in an inpainting manner. We conducted extensive experiments on the SUSTeck1K dataset to evaluate the generative quality and recognition performance of the proposed method. Furthermore, we demonstrate the applicability of our upsampling model using a real-world dataset, captured with a low-resolution sensor across varying measurement distances.
♻ ☆ Accelerating Diffusion Transformers with Token-wise Feature Caching
Diffusion transformers have shown significant effectiveness in both image and video synthesis at the expense of huge computation costs. To address this problem, feature caching methods have been introduced to accelerate diffusion transformers by caching the features in previous timesteps and reusing them in the following timesteps. However, previous caching methods ignore that different tokens exhibit different sensitivities to feature caching, and feature caching on some tokens may lead to 10$\times$ more destruction to the overall generation quality compared with other tokens. In this paper, we introduce token-wise feature caching, allowing us to adaptively select the most suitable tokens for caching, and further enable us to apply different caching ratios to neural layers in different types and depths. Extensive experiments on PixArt-$\alpha$, OpenSora, and DiT demonstrate our effectiveness in both image and video generation with no requirements for training. For instance, 2.36$\times$ and 1.93$\times$ acceleration are achieved on OpenSora and PixArt-$\alpha$ with almost no drop in generation quality.
♻ ☆ Exploring Annotation-free Image Captioning with Retrieval-augmented Pseudo Sentence Generation
Recently, training an image captioner without annotated image-sentence pairs has gained traction. Previous methods have faced limitations due to either using mismatched corpora for inaccurate pseudo annotations or relying on resource-intensive pre-training. To alleviate these challenges, we propose a new strategy where the prior knowledge from large pre-trained models (LPMs) is distilled and leveraged as supervision, and a retrieval process is integrated to further reinforce its effectiveness. Specifically, we introduce Retrieval-augmented Pseudo Sentence Generation (RaPSG), which can efficiently retrieve highly relevant short region descriptions from the mismatching corpora and use them to generate a variety of high-quality pseudo sentences via LPMs. Additionally, we introduce a fluency filter and a CLIP guidance objective to enhance contrastive information learning. Experimental results indicate that our method outperforms SOTA captioning models across various settings including zero-shot, unsupervised, semi-supervised, and cross-domain scenarios. Code is available at: https://github.com/Zhiyuan-Li-John/RaPSG.
comment: Accepted by ACM Multimedia Asia 2024
♻ ☆ Improving Consistency Models with Generator-Induced Flows
Consistency models imitate the multi-step sampling of score-based diffusion in a single forward pass of a neural network. They can be learned in two ways: consistency distillation and consistency training. The former relies on the true velocity field of the corresponding differential equation, approximated by a pre-trained neural network. In contrast, the latter uses a single-sample Monte Carlo estimate of this velocity field. The related estimation error induces a discrepancy between consistency distillation and training that, we show, still holds in the continuous-time limit. To alleviate this issue, we propose a novel flow that transports noisy data towards their corresponding outputs derived from the currently trained model --~as a proxy of the true flow. Our empirical findings demonstrate that this approach mitigates the previously identified discrepancy. Furthermore, we present theoretical and empirical evidence indicating that our generator-induced flow surpasses dedicated optimal transport-based consistency models in effectively reducing the noise-data transport cost. Consequently, our method not only accelerates consistency training convergence but also enhances its overall performance. The code is available at: https://github.com/thibautissenhuth/consistency_GC.
♻ ☆ Learning to Balance: Diverse Normalization for Cloth-Changing Person Re-Identification
Cloth-Changing Person Re-Identification (CC-ReID) involves recognizing individuals in images regardless of clothing status. In this paper, we empirically and experimentally demonstrate that completely eliminating or fully retaining clothing features is detrimental to the task. Existing work, either relying on clothing labels, silhouettes, or other auxiliary data, fundamentally aim to balance the learning of clothing and identity features. However, we practically find that achieving this balance is challenging and nuanced. In this study, we introduce a novel module called Diverse Norm, which expands personal features into orthogonal spaces and employs channel attention to separate clothing and identity features. A sample re-weighting optimization strategy is also introduced to guarantee the opposite optimization direction. Diverse Norm presents a simple yet effective approach that does not require additional data. Furthermore, Diverse Norm can be seamlessly integrated ResNet50 and significantly outperforms the state-of-the-art methods.
♻ ☆ Generative inpainting of incomplete Euclidean distance matrices of trajectories generated by a fractional Brownian motion
Fractional Brownian motion (fBm) features both randomness and strong scale-free correlations, challenging generative models to reproduce the intrinsic memory characterizing the underlying stochastic process. Here we examine a zoo of diffusion-based inpainting methods on a specific dataset of corrupted images, which represent incomplete Euclidean distance matrices (EDMs) of fBm at various memory exponents $H$. Our dataset implies uniqueness of the data imputation in the regime of low missing ratio, where the remaining partial graph is rigid, providing the ground truth for the inpainting. We find that the conditional diffusion generation readily reproduces the built-in correlations of fBm paths in different memory regimes (i.e., for sub-, Brownian and super-diffusion trajectories), providing a robust tool for the statistical imputation at high missing ratio. Furthermore, while diffusion models have been recently shown to memorize samples from the training database, we demonstrate that diffusion behaves qualitatively different from the database search and thus generalize rather than memorize the training dataset. As a biological application, we apply our fBm-trained diffusion model for the imputation of microscopy-derived distance matrices of chromosomal segments (FISH data) - incomplete due to experimental imperfections - and demonstrate its superiority over the standard approaches used in bioinformatics.
♻ ☆ Improving Multimodal Learning with Multi-Loss Gradient Modulation
Learning from multiple modalities, such as audio and video, offers opportunities for leveraging complementary information, enhancing robustness, and improving contextual understanding and performance. However, combining such modalities presents challenges, especially when modalities differ in data structure, predictive contribution, and the complexity of their learning processes. It has been observed that one modality can potentially dominate the learning process, hindering the effective utilization of information from other modalities and leading to sub-optimal model performance. To address this issue the vast majority of previous works suggest to assess the unimodal contributions and dynamically adjust the training to equalize them. We improve upon previous work by introducing a multi-loss objective and further refining the balancing process, allowing it to dynamically adjust the learning pace of each modality in both directions, acceleration and deceleration, with the ability to phase out balancing effects upon convergence. We achieve superior results across three audio-video datasets: on CREMA-D, models with ResNet backbone encoders surpass the previous best by 1.9% to 12.4%, and Conformer backbone models deliver improvements ranging from 2.8% to 14.1% across different fusion methods. On AVE, improvements range from 2.7% to 7.7%, while on UCF101, gains reach up to 6.1%.
♻ ☆ LIME-Eval: Rethinking Low-light Image Enhancement Evaluation via Object Detection
Due to the nature of enhancement--the absence of paired ground-truth information, high-level vision tasks have been recently employed to evaluate the performance of low-light image enhancement. A widely-used manner is to see how accurately an object detector trained on enhanced low-light images by different candidates can perform with respect to annotated semantic labels. In this paper, we first demonstrate that the mentioned approach is generally prone to overfitting, and thus diminishes its measurement reliability. In search of a proper evaluation metric, we propose LIME-Bench, the first online benchmark platform designed to collect human preferences for low-light enhancement, providing a valuable dataset for validating the correlation between human perception and automated evaluation metrics. We then customize LIME-Eval, a novel evaluation framework that utilizes detectors pre-trained on standard-lighting datasets without object annotations, to judge the quality of enhanced images. By adopting an energy-based strategy to assess the accuracy of output confidence maps, our LIME-Eval can simultaneously bypass biases associated with retraining detectors and circumvent the reliance on annotations for dim images. Comprehensive experiments are provided to reveal the effectiveness of our LIME-Eval. Our benchmark platform (https://huggingface.co/spaces/lime-j/eval) and code (https://github.com/lime-j/lime-eval) are available online.
♻ ☆ RealCompo: Balancing Realism and Compositionality Improves Text-to-Image Diffusion Models NeurIPS 2024
Diffusion models have achieved remarkable advancements in text-to-image generation. However, existing models still have many difficulties when faced with multiple-object compositional generation. In this paper, we propose RealCompo, a new training-free and transferred-friendly text-to-image generation framework, which aims to leverage the respective advantages of text-to-image models and spatial-aware image diffusion models (e.g., layout, keypoints and segmentation maps) to enhance both realism and compositionality of the generated images. An intuitive and novel balancer is proposed to dynamically balance the strengths of the two models in denoising process, allowing plug-and-play use of any model without extra training. Extensive experiments show that our RealCompo consistently outperforms state-of-the-art text-to-image models and spatial-aware image diffusion models in multiple-object compositional generation while keeping satisfactory realism and compositionality of the generated images. Notably, our RealCompo can be seamlessly extended with a wide range of spatial-aware image diffusion models and stylized diffusion models. Our code is available at: https://github.com/YangLing0818/RealCompo
comment: NeurIPS 2024. Project: https://github.com/YangLing0818/RealCompo
♻ ☆ ADFQ-ViT: Activation-Distribution-Friendly Post-Training Quantization for Vision Transformers
Vision Transformers (ViTs) have exhibited exceptional performance across diverse computer vision tasks, while their substantial parameter size incurs significantly increased memory and computational demands, impeding effective inference on resource-constrained devices. Quantization has emerged as a promising solution to mitigate these challenges, yet existing methods still suffer from significant accuracy loss at low-bit. We attribute this issue to the distinctive distributions of post-LayerNorm and post-GELU activations within ViTs, rendering conventional hardware-friendly quantizers ineffective, particularly in low-bit scenarios. To address this issue, we propose a novel framework called Activation-Distribution-Friendly post-training Quantization for Vision Transformers, ADFQ-ViT. Concretely, we introduce the Per-Patch Outlier-aware Quantizer to tackle irregular outliers in post-LayerNorm activations. This quantizer refines the granularity of the uniform quantizer to a per-patch level while retaining a minimal subset of values exceeding a threshold at full-precision. To handle the non-uniform distributions of post-GELU activations between positive and negative regions, we design the Shift-Log2 Quantizer, which shifts all elements to the positive region and then applies log2 quantization. Moreover, we present the Attention-score enhanced Module-wise Optimization which adjusts the parameters of each quantizer by reconstructing errors to further mitigate quantization error. Extensive experiments demonstrate ADFQ-ViT provides significant improvements over various baselines in image classification, object detection, and instance segmentation tasks at 4-bit. Specifically, when quantizing the ViT-B model to 4-bit, we achieve a 10.23% improvement in Top-1 accuracy on the ImageNet dataset.
comment: 29 pages,9 figures
♻ ☆ VideoTetris: Towards Compositional Text-to-Video Generation NeurIPS 2024
Diffusion models have demonstrated great success in text-to-video (T2V) generation. However, existing methods may face challenges when handling complex (long) video generation scenarios that involve multiple objects or dynamic changes in object numbers. To address these limitations, we propose VideoTetris, a novel framework that enables compositional T2V generation. Specifically, we propose spatio-temporal compositional diffusion to precisely follow complex textual semantics by manipulating and composing the attention maps of denoising networks spatially and temporally. Moreover, we propose an enhanced video data preprocessing to enhance the training data regarding motion dynamics and prompt understanding, equipped with a new reference frame attention mechanism to improve the consistency of auto-regressive video generation. Extensive experiments demonstrate that our VideoTetris achieves impressive qualitative and quantitative results in compositional T2V generation. Code is available at: https://github.com/YangLing0818/VideoTetris
comment: NeurIPS 2024. Code: https://github.com/YangLing0818/VideoTetris
♻ ☆ AdaPose: Towards Cross-Site Device-Free Human Pose Estimation with Commodity WiFi
WiFi-based pose estimation is a technology with great potential for the development of smart homes and metaverse avatar generation. However, current WiFi-based pose estimation methods are predominantly evaluated under controlled laboratory conditions with sophisticated vision models to acquire accurately labeled data. Furthermore, WiFi CSI is highly sensitive to environmental variables, and direct application of a pre-trained model to a new environment may yield suboptimal results due to domain shift. In this paper, we proposes a domain adaptation algorithm, AdaPose, designed specifically for weakly-supervised WiFi-based pose estimation. The proposed method aims to identify consistent human poses that are highly resistant to environmental dynamics. To achieve this goal, we introduce a Mapping Consistency Loss that aligns the domain discrepancy of source and target domains based on inner consistency between input and output at the mapping level. We conduct extensive experiments on domain adaptation in two different scenes using our self-collected pose estimation dataset containing WiFi CSI frames. The results demonstrate the effectiveness and robustness of AdaPose in eliminating domain shift, thereby facilitating the widespread application of WiFi-based pose estimation in smart cities.
♻ ☆ Multi-scale direction-aware SAR object detection network via global information fusion
Deep learning has driven significant progress in object detection using Synthetic Aperture Radar (SAR) imagery. Existing methods, while achieving promising results, often struggle to effectively integrate local and global information, particularly direction-aware features. This paper proposes SAR-Net, a novel framework specifically designed for global fusion of direction-aware information in SAR object detection. SAR-Net leverages two key innovations: the Unity Compensation Mechanism (UCM) and the Direction-aware Attention Module (DAM). UCM facilitates the establishment of complementary relationships among features across different scales, enabling efficient global information fusion and transmission. Additionally, DAM, through bidirectional attention polymerization, captures direction-aware information, effectively eliminating background interference. Extensive experiments demonstrate the effectiveness of SAR-Net, achieving state-of-the-art results on aircraft (SAR-AIRcraft-1.0) and ship datasets (SSDD, HRSID), confirming its generalization capability and robustness.
♻ ☆ The Surprising Effectiveness of Multimodal Large Language Models for Video Moment Retrieval
Recent studies have shown promising results in utilizing multimodal large language models (MLLMs) for computer vision tasks such as object detection and semantic segmentation. However, many challenging video tasks remain under-explored. Video-language tasks necessitate spatial and temporal comprehension and require significant compute. Therefore, prior works have developed complex, highly specialized architectures or leveraged additional input signals such as video transcripts to best encode contextual and temporal information, which limits their generality and can be impractical. One particularly challenging task is video moment retrieval, which requires precise temporal and contextual grounding. This work demonstrates the surprising effectiveness of leveraging image-text pretrained MLLMs for moment retrieval. We introduce Mr. BLIP (Mr. as in Moment Retrieval), a multimodal, single-stage model that requires no expensive video-language pretraining, no additional input signal (e.g., no transcript or audio), and has a simpler and more versatile design than prior state-of-the-art methods. We achieve a new state-of-the-art in moment retrieval on the widely used benchmarks Charades-STA, QVHighlights, and ActivityNet Captions. Notably, we attain over 9% (absolute) higher Recall (at 0.5 and 0.7 IoU) on the challenging long-video multi-moment QVHighlights benchmark. Our code is publicly available.
comment: Code: https://github.com/sudo-Boris/mr-Blip
♻ ☆ DD-RobustBench: An Adversarial Robustness Benchmark for Dataset Distillation
Dataset distillation is an advanced technique aimed at compressing datasets into significantly smaller counterparts, while preserving formidable training performance. Significant efforts have been devoted to promote evaluation accuracy under limited compression ratio while overlooked the robustness of distilled dataset. In this work, we introduce a comprehensive benchmark that, to the best of our knowledge, is the most extensive to date for evaluating the adversarial robustness of distilled datasets in a unified way. Our benchmark significantly expands upon prior efforts by incorporating a wider range of dataset distillation methods, including the latest advancements such as TESLA and SRe2L, a diverse array of adversarial attack methods, and evaluations across a broader and more extensive collection of datasets such as ImageNet-1K. Moreover, we assessed the robustness of these distilled datasets against representative adversarial attack algorithms like PGD and AutoAttack, while exploring their resilience from a frequency perspective. We also discovered that incorporating distilled data into the training batches of the original dataset can yield to improvement of robustness.
comment: * denotes equal contributions; ^ denotes corresponding author. In this updated version, we have expanded our research to include more experiments on various adversarial attack methods and latest dataset distillation studies. All new results have been incorporated into the document
♻ ☆ MT-SNN: Enhance Spiking Neural Network with Multiple Thresholds
Spiking neural networks (SNNs) present a promising energy efficient alternative to traditional Artificial Neural Networks (ANNs) due to their multiplication-free operations enabled by binarized intermediate activations. However, this binarization leads to precision loss, hindering the SNN performance. In this paper, we introduce Multiple Threshold (MT) approaches to significantly enhance SNN accuracy by mitigating precision loss. We propose two distinct modes for MT implementation, depending on the membrane update rule: parallel mode and cascade mode. MT-SNN models can be efficiently trained on standard hardwares like GPUs and TPUs, while retaining the multiplication-free advantage crucial for deployment on neuromorphic devices. Our extensive experiments on CIFAR10, CIFAR100, ImageNet, and DVS-CIFAR10 datasets demonstrate that both MT modes substantially improve the performance of single-threshold SNNs, achieving higher accuracy with fewer time steps and comparable energy consumption. Moreover, MT-SNNs outperform state-of-the-art (SOTA) results. Notably, with MT, a Parametric-Leaky-Integrate-Fire (PLIF) based ResNet-34 architecture reaches 72.17\% accuracy on ImageNet with a single time step, surpassing the previous SOTA by 2.75\% despite using 4 steps.
comment: 16 pages, 5 figures, 4 tables
♻ ☆ Twisting Lids Off with Two Hands
Manipulating objects with two multi-fingered hands has been a long-standing challenge in robotics, due to the contact-rich nature of many manipulation tasks and the complexity inherent in coordinating a high-dimensional bimanual system. In this work, we share novel insights into physical modeling, real-time perception, and reward design that enable policies trained in simulation using deep reinforcement learning (RL) to be effectively and efficiently transferred to the real world. Specifically, we consider the problem of twisting lids of various bottle-like objects with two hands, demonstrating policies with generalization capabilities across a diverse set of unseen objects as well as dynamic and dexterous behaviors. To the best of our knowledge, this is the first sim-to-real RL system that enables such capabilities on bimanual multi-fingered hands.
comment: Project page can be found at https://toruowo.github.io/bimanual-twist
♻ ☆ Advancing Supervised Local Learning Beyond Classification with Long-term Feature Bank
Local learning offers an alternative to traditional end-to-end back-propagation in deep neural networks, significantly reducing GPU memory usage. While local learning has shown promise in image classification tasks, its application to other visual tasks remains limited. This limitation arises primarily from two factors: 1) architectures tailored for classification are often not transferable to other tasks, leading to a lack of reusability of task-specific knowledge; 2) the absence of cross-scale feature communication results in degraded performance in tasks such as object detection and super-resolution. To address these challenges, we propose the Memory-augmented Auxiliary Network (MAN), which introduces a simplified design principle and incorporates a feature bank to enhance cross-task adaptability and communication. This work represents the first successful application of local learning methods beyond classification, demonstrating that MAN not only conserves GPU memory but also achieves performance on par with end-to-end approaches across multiple datasets for various visual tasks.
Fusion-Driven Tree Reconstruction and Fruit Localization: Advancing Precision in Agriculture IROS
Fruit distribution is pivotal in shaping the future of both agriculture and agricultural robotics, paving the way for a streamlined supply chain. This study introduces an innovative methodology that harnesses the synergy of RGB imagery, LiDAR, and IMU data, to achieve intricate tree reconstructions and the pinpoint localization of fruits. Such integration not only offers insights into the fruit distribution, which enhances the precision of guidance for agricultural robotics and automation systems, but also sets the stage for simulating synthetic fruit patterns across varied tree architectures. To validate this approach, experiments have been carried out in both a controlled environment and an actual peach orchard. The results underscore the robustness and efficacy of this fusion-driven methodology, highlighting its potential as a transformative tool for future agricultural robotics and precision farming.
comment: This work was presented at IEEE/RSI International Conference on Intelligent Robots and Systems (IROS) Workshop
♻ ☆ AnyDesign: Versatile Area Fashion Editing via Mask-Free Diffusion
Fashion image editing aims to modify a person's appearance based on a given instruction. Existing methods require auxiliary tools like segmenters and keypoint extractors, lacking a flexible and unified framework. Moreover, these methods are limited in the variety of clothing types they can handle, as most datasets focus on people in clean backgrounds and only include generic garments such as tops, pants, and dresses. These limitations restrict their applicability in real-world scenarios. In this paper, we first extend an existing dataset for human generation to include a wider range of apparel and more complex backgrounds. This extended dataset features people wearing diverse items such as tops, pants, dresses, skirts, headwear, scarves, shoes, socks, and bags. Additionally, we propose AnyDesign, a diffusion-based method that enables mask-free editing on versatile areas. Users can simply input a human image along with a corresponding prompt in either text or image format. Our approach incorporates Fashion DiT, equipped with a Fashion-Guidance Attention (FGA) module designed to fuse explicit apparel types and CLIP-encoded apparel features. Both Qualitative and quantitative experiments demonstrate that our method delivers high-quality fashion editing and outperforms contemporary text-guided fashion editing methods.
♻ ☆ Sim-to-Real Transfer via 3D Feature Fields for Vision-and-Language Navigation
Vision-and-language navigation (VLN) enables the agent to navigate to a remote location in 3D environments following the natural language instruction. In this field, the agent is usually trained and evaluated in the navigation simulators, lacking effective approaches for sim-to-real transfer. The VLN agents with only a monocular camera exhibit extremely limited performance, while the mainstream VLN models trained with panoramic observation, perform better but are difficult to deploy on most monocular robots. For this case, we propose a sim-to-real transfer approach to endow the monocular robots with panoramic traversability perception and panoramic semantic understanding, thus smoothly transferring the high-performance panoramic VLN models to the common monocular robots. In this work, the semantic traversable map is proposed to predict agent-centric navigable waypoints, and the novel view representations of these navigable waypoints are predicted through the 3D feature fields. These methods broaden the limited field of view of the monocular robots and significantly improve navigation performance in the real world. Our VLN system outperforms previous SOTA monocular VLN methods in R2R-CE and RxR-CE benchmarks within the simulation environments and is also validated in real-world environments, providing a practical and high-performance solution for real-world VLN.
comment: Accepted by CoRL 2024. The code is available at https://github.com/MrZihan/Sim2Real-VLN-3DFF
♻ ☆ Designing Extremely Memory-Efficient CNNs for On-device Vision Tasks
In this paper, we introduce a memory-efficient CNN (convolutional neural network), which enables resource-constrained low-end embedded and IoT devices to perform on-device vision tasks, such as image classification and object detection, using extremely low memory, i.e., only 63 KB on ImageNet classification. Based on the bottleneck block of MobileNet, we propose three design principles that significantly curtail the peak memory usage of a CNN so that it can fit the limited KB memory of the low-end device. First, 'input segmentation' divides an input image into a set of patches, including the central patch overlapped with the others, reducing the size (and memory requirement) of a large input image. Second, 'patch tunneling' builds independent tunnel-like paths consisting of multiple bottleneck blocks per patch, penetrating through the entire model from an input patch to the last layer of the network, maintaining lightweight memory usage throughout the whole network. Lastly, 'bottleneck reordering' rearranges the execution order of convolution operations inside the bottleneck block such that the memory usage remains constant regardless of the size of the convolution output channels. The experiment result shows that the proposed network classifies ImageNet with extremely low memory (i.e., 63 KB) while achieving competitive top-1 accuracy (i.e., 61.58\%). To the best of our knowledge, the memory usage of the proposed network is far smaller than state-of-the-art memory-efficient networks, i.e., up to 89x and 3.1x smaller than MobileNet (i.e., 5.6 MB) and MCUNet (i.e., 196 KB), respectively.
♻ ☆ Unsupervised Point Cloud Completion through Unbalanced Optimal Transport
Unpaired point cloud completion explores methods for learning a completion map from unpaired incomplete and complete point cloud data. In this paper, we propose a novel approach for unpaired point cloud completion using the unbalanced optimal transport map, called Unbalanced Optimal Transport Map for Unpaired Point Cloud Completion (UOT-UPC). We demonstrate that the unpaired point cloud completion can be naturally interpreted as the Optimal Transport (OT) problem and introduce the Unbalanced Optimal Transport (UOT) approach to address the class imbalance problem, which is prevalent in unpaired point cloud completion datasets. Moreover, we analyze the appropriate cost function for unpaired completion tasks. This analysis shows that the InfoCD cost function is particularly well-suited for this task. Our model is the first attempt to leverage UOT for unpaired point cloud completion, achieving competitive or superior results on both single-category and multi-category datasets. In particular, our model is especially effective in scenarios with class imbalance, where the proportions of categories are different between the incomplete and complete point cloud datasets.
comment: 20 pages, 10 figures
♻ ☆ Neural Collapse Meets Differential Privacy: Curious Behaviors of NoisyGD with Near-perfect Representation Learning ICML 2024
A recent study by De et al. (2022) has reported that large-scale representation learning through pre-training on a public dataset significantly enhances differentially private (DP) learning in downstream tasks, despite the high dimensionality of the feature space. To theoretically explain this phenomenon, we consider the setting of a layer-peeled model in representation learning, which results in interesting phenomena related to learned features in deep learning and transfer learning, known as Neural Collapse (NC). Within the framework of NC, we establish an error bound indicating that the misclassification error is independent of dimension when the distance between actual features and the ideal ones is smaller than a threshold. Additionally, the quality of the features in the last layer is empirically evaluated under different pre-trained models within the framework of NC, showing that a more powerful transformer leads to a better feature representation. Furthermore, we reveal that DP fine-tuning is less robust compared to fine-tuning without DP, particularly in the presence of perturbations. These observations are supported by both theoretical analyses and experimental evaluation. Moreover, to enhance the robustness of DP fine-tuning, we suggest several strategies, such as feature normalization or employing dimension reduction methods like Principal Component Analysis (PCA). Empirically, we demonstrate a significant improvement in testing accuracy by conducting PCA on the last-layer features.
comment: ICML 2024 (oral)
♻ ☆ VideoScore: Building Automatic Metrics to Simulate Fine-grained Human Feedback for Video Generation
The recent years have witnessed great advances in video generation. However, the development of automatic video metrics is lagging significantly behind. None of the existing metric is able to provide reliable scores over generated videos. The main barrier is the lack of large-scale human-annotated dataset. In this paper, we release VideoFeedback, the first large-scale dataset containing human-provided multi-aspect score over 37.6K synthesized videos from 11 existing video generative models. We train VideoScore (initialized from Mantis) based on VideoFeedback to enable automatic video quality assessment. Experiments show that the Spearman correlation between VideoScore and humans can reach 77.1 on VideoFeedback-test, beating the prior best metrics by about 50 points. Further result on other held-out EvalCrafter, GenAI-Bench, and VBench show that VideoScore has consistently much higher correlation with human judges than other metrics. Due to these results, we believe VideoScore can serve as a great proxy for human raters to (1) rate different video models to track progress (2) simulate fine-grained human feedback in Reinforcement Learning with Human Feedback (RLHF) to improve current video generation models.
♻ ☆ Unraveling the Smoothness Properties of Diffusion Models: A Gaussian Mixture Perspective
Diffusion models have made rapid progress in generating high-quality samples across various domains. However, a theoretical understanding of the Lipschitz continuity and second momentum properties of the diffusion process is still lacking. In this paper, we bridge this gap by providing a detailed examination of these smoothness properties for the case where the target data distribution is a mixture of Gaussians, which serves as a universal approximator for smooth densities such as image data. We prove that if the target distribution is a $k$-mixture of Gaussians, the density of the entire diffusion process will also be a $k$-mixture of Gaussians. We then derive tight upper bounds on the Lipschitz constant and second momentum that are independent of the number of mixture components $k$. Finally, we apply our analysis to various diffusion solvers, both SDE and ODE based, to establish concrete error guarantees in terms of the total variation distance and KL divergence between the target and learned distributions. Our results provide deeper theoretical insights into the dynamics of the diffusion process under common data distributions.
♻ ☆ ParCon: Noise-Robust Collaborative Perception via Multi-module Parallel Connection ICLR 2025
In this paper, we investigate improving the perception performance of autonomous vehicles through communication with other vehicles and road infrastructures. To this end, we introduce a novel collaborative perception architecture, called ParCon, which connects multiple modules in parallel, as opposed to the sequential connections used in most other collaborative perception methods. Through extensive experiments, we demonstrate that ParCon inherits the advantages of parallel connection. Specifically, ParCon is robust to noise, as the parallel architecture allows each module to manage noise independently and complement the limitations of other modules. As a result, ParCon achieves state-of-the-art accuracy, particularly in noisy environments, such as real-world datasets, increasing detection accuracy by 6.91%. Additionally, ParCon is computationally efficient, reducing floating-point operations (FLOPs) by 11.46%.
comment: 20pages, under review at ICLR 2025
♻ ☆ JointDreamer: Ensuring Geometry Consistency and Text Congruence in Text-to-3D Generation via Joint Score Distillation ECCV2024
Score Distillation Sampling (SDS) by well-trained 2D diffusion models has shown great promise in text-to-3D generation. However, this paradigm distills view-agnostic 2D image distributions into the rendering distribution of 3D representation for each view independently, overlooking the coherence across views and yielding 3D inconsistency in generations. In this work, we propose \textbf{J}oint \textbf{S}core \textbf{D}istillation (JSD), a new paradigm that ensures coherent 3D generations. Specifically, we model the joint image distribution, which introduces an energy function to capture the coherence among denoised images from the diffusion model. We then derive the joint score distillation on multiple rendered views of the 3D representation, as opposed to a single view in SDS. In addition, we instantiate three universal view-aware models as energy functions, demonstrating compatibility with JSD. Empirically, JSD significantly mitigates the 3D inconsistency problem in SDS, while maintaining text congruence. Moreover, we introduce the Geometry Fading scheme and Classifier-Free Guidance (CFG) Switching strategy to enhance generative details. Our framework, JointDreamer, establishes a new benchmark in text-to-3D generation, achieving outstanding results with an 88.5\% CLIP R-Precision and 27.7\% CLIP Score. These metrics demonstrate exceptional text congruence, as well as remarkable geometric consistency and texture fidelity.
comment: 29 pages, ECCV2024
♻ ☆ White-box Multimodal Jailbreaks Against Large Vision-Language Models
Recent advancements in Large Vision-Language Models (VLMs) have underscored their superiority in various multimodal tasks. However, the adversarial robustness of VLMs has not been fully explored. Existing methods mainly assess robustness through unimodal adversarial attacks that perturb images, while assuming inherent resilience against text-based attacks. Different from existing attacks, in this work we propose a more comprehensive strategy that jointly attacks both text and image modalities to exploit a broader spectrum of vulnerability within VLMs. Specifically, we propose a dual optimization objective aimed at guiding the model to generate affirmative responses with high toxicity. Our attack method begins by optimizing an adversarial image prefix from random noise to generate diverse harmful responses in the absence of text input, thus imbuing the image with toxic semantics. Subsequently, an adversarial text suffix is integrated and co-optimized with the adversarial image prefix to maximize the probability of eliciting affirmative responses to various harmful instructions. The discovered adversarial image prefix and text suffix are collectively denoted as a Universal Master Key (UMK). When integrated into various malicious queries, UMK can circumvent the alignment defenses of VLMs and lead to the generation of objectionable content, known as jailbreaks. The experimental results demonstrate that our universal attack strategy can effectively jailbreak MiniGPT-4 with a 96% success rate, highlighting the vulnerability of VLMs and the urgent need for new alignment strategies.
♻ ☆ A Review of Electromagnetic Elimination Methods for low-field portable MRI scanner
This paper analyzes conventional and deep learning methods for eliminating electromagnetic interference (EMI) in MRI systems. We compare traditional analytical and adaptive techniques with advanced deep learning approaches. Key strengths and limitations of each method are highlighted. Recent advancements in active EMI elimination, such as external EMI receiver coils, are discussed alongside deep learning methods, which show superior EMI suppression by leveraging neural networks trained on MRI data. While deep learning improves EMI elimination and diagnostic capabilities, it introduces security and safety concerns, particularly in commercial applications. A balanced approach, integrating conventional reliability with deep learning's advanced capabilities, is proposed for more effective EMI suppression in MRI systems.
comment: Accepted by 2024 5th International Conference on Machine Learning and Computer Application
♻ ☆ Bridging the Gap between Text, Audio, Image, and Any Sequence: A Novel Approach using Gloss-based Annotation
This paper presents an innovative approach called BGTAI to simplify multimodal understanding by utilizing gloss-based annotation as an intermediate step in aligning Text and Audio with Images. While the dynamic temporal factors in textual and audio inputs contain various predicate adjectives that influence the meaning of the entire sentence, images, on the other hand, present static scenes. By representing text and audio as gloss notations that omit complex semantic nuances, a better alignment with images can potentially be achieved. This study explores the feasibility of this idea, specifically, we first propose the first Langue2Gloss model and then integrate it into the multimodal model UniBriVL for joint training. To strengthen the adaptability of gloss with text/audio and overcome the efficiency and instability issues in multimodal training, we propose a DS-Net (Data-Pair Selection Network), an Result Filter module, and a novel SP-Loss function. Our approach outperforms previous multimodal models in the main experiments, demonstrating its efficacy in enhancing multimodal representations and improving compatibility among text, audio, visual, and any sequence modalities.
♻ ☆ Edit-Your-Motion: Space-Time Diffusion Decoupling Learning for Video Motion Editing
Existing diffusion-based video editing methods have achieved impressive results in motion editing. Most of the existing methods focus on the motion alignment between the edited video and the reference video. However, these methods do not constrain the background and object content of the video to remain unchanged, which makes it possible for users to generate unexpected videos. In this paper, we propose a one-shot video motion editing method called Edit-Your-Motion that requires only a single text-video pair for training. Specifically, we design the Detailed Prompt-Guided Learning Strategy (DPL) to decouple spatio-temporal features in space-time diffusion models. DPL separates learning object content and motion into two training stages. In the first training stage, we focus on learning the spatial features (the features of object content) and breaking down the temporal relationships in the video frames by shuffling them. We further propose Recurrent-Causal Attention (RC-Attn) to learn the consistent content features of the object from unordered video frames. In the second training stage, we restore the temporal relationship in video frames to learn the temporal feature (the features of the background and object's motion). We also adopt the Noise Constraint Loss to smooth out inter-frame differences. Finally, in the inference stage, we inject the content features of the source object into the editing branch through a two-branch structure (editing branch and reconstruction branch). With Edit-Your-Motion, users can edit the motion of objects in the source video to generate more exciting and diverse videos. Comprehensive qualitative experiments, quantitative experiments and user preference studies demonstrate that Edit-Your-Motion performs better than other methods.
♻ ☆ MedSegMamba: 3D CNN-Mamba Hybrid Architecture for Brain Segmentation
Widely used traditional pipelines for subcortical brain segmentation are often inefficient and slow, particularly when processing large datasets. Furthermore, deep learning models face challenges due to the high resolution of MRI images and the large number of anatomical classes involved. To address these limitations, we developed a 3D patch-based hybrid CNN-Mamba model that leverages Mamba's selective scan algorithm, thereby enhancing segmentation accuracy and efficiency for 3D inputs. This retrospective study utilized 1784 T1-weighted MRI scans from a diverse, multi-site dataset of healthy individuals. The dataset was divided into training, validation, and testing sets with a 1076/345/363 split. The scans were obtained from 1.5T and 3T MRI machines. Our model's performance was validated against several benchmarks, including other CNN-Mamba, CNN-Transformer, and pure CNN networks, using FreeSurfer-generated ground truths. We employed the Dice Similarity Coefficient (DSC), Volume Similarity (VS), and Average Symmetric Surface Distance (ASSD) as evaluation metrics. Statistical significance was determined using the Wilcoxon signed-rank test with a threshold of P < 0.05. The proposed model achieved the highest overall performance across all metrics (DSC 0.88383; VS 0.97076; ASSD 0.33604), significantly outperforming all non-Mamba-based models (P < 0.001). While the model did not show significant improvement in DSC or VS compared to another Mamba-based model (P-values of 0.114 and 0.425), it demonstrated a significant enhancement in ASSD (P < 0.001) with approximately 20% fewer parameters. In conclusion, our proposed hybrid CNN-Mamba architecture offers an efficient and accurate approach for 3D subcortical brain segmentation, demonstrating potential advantages over existing methods. Code is available at: https://github.com/aaroncao06/MedSegMamba.
comment: 14 pages, 8 figures
♻ ☆ Show-o: One Single Transformer to Unify Multimodal Understanding and Generation
We present a unified transformer, i.e., Show-o, that unifies multimodal understanding and generation. Unlike fully autoregressive models, Show-o unifies autoregressive and (discrete) diffusion modeling to adaptively handle inputs and outputs of various and mixed modalities. The unified model flexibly supports a wide range of vision-language tasks including visual question-answering, text-to-image generation, text-guided inpainting/extrapolation, and mixed-modality generation. Across various benchmarks, it demonstrates comparable or superior performance to existing individual models with an equivalent or larger number of parameters tailored for understanding or generation. This significantly highlights its potential as a next-generation foundation model. Code and models are released at https://github.com/showlab/Show-o.
comment: Technical Report
♻ ☆ Exploring the Potential of Polynomial Basis Functions in Kolmogorov-Arnold Networks: A Comparative Study of Different Groups of Polynomials
This paper presents a comprehensive survey of 18 distinct polynomials and their potential applications in Kolmogorov-Arnold Network (KAN) models as an alternative to traditional spline-based methods. The polynomials are classified into various groups based on their mathematical properties, such as orthogonal polynomials, hypergeometric polynomials, q-polynomials, Fibonacci-related polynomials, combinatorial polynomials, and number-theoretic polynomials. The study aims to investigate the suitability of these polynomials as basis functions in KAN models for complex tasks like handwritten digit classification on the MNIST dataset. The performance metrics of the KAN models, including overall accuracy, Kappa, and F1 score, are evaluated and compared. The Gottlieb-KAN model achieves the highest performance across all metrics, suggesting its potential as a suitable choice for the given task. However, further analysis and tuning of these polynomials on more complex datasets are necessary to fully understand their capabilities in KAN models. The source code for the implementation of these KAN models is available at https://github.com/seydi1370/Basis_Functions .
♻ ☆ SceneDreamer360: Text-Driven 3D-Consistent Scene Generation with Panoramic Gaussian Splatting
Text-driven 3D scene generation has seen significant advancements recently. However, most existing methods generate single-view images using generative models and then stitch them together in 3D space. This independent generation for each view often results in spatial inconsistency and implausibility in the 3D scenes. To address this challenge, we proposed a novel text-driven 3D-consistent scene generation model: SceneDreamer360. Our proposed method leverages a text-driven panoramic image generation model as a prior for 3D scene generation and employs 3D Gaussian Splatting (3DGS) to ensure consistency across multi-view panoramic images. Specifically, SceneDreamer360 enhances the fine-tuned Panfusion generator with a three-stage panoramic enhancement, enabling the generation of high-resolution, detail-rich panoramic images. During the 3D scene construction, a novel point cloud fusion initialization method is used, producing higher quality and spatially consistent point clouds. Our extensive experiments demonstrate that compared to other methods, SceneDreamer360 with its panoramic image generation and 3DGS can produce higher quality, spatially consistent, and visually appealing 3D scenes from any text prompt. Our codes are available at \url{https://github.com/liwrui/SceneDreamer360}.
♻ ☆ Event3DGS: Event-Based 3D Gaussian Splatting for High-Speed Robot Egomotion
By combining differentiable rendering with explicit point-based scene representations, 3D Gaussian Splatting (3DGS) has demonstrated breakthrough 3D reconstruction capabilities. However, to date 3DGS has had limited impact on robotics, where high-speed egomotion is pervasive: Egomotion introduces motion blur and leads to artifacts in existing frame-based 3DGS reconstruction methods. To address this challenge, we introduce Event3DGS, an {\em event-based} 3DGS framework. By exploiting the exceptional temporal resolution of event cameras, Event3GDS can reconstruct high-fidelity 3D structure and appearance under high-speed egomotion. Extensive experiments on multiple synthetic and real-world datasets demonstrate the superiority of Event3DGS compared with existing event-based dense 3D scene reconstruction frameworks; Event3DGS substantially improves reconstruction quality (+3dB) while reducing computational costs by 95\%. Our framework also allows one to incorporate a few motion-blurred frame-based measurements into the reconstruction process to further improve appearance fidelity without loss of structural accuracy.
comment: In the 8th Annual Conference on Robot Learning (CoRL 2024)
♻ ☆ Finetuning YOLOv9 for Vehicle Detection: Deep Learning for Intelligent Transportation Systems in Dhaka, Bangladesh
Rapid urbanization in megacities around the world, like Dhaka, has caused numerous transportation challenges that need to be addressed. Emerging technologies of deep learning and artificial intelligence can help us solve these problems to move towards Intelligent Transportation Systems (ITS) in the city. The government of Bangladesh recognizes the integration of ITS to ensure smart mobility as a vital step towards the development plan "Smart Bangladesh Vision 2041", but faces challenges in understanding ITS, its effects, and directions to implement. A vehicle detection system can pave the way to understanding traffic congestion, finding mobility patterns, and ensuring traffic surveillance. So, this paper proposes a fine-tuned object detector, the YOLOv9 model to detect native vehicles trained on a Bangladesh-based dataset. Results show that the fine-tuned YOLOv9 model achieved a mean Average Precision (mAP) of 0.934 at the Intersection over Union (IoU) threshold of 0.5, achieving state-of-the-art performance over past studies on Bangladesh-based datasets, shown through a comparison. Later, by suggesting the model to be deployed on CCTVs (closed circuit television) on the roads, a conceptual technique is proposed to process the vehicle detection model output data in a graph structure creating a vehicle detection system in the city. Finally, applications of such vehicle detection system are discussed showing a framework on how it can solve further ITS research questions, to provide a rationale for policymakers to implement the proposed vehicle detection system in the city.
comment: 16 pages, 10 figures
♻ ☆ SynCo: Synthetic Hard Negatives in Contrastive Learning for Better Unsupervised Visual Representations
Contrastive learning has become a dominant approach in self-supervised visual representation learning. Hard negatives - samples closely resembling the anchor - are key to enhancing learned representations' discriminative power. However, efficiently leveraging hard negatives remains challenging. We introduce SynCo (sYnthetic Negatives in Contrastive learning), a novel approach that improves model performance by generating synthetic hard negatives on the representation space. Building on the MoCo framework, SynCo introduces six strategies for creating diverse synthetic hard negatives on-the-fly with minimal computational overhead. SynCo achieves faster training and better representation learning, reaching 67.9% top-1 accuracy on ImageNet ILSVRC-201 linear evaluation after 200 pretraining epochs, surpassing MoCo's 67.5% using the same ResNet-50 encoder. It also transfers more effectively to detection tasks: on PASCAL VOC, it outperforms both the supervised baseline and MoCo with 82.6% AP; on COCO, it sets new benchmarks with 41.0% AP for bounding box detection and 35.7% AP for instance segmentation. Our synthetic hard negative generation approach significantly enhances visual representations learned through self-supervised contrastive learning. Code is available at https://github.com/giakoumoglou/synco.
comment: 10 pages, 6 figures, 4 tables
♻ ☆ Adapting Large Multimodal Models to Distribution Shifts: The Role of In-Context Learning
Recent studies indicate that large multimodal models (LMMs) potentially act as general-purpose assistants and are highly robust against different distributions. Despite this, domain-specific adaptation is still necessary particularly in specialized areas like healthcare. Due to the impracticality of fine-tuning LMMs given their vast parameter space, this work investigates in-context learning (ICL) as an effective alternative for enhancing LMMs' adaptability. Our study addresses this by evaluating an unsupervised ICL method which selects in-context examples through a nearest example search based on feature similarity. We uncover that its effectiveness is limited by the deficiencies of pre-trained vision encoders under distribution shift scenarios. To address these challenges, we propose InvariantSelectPR, a novel method leveraging Class-conditioned Contrastive Invariance (CCI) for more robust demonstration selection. Specifically, CCI enhances pre-trained vision encoders by improving their discriminative capabilities across different classes and ensuring invariance to domain-specific variations. This enhancement allows the encoders to effectively identify and retrieve the most informative examples, which are then used to guide LMMs in adapting to new query samples under varying distributions. Our experiments show that InvariantSelectPR substantially improves the adaptability of LMMs, achieving significant performance gains on benchmark datasets, with a 34.2%$\uparrow$ accuracy increase in 7-shot on Camelyon17 and 16.9%$\uparrow$ increase in 7-shot on HAM10000 compared to the baseline zero-shot performance.
comment: 10 pages, 9 figures, 7 tables
♻ ☆ On Feature Decorrelation in Cloth-Changing Person Re-identification
Cloth-changing person re-identification (CC-ReID) poses a significant challenge in computer vision. A prevailing approach is to prompt models to concentrate on causal attributes, like facial features and hairstyles, rather than confounding elements such as clothing appearance. Traditional methods to achieve this involve integrating multi-modality data or employing manually annotated clothing labels, which tend to complicate the model and require extensive human effort. In our study, we demonstrate that simply reducing feature correlations during training can significantly enhance the baseline model's performance. We theoretically elucidate this effect and introduce a novel regularization technique based on density ratio estimation. This technique aims to minimize feature correlation in the training process of cloth-changing ReID baselines. Our approach is model-independent, offering broad enhancements without needing additional data or labels. We validate our method through comprehensive experiments on prevalent CC-ReID datasets, showing its effectiveness in improving baseline models' generalization capabilities.
♻ ☆ Dragonfly: Multi-Resolution Zoom-In Encoding Enhances Vision-Language Models
Recent advances in vision-language models (VLMs) have demonstrated the advantages of processing images at higher resolutions and utilizing multi-crop features to preserve native resolution details. However, despite these improvements, existing vision transformers (ViTs) still struggle to capture fine-grained details from less prominent objects, charts, and embedded text, limiting their effectiveness in certain tasks. In this paper, we extend recent high-resolution and multi-crop techniques by not only preserving the native resolution, but zooming in beyond it and extracting features from a large number of image sub-crops. This enhancement allows our model to better capture fine-grained details, overcoming the limitations of current ViTs. To manage the increased token count and computational complexity, we demonstrate that a simple mean-pooling aggregation over tokens is effective. Our model, Dragonfly, achieves competitive performance on general-domain tasks such as ScienceQA and AI2D, and excels in tasks requiring fine-grained image understanding, including TextVQA and ChartQA. Among models in the 7-8B parameter range, Dragonfly consistently ranks at the top across ten general-domain benchmarks, achieving the highest or second-highest scores in most cases, outperforming models that are significantly larger or trained on larger datasets. Our biomedical model, Dragonfly-Med, sets new benchmarks on several medical tasks, achieving 91.6% accuracy on SLAKE (compared to 84.8% for Med-Gemini), a 67.1% token F1 score on Path-VQA (compared to 62.7% for Med-PaLM M), and state-of-the-art results across the majority of image captioning tasks. Overall, our work highlights the persistent challenge of engineering visual representations with fixed-resolution ViTs, and proposes a simple yet effective solution to address this issue and boost performance in both general and specialized domains.
♻ ☆ Weakly Supervised Learning for Facial Behavior Analysis : A Review
In the recent years, there has been a shift in facial behavior analysis from the laboratory-controlled conditions to the challenging in-the-wild conditions due to the superior performance of deep learning based approaches for many real world applications.However, the performance of deep learning approaches relies on the amount of training data. One of the major problems with data acquisition is the requirement of annotations for large amount of training data. Labeling process of huge training data demands lot of human support with strong domain expertise for facial expressions or action units, which is difficult to obtain in real-time environments.Moreover, labeling process is highly vulnerable to ambiguity of expressions or action units, especially for intensities due to the bias induced by the domain experts. Therefore, there is an imperative need to address the problem of facial behavior analysis with weak annotations. In this paper, we provide a comprehensive review of weakly supervised learning (WSL) approaches for facial behavior analysis with both categorical as well as dimensional labels along with the challenges and potential research directions associated with it. First, we introduce various types of weak annotations in the context of facial behavior analysis and the corresponding challenges associated with it. We then systematically review the existing state-of-the-art approaches and provide a taxonomy of these approaches along with their insights and limitations. In addition, widely used data-sets in the reviewed literature and the performance of these approaches along with evaluation principles are summarized. Finally, we discuss the remaining challenges and opportunities along with the potential research directions in order to apply facial behavior analysis with weak labels in real life situations.
comment: Provided a link of constantly updated papers \url{https://github.com/praveena2j/ awesome-Weakly-Supervised-Facial-Behavior-Analysis}
♻ ☆ Words in Motion: Extracting Interpretable Control Vectors for Motion Transformers NeurIPS'24
Transformer-based models generate hidden states that are difficult to interpret. In this work, we aim to interpret these hidden states and control them at inference, with a focus on motion forecasting. We leverage the phenomenon of neural collapse and use linear probes to measure interpretable features in hidden states. Our experiments reveal meaningful directions and distances between hidden states of opposing features, which we use to fit control vectors for activation steering. Consequently, our method enables controlling transformer-based motion forecasting models with interpretable features, providing a unique interface to interact with and understand these models. Our implementation is available at https://github.com/kit-mrt/future-motion
comment: NeurIPS'24 workshop camera-ready
♻ ☆ Dense Video Object Captioning from Disjoint Supervision
We propose a new task and model for dense video object captioning -- detecting, tracking and captioning trajectories of objects in a video. This task unifies spatial and temporal localization in video, whilst also requiring fine-grained visual understanding that is best described by natural language. We propose a unified model, and demonstrate how our end-to-end approach is more accurate and temporally coherent than a multi-stage pipeline combining state-of-the-art detection, tracking, and captioning models. Moreover, we propose a training strategy based on a mixture of disjoint tasks, which allows us to leverage diverse, large-scale datasets which supervise different parts of our model. Although each pretraining task only provides weak supervision, they are complementary and, when combined, result in noteworthy zero-shot ability and serve as strong initialization for additional finetuning to further improve accuracy. We carefully design new metrics capturing all components of our task, and show how we can repurpose existing video grounding datasets (e.g. VidSTG and VLN) for our new task. We show that our model improves upon a number of strong baselines for this new task. Furthermore, we can apply our model to the task of spatial grounding, outperforming prior state-of-the-art on VidSTG and VLN, without explicitly training for it. Code is available at https://github.com/google-research/scenic/tree/main/scenic/projects/densevoc.
comment: Code is available at https://github.com/google-research/scenic/tree/main/scenic/projects/densevoc
♻ ☆ CYCLO: Cyclic Graph Transformer Approach to Multi-Object Relationship Modeling in Aerial Videos NeurIPS 2024
Video scene graph generation (VidSGG) has emerged as a transformative approach to capturing and interpreting the intricate relationships among objects and their temporal dynamics in video sequences. In this paper, we introduce the new AeroEye dataset that focuses on multi-object relationship modeling in aerial videos. Our AeroEye dataset features various drone scenes and includes a visually comprehensive and precise collection of predicates that capture the intricate relationships and spatial arrangements among objects. To this end, we propose the novel Cyclic Graph Transformer (CYCLO) approach that allows the model to capture both direct and long-range temporal dependencies by continuously updating the history of interactions in a circular manner. The proposed approach also allows one to handle sequences with inherent cyclical patterns and process object relationships in the correct sequential order. Therefore, it can effectively capture periodic and overlapping relationships while minimizing information loss. The extensive experiments on the AeroEye dataset demonstrate the effectiveness of the proposed CYCLO model, demonstrating its potential to perform scene understanding on drone videos. Finally, the CYCLO method consistently achieves State-of-the-Art (SOTA) results on two in-the-wild scene graph generation benchmarks, i.e., PVSG and ASPIRe.
comment: Accepted to NeurIPS 2024
♻ ☆ Organizing Unstructured Image Collections using Natural Language
Organizing unstructured visual data into semantic clusters is a key challenge in computer vision. Traditional deep clustering (DC) approaches focus on a single partition of data, while multiple clustering (MC) methods address this limitation by uncovering distinct clustering solutions. The rise of large language models (LLMs) and multimodal LLMs (MLLMs) has enhanced MC by allowing users to define clustering criteria in natural language. However, manually specifying criteria for large datasets is impractical. In this work, we introduce the task Semantic Multiple Clustering (SMC) that aims to automatically discover clustering criteria from large image collections, uncovering interpretable substructures without requiring human input. Our framework, Text Driven Semantic Multiple Clustering (TeDeSC), uses text as a proxy to concurrently reason over large image collections, discover partitioning criteria, expressed in natural language, and reveal semantic substructures. To evaluate TeDeSC, we introduce the COCO-4c and Food-4c benchmarks, each containing four grouping criteria and ground-truth annotations. We apply TeDeSC to various applications, such as discovering biases and analyzing social media image popularity, demonstrating its utility as a tool for automatically organizing image collections and revealing novel insights.
comment: Preprint. Project webpage: https://oatmealliu.github.io/smc.html
Information Retrieval 23
☆ Generating Model Parameters for Controlling: Parameter Diffusion for Controllable Multi-Task Recommendation
Commercial recommender systems face the challenge that task requirements from platforms or users often change dynamically (e.g., varying preferences for accuracy or diversity). Ideally, the model should be re-trained after resetting a new objective function, adapting to these changes in task requirements. However, in practice, the high computational costs associated with retraining make this process impractical for models already deployed to online environments. This raises a new challenging problem: how to efficiently adapt the learning model to different task requirements by controlling model parameters after deployment, without the need for retraining. To address this issue, we propose a novel controllable learning approach via Parameter Diffusion for controllable multi-task Recommendation (PaDiRec), which allows the customization and adaptation of recommendation model parameters to new task requirements without retraining. Specifically, we first obtain the optimized model parameters through adapter tunning based on the feasible task requirements. Then, we utilize the diffusion model as a parameter generator, employing classifier-free guidance in conditional training to learn the distribution of optimized model parameters under various task requirements. Finally, the diffusion model is applied to effectively generate model parameters in a test-time adaptation manner given task requirements. As a model-agnostic approach, PaDiRec can leverage existing recommendation models as backbones to enhance their controllability. Extensive experiments on public datasets and a dataset from a commercial app, indicate that PaDiRec can effectively enhance controllability through efficient model parameter generation. The code is released at https://anonymous.4open.science/r/PaDiRec-DD13.
☆ VisRAG: Vision-based Retrieval-augmented Generation on Multi-modality Documents
Retrieval-augmented generation (RAG) is an effective technique that enables large language models (LLMs) to utilize external knowledge sources for generation. However, current RAG systems are solely based on text, rendering it impossible to utilize vision information like layout and images that play crucial roles in real-world multi-modality documents. In this paper, we introduce VisRAG, which tackles this issue by establishing a vision-language model (VLM)-based RAG pipeline. In this pipeline, instead of first parsing the document to obtain text, the document is directly embedded using a VLM as an image and then retrieved to enhance the generation of a VLM. Compared to traditional text-based RAG, VisRAG maximizes the retention and utilization of the data information in the original documents, eliminating the information loss introduced during the parsing process. We collect both open-source and synthetic data to train the retriever in VisRAG and explore a variety of generation methods. Experiments demonstrate that VisRAG outperforms traditional RAG in both the retrieval and generation stages, achieving a 25--39\% end-to-end performance gain over traditional text-based RAG pipeline. Further analysis reveals that VisRAG is effective in utilizing training data and demonstrates strong generalization capability, positioning it as a promising solution for RAG on multi-modality documents. Our code and data are available at https://github.com/openbmb/visrag .
☆ Rethinking Legal Judgement Prediction in a Realistic Scenario in the Era of Large Language Models EMNLP 2024
This study investigates judgment prediction in a realistic scenario within the context of Indian judgments, utilizing a range of transformer-based models, including InLegalBERT, BERT, and XLNet, alongside LLMs such as Llama-2 and GPT-3.5 Turbo. In this realistic scenario, we simulate how judgments are predicted at the point when a case is presented for a decision in court, using only the information available at that time, such as the facts of the case, statutes, precedents, and arguments. This approach mimics real-world conditions, where decisions must be made without the benefit of hindsight, unlike retrospective analyses often found in previous studies. For transformer models, we experiment with hierarchical transformers and the summarization of judgment facts to optimize input for these models. Our experiments with LLMs reveal that GPT-3.5 Turbo excels in realistic scenarios, demonstrating robust performance in judgment prediction. Furthermore, incorporating additional legal information, such as statutes and precedents, significantly improves the outcome of the prediction task. The LLMs also provide explanations for their predictions. To evaluate the quality of these predictions and explanations, we introduce two human evaluation metrics: Clarity and Linking. Our findings from both automatic and human evaluations indicate that, despite advancements in LLMs, they are yet to achieve expert-level performance in judgment prediction and explanation tasks.
comment: Accepted on NLLP at EMNLP 2024
☆ Advancing Academic Knowledge Retrieval via LLM-enhanced Representation Similarity Fusion KDD
In an era marked by robust technological growth and swift information renewal, furnishing researchers and the populace with top-tier, avant-garde academic insights spanning various domains has become an urgent necessity. The KDD Cup 2024 AQA Challenge is geared towards advancing retrieval models to identify pertinent academic terminologies from suitable papers for scientific inquiries. This paper introduces the LLM-KnowSimFuser proposed by Robo Space, which wins the 2nd place in the competition. With inspirations drawed from the superior performance of LLMs on multiple tasks, after careful analysis of the provided datasets, we firstly perform fine-tuning and inference using LLM-enhanced pre-trained retrieval models to introduce the tremendous language understanding and open-domain knowledge of LLMs into this task, followed by a weighted fusion based on the similarity matrix derived from the inference results. Finally, experiments conducted on the competition datasets show the superiority of our proposal, which achieved a score of 0.20726 on the final leaderboard.
comment: The 2nd Place of KDD Cup 2024 OAG-Challenge AQA
☆ Medico: Towards Hallucination Detection and Correction with Multi-source Evidence Fusion EMNLP 2024
As we all know, hallucinations prevail in Large Language Models (LLMs), where the generated content is coherent but factually incorrect, which inflicts a heavy blow on the widespread application of LLMs. Previous studies have shown that LLMs could confidently state non-existent facts rather than answering ``I don't know''. Therefore, it is necessary to resort to external knowledge to detect and correct the hallucinated content. Since manual detection and correction of factual errors is labor-intensive, developing an automatic end-to-end hallucination-checking approach is indeed a needful thing. To this end, we present Medico, a Multi-source evidence fusion enhanced hallucination detection and correction framework. It fuses diverse evidence from multiple sources, detects whether the generated content contains factual errors, provides the rationale behind the judgment, and iteratively revises the hallucinated content. Experimental results on evidence retrieval (0.964 HR@5, 0.908 MRR@5), hallucination detection (0.927-0.951 F1), and hallucination correction (0.973-0.979 approval rate) manifest the great potential of Medico. A video demo of Medico can be found at https://youtu.be/RtsO6CSesBI.
comment: 12 pages, 3 figures, 6 tables. Accepted by EMNLP 2024's demo track
☆ Collaborative filtering based on nonnegative/binary matrix factorization
Collaborative filtering generates recommendations based on user-item similarities through rating data, which may involve numerous unrated items. To predict scores for unrated items, matrix factorization techniques, such as nonnegative matrix factorization (NMF), are often employed to predict scores for unrated items. Nonnegative/binary matrix factorization (NBMF), which is an extension of NMF, approximates a nonnegative matrix as the product of nonnegative and binary matrices. Previous studies have employed NBMF for image analysis where the data were dense. In this paper, we propose a modified NBMF algorithm that can be applied to collaborative filtering where data are sparse. In the modified method, unrated elements in a rating matrix are masked, which improves the collaborative filtering performance. Utilizing a low-latency Ising machine in NBMF is advantageous in terms of the computation time, making the proposed method beneficial.
comment: 12 pages, 7 figures
☆ BookWorm: A Dataset for Character Description and Analysis EMNLP 2024
Characters are at the heart of every story, driving the plot and engaging readers. In this study, we explore the understanding of characters in full-length books, which contain complex narratives and numerous interacting characters. We define two tasks: character description, which generates a brief factual profile, and character analysis, which offers an in-depth interpretation, including character development, personality, and social context. We introduce the BookWorm dataset, pairing books from the Gutenberg Project with human-written descriptions and analyses. Using this dataset, we evaluate state-of-the-art long-context models in zero-shot and fine-tuning settings, utilizing both retrieval-based and hierarchical processing for book-length inputs. Our findings show that retrieval-based approaches outperform hierarchical ones in both tasks. Additionally, fine-tuned models using coreference-based retrieval produce the most factual descriptions, as measured by fact- and entailment-based metrics. We hope our dataset, experiments, and analysis will inspire further research in character-based narrative understanding.
comment: 30 pages, 2 figures, EMNLP 2024 Findings
☆ A Hybrid Filtering for Micro-video Hashtag Recommendation using Graph-based Deep Neural Network
Due to the growing volume of user generated content, hashtags are employed as topic indicators to manage content efficiently on social media platforms. However, finding these vital topics is challenging in microvideos since they contain substantial information in a short duration. Existing methods that recommend hashtags for microvideos primarily focus on content and personalization while disregarding relatedness among users. Moreover, the cold start user issue prevails in hashtag recommendation systems. Considering the above, we propose a hybrid filtering based MIcro-video haSHtag recommendatiON MISHON technique to recommend hashtags for micro-videos. Besides content based filtering, we employ user-based collaborative filtering to enhance recommendations. Since hashtags reflect users topical interests, we find similar users based on historical tagging behavior to model user relatedness. We employ a graph-based deep neural network to model user to user, modality to modality, and user to modality interactions. We then use refined modality specific and user representations to recommend pertinent hashtags for microvideos. The empirical results on three real world datasets demonstrate that MISHON attains a comparative enhancement of 3.6, 2.8, and 6.5 reported in percentage concerning the F1 score, respectively. Since cold start users exist whose historical tagging information is unavailable, we also propose a content and social influence based technique to model the relatedness of cold start users with influential users. The proposed solution shows a relative improvement of 15.8 percent in the F1 score over its content only counterpart. These results show that the proposed framework mitigates the cold start user problem.
☆ Parenting: Optimizing Knowledge Selection of Retrieval-Augmented Language Models with Parameter Decoupling and Tailored Tuning
Retrieval-Augmented Generation (RAG) offers an effective solution to the issues faced by Large Language Models (LLMs) in hallucination generation and knowledge obsolescence by incorporating externally retrieved knowledge. However, due to potential conflicts between internal and external knowledge, as well as retrieval noise, LLMs often struggle to effectively integrate external evidence, leading to a decline in performance. Although existing methods attempt to tackle these challenges, they often struggle to strike a balance between model adherence and robustness, resulting in significant learning variance. Inspired by human cognitive processes, we propose Parenting, a novel framework that decouples adherence and robustness within the parameter space of LLMs. Specifically, Parenting utilizes a key parameter mining method based on forward activation gain to identify and isolate the crucial parameter units that are strongly linked to adherence and robustness. Then, Parenting employs a type-guided tailored tuning strategy, applying specific and appropriate fine-tuning methods to parameter units representing different capabilities, aiming to achieve a balanced enhancement of adherence and robustness. Extensive experiments on various datasets and models validate the effectiveness and generalizability of our methods.
☆ Enhancing Attributed Graph Networks with Alignment and Uniformity Constraints for Session-based Recommendation
Session-based Recommendation (SBR), seeking to predict a user's next action based on an anonymous session, has drawn increasing attention for its practicability. Most SBR models only rely on the contextual transitions within a short session to learn item representations while neglecting additional valuable knowledge. As such, their model capacity is largely limited by the data sparsity issue caused by short sessions. A few studies have exploited the Modeling of Item Attributes (MIA) to enrich item representations. However, they usually involve specific model designs that can hardly transfer to existing attribute-agnostic SBR models and thus lack universality. In this paper, we propose a model-agnostic framework, named AttrGAU (Attributed Graph Networks with Alignment and Uniformity Constraints), to bring the MIA's superiority into existing attribute-agnostic models, to improve their accuracy and robustness for recommendation. Specifically, we first build a bipartite attributed graph and design an attribute-aware graph convolution to exploit the rich attribute semantics hidden in the heterogeneous item-attribute relationship. We then decouple existing attribute-agnostic SBR models into the graph neural network and attention readout sub-modules to satisfy the non-intrusive requirement. Lastly, we design two representation constraints, i.e., alignment and uniformity, to optimize distribution discrepancy in representation between the attribute semantics and collaborative semantics. Extensive experiments on three public benchmark datasets demonstrate that the proposed AttrGAU framework can significantly enhance backbone models' recommendation performance and robustness against data sparsity and data noise issues. Our implementation codes will be available at https://github.com/ItsukiFujii/AttrGAU.
comment: 11 pages, 4 figures, 5 tables. Accepted by ICWS 2024
☆ FunnelRAG: A Coarse-to-Fine Progressive Retrieval Paradigm for RAG
Retrieval-Augmented Generation (RAG) prevails in Large Language Models. It mainly consists of retrieval and generation. The retrieval modules (a.k.a. retrievers) aim to find useful information used to facilitate generation modules (a.k.a. generators). As such, generators' performance largely depends on the effectiveness and efficiency of retrievers. However, the retrieval paradigm that we design and use remains flat, which treats the retrieval procedures as a one-off deal with constant granularity. Despite effectiveness, we argue that they suffer from two limitations: (1) flat retrieval exerts a significant burden on one retriever; (2) constant granularity limits the ceiling of retrieval performance. In this work, we propose a progressive retrieval paradigm with coarse-to-fine granularity for RAG, termed FunnelRAG, so as to balance effectiveness and efficiency. Specifically, FunnelRAG establishes a progressive retrieval pipeline by collaborating coarse-to-fine granularity, large-to-small quantity, and low-to-high capacity, which can relieve the burden on one retriever and also promote the ceiling of retrieval performance. Extensive experiments manifest that FunnelRAG achieves comparable retrieval performance while the time overhead is reduced by nearly 40 percent.
comment: 18 pages, 6 figures, 13 tables
☆ Back-of-the-Book Index Automation for Arabic Documents
Back-of-the-book indexes are crucial for book readability. Their manual creation is laborious and error prone. In this paper, we consider automating back-of-the-book index extraction for Arabic books to help simplify both the creation and review tasks. Given a back-of-the-book index, we aim to check and identify the accurate occurrences of index terms relative to the associated pages. To achieve this, we first define a pool of candidates for each term by extracting all possible noun phrases from paragraphs appearing on the relevant index pages. These noun phrases, identified through part-of-speech analysis, are stored in a vector database for efficient retrieval. We use several metrics, including exact matches, lexical similarity, and semantic similarity, to determine the most appropriate occurrence. The candidate with the highest score based on these metrics is chosen as the occurrence of the term. We fine-tuned a heuristic method, that considers the above metrics and that achieves an F1-score of .966 (precision=.966, recall=.966). These excellent results open the door for future work related to automation of back-of-the-book index generation and checking.
☆ DecKG: Decentralized Collaborative Learning with Knowledge Graph Enhancement for POI Recommendation
Decentralized collaborative learning for Point-of-Interest (POI) recommendation has gained research interest due to its advantages in privacy preservation and efficiency, as it keeps data locally and leverages collaborative learning among clients to train models in a decentralized manner. However, since local data is often limited and insufficient for training accurate models, a common solution is integrating external knowledge as auxiliary information to enhance model performance. Nevertheless, this solution poses challenges for decentralized collaborative learning. Due to private nature of local data, identifying relevant auxiliary information specific to each user is non-trivial. Furthermore, resource-constrained local devices struggle to accommodate all auxiliary information, which places heavy burden on local storage. To fill the gap, we propose a novel decentralized collaborative learning with knowledge graph enhancement framework for POI recommendation (DecKG). Instead of directly uploading interacted items, users generate desensitized check-in data by uploading general categories of interacted items and sampling similar items from same category. The server then pretrains KG without sensitive user-item interactions and deploys relevant partitioned sub-KGs to individual users. Entities are further refined on the device, allowing client to client communication to exchange knowledge learned from local data and sub-KGs. Evaluations across two real-world datasets demonstrate DecKG's effectiveness recommendation performance.
☆ MAIR: A Massive Benchmark for Evaluating Instructed Retrieval EMNLP 2024
Recent information retrieval (IR) models are pre-trained and instruction-tuned on massive datasets and tasks, enabling them to perform well on a wide range of tasks and potentially generalize to unseen tasks with instructions. However, existing IR benchmarks focus on a limited scope of tasks, making them insufficient for evaluating the latest IR models. In this paper, we propose MAIR (Massive Instructed Retrieval Benchmark), a heterogeneous IR benchmark that includes 126 distinct IR tasks across 6 domains, collected from existing datasets. We benchmark state-of-the-art instruction-tuned text embedding models and re-ranking models. Our experiments reveal that instruction-tuned models generally achieve superior performance compared to non-instruction-tuned models on MAIR. Additionally, our results suggest that current instruction-tuned text embedding models and re-ranking models still lack effectiveness in specific long-tail tasks. MAIR is publicly available at https://github.com/sunnweiwei/Mair.
comment: EMNLP 2024
☆ SGUQ: Staged Graph Convolution Neural Network for Alzheimer's Disease Diagnosis using Multi-Omics Data
Alzheimer's disease (AD) is a chronic neurodegenerative disorder and the leading cause of dementia, significantly impacting cost, mortality, and burden worldwide. The advent of high-throughput omics technologies, such as genomics, transcriptomics, proteomics, and epigenomics, has revolutionized the molecular understanding of AD. Conventional AI approaches typically require the completion of all omics data at the outset to achieve optimal AD diagnosis, which are inefficient and may be unnecessary. To reduce the clinical cost and improve the accuracy of AD diagnosis using multi-omics data, we propose a novel staged graph convolutional network with uncertainty quantification (SGUQ). SGUQ begins with mRNA and progressively incorporates DNA methylation and miRNA data only when necessary, reducing overall costs and exposure to harmful tests. Experimental results indicate that 46.23% of the samples can be reliably predicted using only single-modal omics data (mRNA), while an additional 16.04% of the samples can achieve reliable predictions when combining two omics data types (mRNA + DNA methylation). In addition, the proposed staged SGUQ achieved an accuracy of 0.858 on ROSMAP dataset, which outperformed existing methods significantly. The proposed SGUQ can not only be applied to AD diagnosis using multi-omics data but also has the potential for clinical decision-making using multi-viewed data. Our implementation is publicly available at https://github.com/chenzhao2023/multiomicsuncertainty.
comment: 20 pages, 2 figures
☆ GraFPrint: A GNN-Based Approach for Audio Identification ICASSP 2025
This paper introduces GraFPrint, an audio identification framework that leverages the structural learning capabilities of Graph Neural Networks (GNNs) to create robust audio fingerprints. Our method constructs a k-nearest neighbor (k-NN) graph from time-frequency representations and applies max-relative graph convolutions to encode local and global information. The network is trained using a self-supervised contrastive approach, which enhances resilience to ambient distortions by optimizing feature representation. GraFPrint demonstrates superior performance on large-scale datasets at various levels of granularity, proving to be both lightweight and scalable, making it suitable for real-world applications with extensive reference databases.
comment: Submitted to IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2025)
♻ ☆ Pure Message Passing Can Estimate Common Neighbor for Link Prediction
Message Passing Neural Networks (MPNNs) have emerged as the {\em de facto} standard in graph representation learning. However, when it comes to link prediction, they often struggle, surpassed by simple heuristics such as Common Neighbor (CN). This discrepancy stems from a fundamental limitation: while MPNNs excel in node-level representation, they stumble with encoding the joint structural features essential to link prediction, like CN. To bridge this gap, we posit that, by harnessing the orthogonality of input vectors, pure message-passing can indeed capture joint structural features. Specifically, we study the proficiency of MPNNs in approximating CN heuristics. Based on our findings, we introduce the Message Passing Link Predictor (MPLP), a novel link prediction model. MPLP taps into quasi-orthogonal vectors to estimate link-level structural features, all while preserving the node-level complexities. Moreover, our approach demonstrates that leveraging message-passing to capture structural features could offset MPNNs' expressiveness limitations at the expense of estimation variance. We conduct experiments on benchmark datasets from various domains, where our method consistently outperforms the baseline methods.
comment: Accepted to Neurips'24
♻ ☆ mGTE: Generalized Long-Context Text Representation and Reranking Models for Multilingual Text Retrieval EMNLP 2024
We present systematic efforts in building long-context multilingual text representation model (TRM) and reranker from scratch for text retrieval. We first introduce a text encoder (base size) enhanced with RoPE and unpadding, pre-trained in a native 8192-token context (longer than 512 of previous multilingual encoders). Then we construct a hybrid TRM and a cross-encoder reranker by contrastive learning. Evaluations show that our text encoder outperforms the same-sized previous state-of-the-art XLM-R. Meanwhile, our TRM and reranker match the performance of large-sized state-of-the-art BGE-M3 models and achieve better results on long-context retrieval benchmarks. Further analysis demonstrate that our proposed models exhibit higher efficiency during both training and inference. We believe their efficiency and effectiveness could benefit various researches and industrial applications.
comment: Camera-ready version of EMNLP 2024: Industry Track
♻ ☆ Personalized Item Representations in Federated Multimodal Recommendation
Federated recommendation systems are essential for providing personalized recommendations while protecting user privacy. However, current methods mainly rely on ID-based item embeddings, neglecting the rich multimodal information of items. To address this, we propose a Federated Multimodal Recommendation System, called FedMR. FedMR uses a foundation model on the server to encode multimodal item data, such as images and text. To handle data heterogeneity caused by user preference differences, FedMR introduces a Mixing Feature Fusion Module on each client, which adjusts fusion strategy weights based on user interaction history to generate personalized item representations that capture users' fine-grained preferences. FedMR is compatible with existing ID-based federated recommendation systems, improving performance without modifying the original framework. Experiments on four real-world multimodal datasets demonstrate FedMR's effectiveness. The code is available at https://anonymous.4open.science/r/FedMR.
comment: 12 pages, 4 figures, 5 tables, conference
♻ ☆ LARA: Linguistic-Adaptive Retrieval-Augmentation for Multi-Turn Intent Classification EMNLP'24
Multi-turn intent classification is notably challenging due to the complexity and evolving nature of conversational contexts. This paper introduces LARA, a Linguistic-Adaptive Retrieval-Augmentation framework to enhance accuracy in multi-turn classification tasks across six languages, accommodating a large number of intents in chatbot interactions. LARA combines a fine-tuned smaller model with a retrieval-augmented mechanism, integrated within the architecture of LLMs. The integration allows LARA to dynamically utilize past dialogues and relevant intents, thereby improving the understanding of the context. Furthermore, our adaptive retrieval techniques bolster the cross-lingual capabilities of LLMs without extensive retraining and fine-tuning. Comprehensive experiments demonstrate that LARA achieves state-of-the-art performance on multi-turn intent classification tasks, enhancing the average accuracy by 3.67\% from state-of-the-art single-turn intent classifiers.
comment: Accepted to EMNLP'24
♻ ☆ Language Model Powered Digital Biology
Recent advancements in Large Language Models (LLMs) are transforming biology, computer science, and many other research fields, as well as impacting everyday life. While transformer-based technologies are currently being deployed in biology, no available agentic system has been developed to tackle bioinformatics workflows. We present a prototype Bioinformatics Retrieval Augmented Data (BRAD) digital assistant. BRAD is a chatbot and agentic system that integrates a suite of tools to handle bioinformatics tasks, from code execution to online search. We demonstrate its capabilities through (1) improved question-and-answering with retrieval augmented generation (RAG), (2) the ability to run complex software pipelines, and (3) the ability to organize and distribute tasks in agentic workflows. We use BRAD for automation, performing tasks ranging from gene enrichment and searching the archive to automatic code generation for running biomarker identification pipelines. BRAD is a step toward autonomous, self-driving labs for digital biology.
comment: 49 pages, 3 tables, 12 figures
♻ ☆ Enhancing Dense Retrievers' Robustness with Group-level Reweighting
The anchor-document data derived from web graphs offers a wealth of paired information for training dense retrieval models in an unsupervised manner. However, unsupervised data contains diverse patterns across the web graph and often exhibits significant imbalance, leading to suboptimal performance in underrepresented or difficult groups. In this paper, we introduce WebDRO, an efficient approach for clustering the web graph data and optimizing group weights to enhance the robustness of dense retrieval models. Initially, we build an embedding model for clustering anchor-document pairs. Specifically, we contrastively train the embedding model for link prediction, which guides the embedding model in capturing the document features behind the web graph links. Subsequently, we employ the group distributional robust optimization to recalibrate the weights across different clusters of anchor-document pairs during training retrieval models. During training, we direct the model to assign higher weights to clusters with higher loss and focus more on worst-case scenarios. This approach ensures that the model has strong generalization ability on all data patterns. Our experiments on MS MARCO and BEIR demonstrate that our method can effectively improve retrieval performance in unsupervised training and finetuning settings. Further analysis confirms the stability and validity of group weights learned by WebDRO. The code of this paper can be obtained from https://github.com/Hanpx20/GroupDRO_Dense_Retrieval.
♻ ☆ On Feature Decorrelation in Cloth-Changing Person Re-identification
Cloth-changing person re-identification (CC-ReID) poses a significant challenge in computer vision. A prevailing approach is to prompt models to concentrate on causal attributes, like facial features and hairstyles, rather than confounding elements such as clothing appearance. Traditional methods to achieve this involve integrating multi-modality data or employing manually annotated clothing labels, which tend to complicate the model and require extensive human effort. In our study, we demonstrate that simply reducing feature correlations during training can significantly enhance the baseline model's performance. We theoretically elucidate this effect and introduce a novel regularization technique based on density ratio estimation. This technique aims to minimize feature correlation in the training process of cloth-changing ReID baselines. Our approach is model-independent, offering broad enhancements without needing additional data or labels. We validate our method through comprehensive experiments on prevalent CC-ReID datasets, showing its effectiveness in improving baseline models' generalization capabilities.
Machine Learning 200
☆ When Does Perceptual Alignment Benefit Vision Representations?
Humans judge perceptual similarity according to diverse visual attributes, including scene layout, subject location, and camera pose. Existing vision models understand a wide range of semantic abstractions but improperly weigh these attributes and thus make inferences misaligned with human perception. While vision representations have previously benefited from alignment in contexts like image generation, the utility of perceptually aligned representations in more general-purpose settings remains unclear. Here, we investigate how aligning vision model representations to human perceptual judgments impacts their usability across diverse computer vision tasks. We finetune state-of-the-art models on human similarity judgments for image triplets and evaluate them across standard vision benchmarks. We find that aligning models to perceptual judgments yields representations that improve upon the original backbones across many downstream tasks, including counting, segmentation, depth estimation, instance retrieval, and retrieval-augmented generation. In addition, we find that performance is widely preserved on other tasks, including specialized out-of-distribution domains such as in medical imaging and 3D environment frames. Our results suggest that injecting an inductive bias about human perceptual knowledge into vision models can contribute to better representations.
comment: S.S. and S.F. contributed equally. Website: percep-align.github.io
☆ TemporalBench: Benchmarking Fine-grained Temporal Understanding for Multimodal Video Models
Understanding fine-grained temporal dynamics is crucial for multimodal video comprehension and generation. Due to the lack of fine-grained temporal annotations, existing video benchmarks mostly resemble static image benchmarks and are incompetent at evaluating models for temporal understanding. In this paper, we introduce TemporalBench, a new benchmark dedicated to evaluating fine-grained temporal understanding in videos. TemporalBench consists of ~10K video question-answer pairs, derived from ~2K high-quality human annotations detailing the temporal dynamics in video clips. As a result, our benchmark provides a unique testbed for evaluating various temporal understanding and reasoning abilities such as action frequency, motion magnitude, event order, etc. Moreover, it enables evaluations on various tasks like both video question answering and captioning, both short and long video understanding, as well as different models such as multimodal video embedding models and text generation models. Results show that state-of-the-art models like GPT-4o achieve only 38.5% question answering accuracy on TemporalBench, demonstrating a significant gap (~30%) between humans and AI in temporal understanding. Furthermore, we notice a critical pitfall for multi-choice QA where LLMs can detect the subtle changes in negative captions and find a centralized description as a cue for its prediction, where we propose Multiple Binary Accuracy (MBA) to correct such bias. We hope that TemporalBench can foster research on improving models' temporal reasoning capabilities. Both dataset and evaluation code will be made available.
comment: Project Page: https://temporalbench.github.io/
☆ LVD-2M: A Long-take Video Dataset with Temporally Dense Captions NeurIPS 2024
The efficacy of video generation models heavily depends on the quality of their training datasets. Most previous video generation models are trained on short video clips, while recently there has been increasing interest in training long video generation models directly on longer videos. However, the lack of such high-quality long videos impedes the advancement of long video generation. To promote research in long video generation, we desire a new dataset with four key features essential for training long video generation models: (1) long videos covering at least 10 seconds, (2) long-take videos without cuts, (3) large motion and diverse contents, and (4) temporally dense captions. To achieve this, we introduce a new pipeline for selecting high-quality long-take videos and generating temporally dense captions. Specifically, we define a set of metrics to quantitatively assess video quality including scene cuts, dynamic degrees, and semantic-level quality, enabling us to filter high-quality long-take videos from a large amount of source videos. Subsequently, we develop a hierarchical video captioning pipeline to annotate long videos with temporally-dense captions. With this pipeline, we curate the first long-take video dataset, LVD-2M, comprising 2 million long-take videos, each covering more than 10 seconds and annotated with temporally dense captions. We further validate the effectiveness of LVD-2M by fine-tuning video generation models to generate long videos with dynamic motions. We believe our work will significantly contribute to future research in long video generation.
comment: NeurIPS 2024 Dataset and Benchmark Track. Project page: https://silentview.github.io/LVD-2M/ . Code: https://github.com/SilentView/LVD-2M
☆ Your Mixture-of-Experts LLM Is Secretly an Embedding Model For Free
While large language models (LLMs) excel on generation tasks, their decoder-only architecture often limits their potential as embedding models if no further representation finetuning is applied. Does this contradict their claim of generalists? To answer the question, we take a closer look at Mixture-of-Experts (MoE) LLMs. Our study shows that the expert routers in MoE LLMs can serve as an off-the-shelf embedding model with promising performance on a diverse class of embedding-focused tasks, without requiring any finetuning. Moreover, our extensive analysis shows that the MoE routing weights (RW) is complementary to the hidden state (HS) of LLMs, a widely-used embedding. Compared to HS, we find that RW is more robust to the choice of prompts and focuses on high-level semantics. Motivated by the analysis, we propose MoEE combining RW and HS, which achieves better performance than using either separately. Our exploration of their combination and prompting strategy shed several novel insights, e.g., a weighted sum of RW and HS similarities outperforms the similarity on their concatenation. Our experiments are conducted on 6 embedding tasks with 20 datasets from the Massive Text Embedding Benchmark (MTEB). The results demonstrate the significant improvement brought by MoEE to LLM-based embedding without further finetuning.
comment: 10 pages, 5 figures
☆ HART: Efficient Visual Generation with Hybrid Autoregressive Transformer
We introduce Hybrid Autoregressive Transformer (HART), an autoregressive (AR) visual generation model capable of directly generating 1024x1024 images, rivaling diffusion models in image generation quality. Existing AR models face limitations due to the poor image reconstruction quality of their discrete tokenizers and the prohibitive training costs associated with generating 1024px images. To address these challenges, we present the hybrid tokenizer, which decomposes the continuous latents from the autoencoder into two components: discrete tokens representing the big picture and continuous tokens representing the residual components that cannot be represented by the discrete tokens. The discrete component is modeled by a scalable-resolution discrete AR model, while the continuous component is learned with a lightweight residual diffusion module with only 37M parameters. Compared with the discrete-only VAR tokenizer, our hybrid approach improves reconstruction FID from 2.11 to 0.30 on MJHQ-30K, leading to a 31% generation FID improvement from 7.85 to 5.38. HART also outperforms state-of-the-art diffusion models in both FID and CLIP score, with 4.5-7.7x higher throughput and 6.9-13.4x lower MACs. Our code is open sourced at https://github.com/mit-han-lab/hart.
comment: Demo: https://hart.mit.edu. The first two authors contributed equally to this work
☆ Deep Linear Probe Generators for Weight Space Learning
Weight space learning aims to extract information about a neural network, such as its training dataset or generalization error. Recent approaches learn directly from model weights, but this presents many challenges as weights are high-dimensional and include permutation symmetries between neurons. An alternative approach, Probing, represents a model by passing a set of learned inputs (probes) through the model, and training a predictor on top of the corresponding outputs. Although probing is typically not used as a stand alone approach, our preliminary experiment found that a vanilla probing baseline worked surprisingly well. However, we discover that current probe learning strategies are ineffective. We therefore propose Deep Linear Probe Generators (ProbeGen), a simple and effective modification to probing approaches. ProbeGen adds a shared generator module with a deep linear architecture, providing an inductive bias towards structured probes thus reducing overfitting. While simple, ProbeGen performs significantly better than the state-of-the-art and is very efficient, requiring between 30 to 1000 times fewer FLOPs than other top approaches.
☆ Hard-Constrained Neural Networks with Universal Approximation Guarantees
Incorporating prior knowledge or specifications of input-output relationships into machine learning models has gained significant attention, as it enhances generalization from limited data and leads to conforming outputs. However, most existing approaches use soft constraints by penalizing violations through regularization, which offers no guarantee of constraint satisfaction -- an essential requirement in safety-critical applications. On the other hand, imposing hard constraints on neural networks may hinder their representational power, adversely affecting performance. To address this, we propose HardNet, a practical framework for constructing neural networks that inherently satisfy hard constraints without sacrificing model capacity. Specifically, we encode affine and convex hard constraints, dependent on both inputs and outputs, by appending a differentiable projection layer to the network's output. This architecture allows unconstrained optimization of the network parameters using standard algorithms while ensuring constraint satisfaction by construction. Furthermore, we show that HardNet retains the universal approximation capabilities of neural networks. We demonstrate the versatility and effectiveness of HardNet across various applications: fitting functions under constraints, learning optimization solvers, optimizing control policies in safety-critical systems, and learning safe decision logic for aircraft systems.
☆ TL-PCA: Transfer Learning of Principal Component Analysis
Principal component analysis (PCA) can be significantly limited when there is too few examples of the target data of interest. We propose a transfer learning approach to PCA (TL-PCA) where knowledge from a related source task is used in addition to the scarce data of a target task. Our TL-PCA has two versions, one that uses a pretrained PCA solution of the source task, and another that uses the source data. Our proposed approach extends the PCA optimization objective with a penalty on the proximity of the target subspace and the source subspace as given by the pretrained source model or the source data. This optimization is solved by eigendecomposition for which the number of data-dependent eigenvectors (i.e., principal directions of TL-PCA) is not limited to the number of target data examples, which is a root cause that limits the standard PCA performance. Accordingly, our results for image datasets show that the representation of test data is improved by TL-PCA for dimensionality reduction where the learned subspace dimension is lower or higher than the number of target data examples.
☆ TrajDiffuse: A Conditional Diffusion Model for Environment-Aware Trajectory Prediction ICPR
Accurate prediction of human or vehicle trajectories with good diversity that captures their stochastic nature is an essential task for many applications. However, many trajectory prediction models produce unreasonable trajectory samples that focus on improving diversity or accuracy while neglecting other key requirements, such as collision avoidance with the surrounding environment. In this work, we propose TrajDiffuse, a planning-based trajectory prediction method using a novel guided conditional diffusion model. We form the trajectory prediction problem as a denoising impaint task and design a map-based guidance term for the diffusion process. TrajDiffuse is able to generate trajectory predictions that match or exceed the accuracy and diversity of the SOTA, while adhering almost perfectly to environmental constraints. We demonstrate the utility of our model through experiments on the nuScenes and PFSD datasets and provide an extensive benchmark analysis against the SOTA methods.
comment: Accepted to be published as inpreceedings of the 2024 International Conference on Pattern Recognition (ICPR)
☆ Generalizable Humanoid Manipulation with Improved 3D Diffusion Policies
Humanoid robots capable of autonomous operation in diverse environments have long been a goal for roboticists. However, autonomous manipulation by humanoid robots has largely been restricted to one specific scene, primarily due to the difficulty of acquiring generalizable skills. Recent advances in 3D visuomotor policies, such as the 3D Diffusion Policy (DP3), have shown promise in extending these capabilities to wilder environments. However, 3D visuomotor policies often rely on camera calibration and point-cloud segmentation, which present challenges for deployment on mobile robots like humanoids. In this work, we introduce the Improved 3D Diffusion Policy (iDP3), a novel 3D visuomotor policy that eliminates these constraints by leveraging egocentric 3D visual representations. We demonstrate that iDP3 enables a full-sized humanoid robot to autonomously perform skills in diverse real-world scenarios, using only data collected in the lab. Videos are available at: https://humanoid-manipulation.github.io
comment: Project website: https://humanoid-manipulation.github.io
☆ Mix Data or Merge Models? Optimizing for Diverse Multi-Task Learning
Large Language Models (LLMs) have been adopted and deployed worldwide for a broad variety of applications. However, ensuring their safe use remains a significant challenge. Preference training and safety measures often overfit to harms prevalent in Western-centric datasets, and safety protocols frequently fail to extend to multilingual settings. In this work, we explore model merging in a diverse multi-task setting, combining safety and general-purpose tasks within a multilingual context. Each language introduces unique and varied learning challenges across tasks. We find that objective-based merging is more effective than mixing data, with improvements of up to 8% and 10% in general performance and safety respectively. We also find that language-based merging is highly effective -- by merging monolingually fine-tuned models, we achieve a 4% increase in general performance and 7% reduction in harm across all languages on top of the data mixtures method using the same available data. Overall, our comprehensive study of merging approaches provides a useful framework for building strong and safe multilingual models.
☆ Context-Parametric Inversion: Why Instruction Finetuning May Not Actually Improve Context Reliance
Large language models are instruction-finetuned to enhance their ability to follow user instructions and process the input context. However, even state-of-the-art models often struggle to follow the instruction, especially when the input context is not aligned with the model's parametric knowledge. This manifests as various failures, such as hallucinations where the responses are outdated, biased or contain unverified facts. In this work, we try to understand the underlying reason for this poor context reliance, especially after instruction tuning. We observe an intriguing phenomenon: during instruction tuning, the context reliance initially increases as expected, but then gradually decreases as instruction finetuning progresses. We call this phenomenon context-parametric inversion and observe it across multiple general purpose instruction tuning datasets like TULU, Alpaca and Ultrachat, as well as model families such as Llama, Mistral and Pythia. In a simple theoretical setup, we isolate why context-parametric inversion occurs along the gradient descent trajectory of instruction finetuning. We tie this phenomena to examples in the instruction finetuning data mixture where the input context provides information that is already present in the model's parametric knowledge. Our analysis suggests natural mitigation strategies that provide some limited gains, while also validating our theoretical insights. We hope that our work serves as a starting point in addressing this failure mode in a staple part of LLM training.
comment: Under Review
☆ Semantic Image Inversion and Editing using Rectified Stochastic Differential Equations
Generative models transform random noise into images; their inversion aims to transform images back to structured noise for recovery and editing. This paper addresses two key tasks: (i) inversion and (ii) editing of a real image using stochastic equivalents of rectified flow models (such as Flux). Although Diffusion Models (DMs) have recently dominated the field of generative modeling for images, their inversion presents faithfulness and editability challenges due to nonlinearities in drift and diffusion. Existing state-of-the-art DM inversion approaches rely on training of additional parameters or test-time optimization of latent variables; both are expensive in practice. Rectified Flows (RFs) offer a promising alternative to diffusion models, yet their inversion has been underexplored. We propose RF inversion using dynamic optimal control derived via a linear quadratic regulator. We prove that the resulting vector field is equivalent to a rectified stochastic differential equation. Additionally, we extend our framework to design a stochastic sampler for Flux. Our inversion method allows for state-of-the-art performance in zero-shot inversion and editing, outperforming prior works in stroke-to-image synthesis and semantic image editing, with large-scale human evaluations confirming user preference.
comment: Preprint
☆ On Information-Theoretic Measures of Predictive Uncertainty
Reliable estimation of predictive uncertainty is crucial for machine learning applications, particularly in high-stakes scenarios where hedging against risks is essential. Despite its significance, a consensus on the correct measurement of predictive uncertainty remains elusive. In this work, we return to first principles to develop a fundamental framework of information-theoretic predictive uncertainty measures. Our proposed framework categorizes predictive uncertainty measures according to two factors: (I) The predicting model (II) The approximation of the true predictive distribution. Examining all possible combinations of these two factors, we derive a set of predictive uncertainty measures that includes both known and newly introduced ones. We empirically evaluate these measures in typical uncertainty estimation settings, such as misclassification detection, selective prediction, and out-of-distribution detection. The results show that no single measure is universal, but the effectiveness depends on the specific setting. Thus, our work provides clarity about the suitability of predictive uncertainty measures by clarifying their implicit assumptions and relationships.
☆ When Attention Sink Emerges in Language Models: An Empirical View
Language Models (LMs) assign significant attention to the first token, even if it is not semantically important, which is known as attention sink. This phenomenon has been widely adopted in applications such as streaming/long context generation, KV cache optimization, inference acceleration, model quantization, and others. Despite its widespread use, a deep understanding of attention sink in LMs is still lacking. In this work, we first demonstrate that attention sinks exist universally in LMs with various inputs, even in small models. Furthermore, attention sink is observed to emerge during the LM pre-training, motivating us to investigate how optimization, data distribution, loss function, and model architecture in LM pre-training influence its emergence. We highlight that attention sink emerges after effective optimization on sufficient training data. The sink position is highly correlated with the loss function and data distribution. Most importantly, we find that attention sink acts more like key biases, storing extra attention scores, which could be non-informative and not contribute to the value computation. We also observe that this phenomenon (at least partially) stems from tokens' inner dependence on attention scores as a result of softmax normalization. After relaxing such dependence by replacing softmax attention with other attention operations, such as sigmoid attention without normalization, attention sinks do not emerge in LMs up to 1B parameters. The code is available at https://github.com/sail-sg/Attention-Sink.
☆ Enhancing JEPAs with Spatial Conditioning: Robust and Efficient Representation Learning NeurIPS 2024
Image-based Joint-Embedding Predictive Architecture (IJEPA) offers an attractive alternative to Masked Autoencoder (MAE) for representation learning using the Masked Image Modeling framework. IJEPA drives representations to capture useful semantic information by predicting in latent rather than input space. However, IJEPA relies on carefully designed context and target windows to avoid representational collapse. The encoder modules in IJEPA cannot adaptively modulate the type of predicted and/or target features based on the feasibility of the masked prediction task as they are not given sufficient information of both context and targets. Based on the intuition that in natural images, information has a strong spatial bias with spatially local regions being highly predictive of one another compared to distant ones. We condition the target encoder and context encoder modules in IJEPA with positions of context and target windows respectively. Our "conditional" encoders show performance gains on several image classification benchmark datasets, improved robustness to context window size and sample-efficiency during pretraining.
comment: NeurIPS 2024 Workshop on Self-Supervised Learning - Theory and Practice. Comments welcome!
☆ AFlow: Automating Agentic Workflow Generation
Large language models (LLMs) have demonstrated remarkable potential in solving complex tasks across diverse domains, typically by employing agentic workflows that follow detailed instructions and operational sequences. However, constructing these workflows requires significant human effort, limiting scalability and generalizability. Recent research has sought to automate the generation and optimization of these workflows, but existing methods still rely on initial manual setup and fall short of achieving fully automated and effective workflow generation. To address this challenge, we reformulate workflow optimization as a search problem over code-represented workflows, where LLM-invoking nodes are connected by edges. We introduce AFlow, an automated framework that efficiently explores this space using Monte Carlo Tree Search, iteratively refining workflows through code modification, tree-structured experience, and execution feedback. Empirical evaluations across six benchmark datasets demonstrate AFlow's efficacy, yielding a 5.7% average improvement over state-of-the-art baselines. Furthermore, AFlow enables smaller models to outperform GPT-4o on specific tasks at 4.55% of its inference cost in dollars. The code will be available at https://github.com/geekan/MetaGPT.
☆ SplitLLM: Collaborative Inference of LLMs for Model Placement and Throughput Optimization
Large language models (LLMs) have been a disruptive innovation in recent years, and they play a crucial role in our daily lives due to their ability to understand and generate human-like text. Their capabilities include natural language understanding, information retrieval and search, translation, chatbots, virtual assistance, and many more. However, it is well known that LLMs are massive in terms of the number of parameters. Additionally, the self-attention mechanism in the underlying architecture of LLMs, Transformers, has quadratic complexity in terms of both computation and memory with respect to the input sequence length. For these reasons, LLM inference is resource-intensive, and thus, the throughput of LLM inference is limited, especially for the longer sequences. In this report, we design a collaborative inference architecture between a server and its clients to alleviate the throughput limit. In this design, we consider the available resources on both sides, i.e., the computation and communication costs. We develop a dynamic programming-based algorithm to optimally allocate computation between the server and the client device to increase the server throughput, while not violating the service level agreement (SLA). We show in the experiments that we are able to efficiently distribute the workload allowing for roughly 1/3 reduction in the server workload, while achieving 19 percent improvement over a greedy method. As a result, we are able to demonstrate that, in an environment with different types of LLM inference requests, the throughput of the server is improved.
☆ Adversarially Robust Out-of-Distribution Detection Using Lyapunov-Stabilized Embeddings
Despite significant advancements in out-of-distribution (OOD) detection, existing methods still struggle to maintain robustness against adversarial attacks, compromising their reliability in critical real-world applications. Previous studies have attempted to address this challenge by exposing detectors to auxiliary OOD datasets alongside adversarial training. However, the increased data complexity inherent in adversarial training, and the myriad of ways that OOD samples can arise during testing, often prevent these approaches from establishing robust decision boundaries. To address these limitations, we propose AROS, a novel approach leveraging neural ordinary differential equations (NODEs) with Lyapunov stability theorem in order to obtain robust embeddings for OOD detection. By incorporating a tailored loss function, we apply Lyapunov stability theory to ensure that both in-distribution (ID) and OOD data converge to stable equilibrium points within the dynamical system. This approach encourages any perturbed input to return to its stable equilibrium, thereby enhancing the model's robustness against adversarial perturbations. To not use additional data, we generate fake OOD embeddings by sampling from low-likelihood regions of the ID data feature space, approximating the boundaries where OOD data are likely to reside. To then further enhance robustness, we propose the use of an orthogonal binary layer following the stable feature space, which maximizes the separation between the equilibrium points of ID and OOD samples. We validate our method through extensive experiments across several benchmarks, demonstrating superior performance, particularly under adversarial attacks. Notably, our approach improves robust detection performance from 37.8% to 80.1% on CIFAR-10 vs. CIFAR-100 and from 29.0% to 67.0% on CIFAR-100 vs. CIFAR-10.
comment: Code and pre-trained models are available at https://github.com/AdaptiveMotorControlLab/AROS
☆ SensorBench: Benchmarking LLMs in Coding-Based Sensor Processing
Effective processing, interpretation, and management of sensor data have emerged as a critical component of cyber-physical systems. Traditionally, processing sensor data requires profound theoretical knowledge and proficiency in signal-processing tools. However, recent works show that Large Language Models (LLMs) have promising capabilities in processing sensory data, suggesting their potential as copilots for developing sensing systems. To explore this potential, we construct a comprehensive benchmark, SensorBench, to establish a quantifiable objective. The benchmark incorporates diverse real-world sensor datasets for various tasks. The results show that while LLMs exhibit considerable proficiency in simpler tasks, they face inherent challenges in processing compositional tasks with parameter selections compared to engineering experts. Additionally, we investigate four prompting strategies for sensor processing and show that self-verification can outperform all other baselines in 48% of tasks. Our study provides a comprehensive benchmark and prompting analysis for future developments, paving the way toward an LLM-based sensor processing copilot.
☆ Online Statistical Inference for Time-varying Sample-averaged Q-learning
Reinforcement learning (RL) has emerged as a key approach for training agents in complex and uncertain environments. Incorporating statistical inference in RL algorithms is essential for understanding and managing uncertainty in model performance. This paper introduces a time-varying batch-averaged Q-learning algorithm, termed sampleaveraged Q-learning, which improves upon traditional single-sample Q-learning by aggregating samples of rewards and next states to better account for data variability and uncertainty. We leverage the functional central limit theorem (FCLT) to establish a novel framework that provides insights into the asymptotic normality of the sample-averaged algorithm under mild conditions. Additionally, we develop a random scaling method for interval estimation, enabling the construction of confidence intervals without requiring extra hyperparameters. Numerical experiments conducted on classic OpenAI Gym environments show that the time-varying sample-averaged Q-learning method consistently outperforms both single-sample and constant-batch Q-learning methods, achieving superior accuracy while maintaining comparable learning speeds.
☆ Towards Calibrated Losses for Adversarial Robust Reject Option Classification ACML
Robustness towards adversarial attacks is a vital property for classifiers in several applications such as autonomous driving, medical diagnosis, etc. Also, in such scenarios, where the cost of misclassification is very high, knowing when to abstain from prediction becomes crucial. A natural question is which surrogates can be used to ensure learning in scenarios where the input points are adversarially perturbed and the classifier can abstain from prediction? This paper aims to characterize and design surrogates calibrated in "Adversarial Robust Reject Option" setting. First, we propose an adversarial robust reject option loss $\ell_{d}^{\gamma}$ and analyze it for the hypothesis set of linear classifiers ($\mathcal{H}_{\textrm{lin}}$). Next, we provide a complete characterization result for any surrogate to be $(\ell_{d}^{\gamma},\mathcal{H}_{\textrm{lin}})$- calibrated. To demonstrate the difficulty in designing surrogates to $\ell_{d}^{\gamma}$, we show negative calibration results for convex surrogates and quasi-concave conditional risk cases (these gave positive calibration in adversarial setting without reject option). We also empirically argue that Shifted Double Ramp Loss (DRL) and Shifted Double Sigmoid Loss (DSL) satisfy the calibration conditions. Finally, we demonstrate the robustness of shifted DRL and shifted DSL against adversarial perturbations on a synthetically generated dataset.
comment: Accepted at Asian Conference on Machine Learning (ACML) , 2024
☆ Towards LLM-guided Efficient and Interpretable Multi-linear Tensor Network Rank Selection
We propose a novel framework that leverages large language models (LLMs) to guide the rank selection in tensor network models for higher-order data analysis. By utilising the intrinsic reasoning capabilities and domain knowledge of LLMs, our approach offers enhanced interpretability of the rank choices and can effectively optimise the objective function. This framework enables users without specialised domain expertise to utilise tensor network decompositions and understand the underlying rationale within the rank selection process. Experimental results validate our method on financial higher-order datasets, demonstrating interpretable reasoning, strong generalisation to unseen test data, and its potential for self-enhancement over successive iterations. This work is placed at the intersection of large language models and higher-order data analysis.
☆ SeedLM: Compressing LLM Weights into Seeds of Pseudo-Random Generators
Large Language Models (LLMs) have transformed natural language processing, but face significant challenges in widespread deployment due to their high runtime cost. In this paper, we introduce SeedLM, a novel post-training compression method that uses seeds of pseudo-random generators to encode and compress model weights. Specifically, for each block of weights, we find a seed that is fed into a Linear Feedback Shift Register (LFSR) during inference to efficiently generate a random matrix. This matrix is then linearly combined with compressed coefficients to reconstruct the weight block. SeedLM reduces memory access and leverages idle compute cycles during inference, effectively speeding up memory-bound tasks by trading compute for fewer memory accesses. Unlike state-of-the-art compression methods that rely on calibration data, our approach is data-free and generalizes well across diverse tasks. Our experiments with Llama 3 70B, which is particularly challenging to compress, show that SeedLM achieves significantly better zero-shot accuracy retention at 4- and 3-bit than state-of-the-art techniques, while maintaining performance comparable to FP16 baselines. Additionally, FPGA-based tests demonstrate that 4-bit SeedLM, as model size increases to 70B, approaches a 4x speed-up over an FP16 Llama 2/3 baseline.
☆ Early Diagnoses of Acute Lymphoblastic Leukemia Using YOLOv8 and YOLOv11 Deep Learning Models
Thousands of individuals succumb annually to leukemia alone. This study explores the application of image processing and deep learning techniques for detecting Acute Lymphoblastic Leukemia (ALL), a severe form of blood cancer responsible for numerous annual fatalities. As artificial intelligence technologies advance, the research investigates the reliability of these methods in real-world scenarios. The study focuses on recent developments in ALL detection, particularly using the latest YOLO series models, to distinguish between malignant and benign white blood cells and to identify different stages of ALL, including early stages. Additionally, the models are capable of detecting hematogones, which are often misclassified as ALL. By utilizing advanced deep learning models like YOLOv8 and YOLOv11, the study achieves high accuracy rates reaching 98.8%, demonstrating the effectiveness of these algorithms across multiple datasets and various real-world situations.
comment: 4 pages, 6 figures, 3 tables
☆ Separation of Neural Drives to Muscles from Transferred Polyfunctional Nerves using Implanted Micro-electrode Arrays
Following limb amputation, neural signals for limb functions persist in the residual peripheral nerves. Targeted muscle reinnervation (TMR) allows to redirected these signals into spare muscles to recover the neural information through electromyography (EMG). However, a significant challenge arises in separating distinct neural commands redirected from the transferred nerves to the muscles. Disentangling overlapping signals from EMG recordings remains complex, as they can contain mixed neural information that complicates limb function interpretation. To address this challenge, Regenerative Peripheral Nerve Interfaces (RPNIs) surgically partition the nerve into individual fascicles that reinnervate specific muscle grafts, isolating distinct neural sources for more precise control and interpretation of EMG signals. We introduce a novel biointerface that combines TMR surgery of polyvalent nerves with a high-density micro-electrode array implanted at a single site within a reinnervated muscle. Instead of surgically identifying distinct nerve fascicles, our approach separates all neural signals that are re-directed into a single muscle, using the high spatio-temporal selectivity of the micro-electrode array and mathematical source separation methods. We recorded EMG signals from four reinnervated muscles while volunteers performed phantom limb tasks. The decomposition of these signals into motor unit activity revealed distinct clusters of motor neurons associated with diverse functional tasks. Notably, our method enabled the extraction of multiple neural commands within a single reinnervated muscle, eliminating the need for surgical nerve division. This approach not only has the potential of enhancing prosthesis control but also uncovers mechanisms of motor neuron synergies following TMR, providing valuable insights into how the central nervous system encodes movement after reinnervation.
☆ Dynamical loss functions shape landscape topography and improve learning in artificial neural networks
Dynamical loss functions are derived from standard loss functions used in supervised classification tasks, but they are modified such that the contribution from each class periodically increases and decreases. These oscillations globally alter the loss landscape without affecting the global minima. In this paper, we demonstrate how to transform cross-entropy and mean squared error into dynamical loss functions. We begin by discussing the impact of increasing the size of the neural network or the learning rate on the learning process. Building on this intuition, we propose several versions of dynamical loss functions and show how they significantly improve validation accuracy for networks of varying sizes. Finally, we explore how the landscape of these dynamical loss functions evolves during training, highlighting the emergence of instabilities that may be linked to edge-of-instability minimization.
☆ SAMPa: Sharpness-aware Minimization Parallelized NeurIPS
Sharpness-aware minimization (SAM) has been shown to improve the generalization of neural networks. However, each SAM update requires \emph{sequentially} computing two gradients, effectively doubling the per-iteration cost compared to base optimizers like SGD. We propose a simple modification of SAM, termed SAMPa, which allows us to fully parallelize the two gradient computations. SAMPa achieves a twofold speedup of SAM under the assumption that communication costs between devices are negligible. Empirical results show that SAMPa ranks among the most efficient variants of SAM in terms of computational time. Additionally, our method consistently outperforms SAM across both vision and language tasks. Notably, SAMPa theoretically maintains convergence guarantees even for \emph{fixed} perturbation sizes, which is established through a novel Lyapunov function. We in fact arrive at SAMPa by treating this convergence guarantee as a hard requirement -- an approach we believe is promising for developing SAM-based methods in general. Our code is available at \url{https://github.com/LIONS-EPFL/SAMPa}.
comment: Advances in Neural Information Processing Systems (NeurIPS), 2024
☆ Combinatorial Multi-armed Bandits: Arm Selection via Group Testing
This paper considers the problem of combinatorial multi-armed bandits with semi-bandit feedback and a cardinality constraint on the super-arm size. Existing algorithms for solving this problem typically involve two key sub-routines: (1) a parameter estimation routine that sequentially estimates a set of base-arm parameters, and (2) a super-arm selection policy for selecting a subset of base arms deemed optimal based on these parameters. State-of-the-art algorithms assume access to an exact oracle for super-arm selection with unbounded computational power. At each instance, this oracle evaluates a list of score functions, the number of which grows as low as linearly and as high as exponentially with the number of arms. This can be prohibitive in the regime of a large number of arms. This paper introduces a novel realistic alternative to the perfect oracle. This algorithm uses a combination of group-testing for selecting the super arms and quantized Thompson sampling for parameter estimation. Under a general separability assumption on the reward function, the proposed algorithm reduces the complexity of the super-arm-selection oracle to be logarithmic in the number of base arms while achieving the same regret order as the state-of-the-art algorithms that use exact oracles. This translates to at least an exponential reduction in complexity compared to the oracle-based approaches.
comment: 26 pages
☆ Enhancing Robustness in Deep Reinforcement Learning: A Lyapunov Exponent Approach
Deep reinforcement learning agents achieve state-of-the-art performance in a wide range of simulated control tasks. However, successful applications to real-world problems remain limited. One reason for this dichotomy is because the learned policies are not robust to observation noise or adversarial attacks. In this paper, we investigate the robustness of deep RL policies to a single small state perturbation in deterministic continuous control tasks. We demonstrate that RL policies can be deterministically chaotic as small perturbations to the system state have a large impact on subsequent state and reward trajectories. This unstable non-linear behaviour has two consequences: First, inaccuracies in sensor readings, or adversarial attacks, can cause significant performance degradation; Second, even policies that show robust performance in terms of rewards may have unpredictable behaviour in practice. These two facets of chaos in RL policies drastically restrict the application of deep RL to real-world problems. To address this issue, we propose an improvement on the successful Dreamer V3 architecture, implementing a Maximal Lyapunov Exponent regularisation. This new approach reduces the chaotic state dynamics, rendering the learnt policies more resilient to sensor noise or adversarial attacks and thereby improving the suitability of Deep Reinforcement Learning for real-world applications.
☆ Double Jeopardy and Climate Impact in the Use of Large Language Models: Socio-economic Disparities and Reduced Utility for Non-English Speakers
Artificial Intelligence (AI), particularly large language models (LLMs), holds the potential to bridge language and information gaps, which can benefit the economies of developing nations. However, our analysis of FLORES-200, FLORES+, Ethnologue, and World Development Indicators data reveals that these benefits largely favor English speakers. Speakers of languages in low-income and lower-middle-income countries face higher costs when using OpenAI's GPT models via APIs because of how the system processes the input -- tokenization. Around 1.5 billion people, speaking languages primarily from lower-middle-income countries, could incur costs that are 4 to 6 times higher than those faced by English speakers. Disparities in LLM performance are significant, and tokenization in models priced per token amplifies inequalities in access, cost, and utility. Moreover, using the quality of translation tasks as a proxy measure, we show that LLMs perform poorly in low-resource languages, presenting a ``double jeopardy" of higher costs and poor performance for these users. We also discuss the direct impact of fragmentation in tokenizing low-resource languages on climate. This underscores the need for fairer algorithm development to benefit all linguistic groups.
comment: Project GitHub repository at https://github.com/worldbank/double-jeopardy-in-llms
☆ Cross-Modal Few-Shot Learning: a Generative Transfer Learning Framework
Most existing studies on few-shot learning focus on unimodal settings, where models are trained to generalize on unseen data using only a small number of labeled examples from the same modality. However, real-world data are inherently multi-modal, and unimodal approaches limit the practical applications of few-shot learning. To address this gap, this paper introduces the Cross-modal Few-Shot Learning (CFSL) task, which aims to recognize instances from multiple modalities when only a few labeled examples are available. This task presents additional challenges compared to classical few-shot learning due to the distinct visual characteristics and structural properties unique to each modality. To tackle these challenges, we propose a Generative Transfer Learning (GTL) framework consisting of two stages: the first stage involves training on abundant unimodal data, and the second stage focuses on transfer learning to adapt to novel data. Our GTL framework jointly estimates the latent shared concept across modalities and in-modality disturbance in both stages, while freezing the generative module during the transfer phase to maintain the stability of the learned representations and prevent overfitting to the limited multi-modal samples. Our finds demonstrate that GTL has superior performance compared to state-of-the-art methods across four distinct multi-modal datasets: Sketchy, TU-Berlin, Mask1K, and SKSF-A. Additionally, the results suggest that the model can estimate latent concepts from vast unimodal data and generalize these concepts to unseen modalities using only a limited number of available samples, much like human cognitive processes.
comment: 19 pages, 7 figures
☆ Transforming Game Play: A Comparative Study of DCQN and DTQN Architectures in Reinforcement Learning
In this study, we investigate the performance of Deep Q-Networks utilizing Convolutional Neural Networks (CNNs) and Transformer architectures across three different Atari games. The advent of DQNs has significantly advanced Reinforcement Learning, enabling agents to directly learn optimal policies from high-dimensional sensory inputs from pixel or RAM data. While CNN-based DQNs have been extensively studied and deployed in various domains, Transformer-based DQNs are relatively unexplored. Our research aims to fill this gap by benchmarking the performance of both DCQNs and DTQNs across the Atari games Asteroids, Space Invaders, and Centipede. We find that in the 35-40 million parameter range, the DCQN outperforms the DTQN in speed across both ViT and Projection Architectures. We also find the DCQN outperforms the DTQN in all games except for Centipede.
comment: KSU C-Day Spring 2024
☆ Navigation under uncertainty: Trajectory prediction and occlusion reasoning with switching dynamical systems
Predicting future trajectories of nearby objects, especially under occlusion, is a crucial task in autonomous driving and safe robot navigation. Prior works typically neglect to maintain uncertainty about occluded objects and only predict trajectories of observed objects using high-capacity models such as Transformers trained on large datasets. While these approaches are effective in standard scenarios, they can struggle to generalize to the long-tail, safety-critical scenarios. In this work, we explore a conceptual framework unifying trajectory prediction and occlusion reasoning under the same class of structured probabilistic generative model, namely, switching dynamical systems. We then present some initial experiments illustrating its capabilities using the Waymo open dataset.
☆ QueST: Querying Functional and Structural Niches on Spatial Transcriptomics Data via Contrastive Subgraph Embedding
The functional or structural spatial regions within tissues, referred to as spatial niches, are elements for illustrating the spatial contexts of multicellular organisms. A key challenge is querying shared niches across diverse tissues, which is crucial for achieving a comprehensive understanding of the organization and phenotypes of cell populations. However, current data analysis methods predominantly focus on creating spatial-aware embeddings for cells, neglecting the development of niche-level representations for effective querying. To address this gap, we introduce QueST, a novel niche representation learning model designed for querying spatial niches across multiple samples. QueST utilizes a novel subgraph contrastive learning approach to explicitly capture niche-level characteristics and incorporates adversarial training to mitigate batch effects. We evaluate QueST on established benchmarks using human and mouse datasets, demonstrating its superiority over state-of-the-art graph representation learning methods in accurate niche queries. Overall, QueST offers a specialized model for spatial niche queries, paving the way for deeper insights into the patterns and mechanisms of cell spatial organization across tissues. Source code can be found at https://github.com/cmhimself/QueST.
☆ A Simple Baseline for Predicting Events with Auto-Regressive Tabular Transformers
Many real-world applications of tabular data involve using historic events to predict properties of new ones, for example whether a credit card transaction is fraudulent or what rating a customer will assign a product on a retail platform. Existing approaches to event prediction include costly, brittle, and application-dependent techniques such as time-aware positional embeddings, learned row and field encodings, and oversampling methods for addressing class imbalance. Moreover, these approaches often assume specific use-cases, for example that we know the labels of all historic events or that we only predict a pre-specified label and not the data's features themselves. In this work, we propose a simple but flexible baseline using standard autoregressive LLM-style transformers with elementary positional embeddings and a causal language modeling objective. Our baseline outperforms existing approaches across popular datasets and can be employed for various use-cases. We demonstrate that the same model can predict labels, impute missing values, or model event sequences.
comment: 10 pages, 6 pages of references+appendix
☆ DR-MPC: Deep Residual Model Predictive Control for Real-world Social Navigation
How can a robot safely navigate around people exhibiting complex motion patterns? Reinforcement Learning (RL) or Deep RL (DRL) in simulation holds some promise, although much prior work relies on simulators that fail to precisely capture the nuances of real human motion. To address this gap, we propose Deep Residual Model Predictive Control (DR-MPC), a method to enable robots to quickly and safely perform DRL from real-world crowd navigation data. By blending MPC with model-free DRL, DR-MPC overcomes the traditional DRL challenges of large data requirements and unsafe initial behavior. DR-MPC is initialized with MPC-based path tracking, and gradually learns to interact more effectively with humans. To further accelerate learning, a safety component estimates when the robot encounters out-of-distribution states and guides it away from likely collisions. In simulation, we show that DR-MPC substantially outperforms prior work, including traditional DRL and residual DRL models. Real-world experiments show our approach successfully enables a robot to navigate a variety of crowded situations with few errors using less than 4 hours of training data.
comment: 8 pages, 8 figures, under review for IEEE Robotics and Automation Letters (RA-L)
☆ Echo State Networks for Spatio-Temporal Area-Level Data
Spatio-temporal area-level datasets play a critical role in official statistics, providing valuable insights for policy-making and regional planning. Accurate modeling and forecasting of these datasets can be extremely useful for policymakers to develop informed strategies for future planning. Echo State Networks (ESNs) are efficient methods for capturing nonlinear temporal dynamics and generating forecasts. However, ESNs lack a direct mechanism to account for the neighborhood structure inherent in area-level data. Ignoring these spatial relationships can significantly compromise the accuracy and utility of forecasts. In this paper, we incorporate approximate graph spectral filters at the input stage of the ESN, thereby improving forecast accuracy while preserving the model's computational efficiency during training. We demonstrate the effectiveness of our approach using Eurostat's tourism occupancy dataset and show how it can support more informed decision-making in policy and planning contexts.
comment: 23 pages, 4 figures
☆ High-Dimensional Differential Parameter Inference in Exponential Family using Time Score Matching
This paper addresses differential inference in time-varying parametric probabilistic models, like graphical models with changing structures. Instead of estimating a high-dimensional model at each time and inferring changes later, we directly learn the differential parameter, i.e., the time derivative of the parameter. The main idea is treating the time score function of an exponential family model as a linear model of the differential parameter for direct estimation. We use time score matching to estimate parameter derivatives. We prove the consistency of a regularized score matching objective and demonstrate the finite-sample normality of a debiased estimator in high-dimensional settings. Our methodology effectively infers differential structures in high-dimensional graphical models, verified on simulated and real-world datasets.
comment: Daniel J. Williams and Leyang Wang contributed equally to this work
☆ Adapt-$\infty$: Scalable Lifelong Multimodal Instruction Tuning via Dynamic Data Selection
Visual instruction datasets from various distributors are released at different times and often contain a significant number of semantically redundant text-image pairs, depending on their task compositions (i.e., skills) or reference sources. This redundancy greatly limits the efficient deployment of lifelong adaptable multimodal large language models, hindering their ability to refine existing skills and acquire new competencies over time. To address this, we reframe the problem of Lifelong Instruction Tuning (LiIT) via data selection, where the model automatically selects beneficial samples to learn from earlier and new datasets based on the current state of acquired knowledge in the model. Based on empirical analyses that show that selecting the best data subset using a static importance measure is often ineffective for multi-task datasets with evolving distributions, we propose Adapt-$\infty$, a new multi-way and adaptive data selection approach that dynamically balances sample efficiency and effectiveness during LiIT. We construct pseudo-skill clusters by grouping gradient-based sample vectors. Next, we select the best-performing data selector for each skill cluster from a pool of selector experts, including our newly proposed scoring function, Image Grounding score. This data selector samples a subset of the most important samples from each skill cluster for training. To prevent the continuous increase in the size of the dataset pool during LiIT, which would result in excessive computation, we further introduce a cluster-wise permanent data pruning strategy to remove the most semantically redundant samples from each cluster, keeping computational requirements manageable. Training with samples selected by Adapt-$\infty$ alleviates catastrophic forgetting, especially for rare tasks, and promotes forward transfer across the continuum using only a fraction of the original datasets.
comment: First two authors contributed equally. Code: https://github.com/adymaharana/adapt-inf
☆ Robust Gradient Descent for Phase Retrieval
Recent progress in robust statistical learning has mainly tackled convex problems, like mean estimation or linear regression, with non-convex challenges receiving less attention. Phase retrieval exemplifies such a non-convex problem, requiring the recovery of a signal from only the magnitudes of its linear measurements, without phase (sign) information. While several non-convex methods, especially those involving the Wirtinger Flow algorithm, have been proposed for noiseless or mild noise settings, developing solutions for heavy-tailed noise and adversarial corruption remains an open challenge. In this paper, we investigate an approach that leverages robust gradient descent techniques to improve the Wirtinger Flow algorithm's ability to simultaneously cope with fourth moment bounded noise and adversarial contamination in both the inputs (covariates) and outputs (responses). We address two scenarios: known zero-mean noise and completely unknown noise. For the latter, we propose a preprocessing step that alters the problem into a new format that does not fit traditional phase retrieval approaches but can still be resolved with a tailored version of the algorithm for the zero-mean noise context.
☆ Lambda-Skip Connections: the architectural component that prevents Rank Collapse
Rank collapse, a phenomenon where embedding vectors in sequence models rapidly converge to a uniform token or equilibrium state, has recently gained attention in the deep learning literature. This phenomenon leads to reduced expressivity and potential training instabilities due to vanishing gradients. Empirical evidence suggests that architectural components like skip connections, LayerNorm, and MultiLayer Perceptrons (MLPs) play critical roles in mitigating rank collapse. While this issue is well-documented for transformers, alternative sequence models, such as State Space Models (SSMs), which have recently gained prominence, have not been thoroughly examined for similar vulnerabilities. This paper extends the theory of rank collapse from transformers to SSMs using a unifying framework that captures both architectures. We study how a parametrized version of the classic skip connection component, which we call \emph{lambda-skip connections}, provides guarantees for rank collapse prevention. Through analytical results, we present a sufficient condition to guarantee prevention of rank collapse across all the aforementioned architectures. We also study the necessity of this condition via ablation studies and analytical examples. To our knowledge, this is the first study that provides a general guarantee to prevent rank collapse, and that investigates rank collapse in the context of SSMs, offering valuable understanding for both theoreticians and practitioners. Finally, we validate our findings with experiments demonstrating the crucial role of architectural components such as skip connections and gating mechanisms in preventing rank collapse.
☆ BrainMVP: Multi-modal Vision Pre-training for Brain Image Analysis using Multi-parametric MRI
Accurate diagnosis of brain abnormalities is greatly enhanced by the inclusion of complementary multi-parametric MRI imaging data. There is significant potential to develop a universal pre-training model that can be quickly adapted for image modalities and various clinical scenarios. However, current models often rely on uni-modal image data, neglecting the cross-modal correlations among different image modalities or struggling to scale up pre-training in the presence of missing modality data. In this paper, we propose BrainMVP, a multi-modal vision pre-training framework for brain image analysis using multi-parametric MRI scans. First, we collect 16,022 brain MRI scans (over 2.4 million images), encompassing eight MRI modalities sourced from a diverse range of centers and devices. Then, a novel pre-training paradigm is proposed for the multi-modal MRI data, addressing the issue of missing modalities and achieving multi-modal information fusion. Cross-modal reconstruction is explored to learn distinctive brain image embeddings and efficient modality fusion capabilities. A modality-wise data distillation module is proposed to extract the essence representation of each MR image modality for both the pre-training and downstream application purposes. Furthermore, we introduce a modality-aware contrastive learning module to enhance the cross-modality association within a study. Extensive experiments on downstream tasks demonstrate superior performance compared to state-of-the-art pre-training methods in the medical domain, with Dice Score improvement of 0.28%-14.47% across six segmentation benchmarks and a consistent accuracy improvement of 0.65%-18.07% in four individual classification tasks.
☆ Neural networks that overcome classic challenges through practice
Since the earliest proposals for neural network models of the mind and brain, critics have pointed out key weaknesses in these models compared to human cognitive abilities. Here we review recent work that has used metalearning to help overcome some of these challenges. We characterize their successes as addressing an important developmental problem: they provide machines with an incentive to improve X (where X represents the desired capability) and opportunities to practice it, through explicit optimization for X; unlike conventional approaches that hope for achieving X through generalization from related but different objectives. We review applications of this principle to four classic challenges: systematicity, catastrophic forgetting, few-shot learning and multi-step reasoning; we also discuss related aspects of human development in natural environments.
☆ Online waveform selection for cognitive radar
Designing a cognitive radar system capable of adapting its parameters is challenging, particularly when tasked with tracking a ballistic missile throughout its entire flight. In this work, we focus on proposing adaptive algorithms that select waveform parameters in an online fashion. Our novelty lies in formulating the learning problem using domain knowledge derived from the characteristics of ballistic trajectories. We propose three reinforcement learning algorithms: bandwidth scaling, Q-learning, and Q-learning lookahead. These algorithms dynamically choose the bandwidth for each transmission based on received feedback. Through experiments on synthetically generated ballistic trajectories, we demonstrate that our proposed algorithms achieve the dual objectives of minimizing range error and maintaining continuous tracking without losing the target.
☆ TRESTLE: A Model of Concept Formation in Structured Domains
The literature on concept formation has demonstrated that humans are capable of learning concepts incrementally, with a variety of attribute types, and in both supervised and unsupervised settings. Many models of concept formation focus on a subset of these characteristics, but none account for all of them. In this paper, we present TRESTLE, an incremental account of probabilistic concept formation in structured domains that unifies prior concept learning models. TRESTLE works by creating a hierarchical categorization tree that can be used to predict missing attribute values and cluster sets of examples into conceptually meaningful groups. It updates its knowledge by partially matching novel structures and sorting them into its categorization tree. Finally, the system supports mixed-data representations, including nominal, numeric, relational, and component attributes. We evaluate TRESTLE's performance on a supervised learning task and an unsupervised clustering task. For both tasks, we compare it to a nonincremental model and to human participants. We find that this new categorization model is competitive with the nonincremental approach and more closely approximates human behavior on both tasks. These results serve as an initial demonstration of TRESTLE's capabilities and show that, by taking key characteristics of human learning into account, it can better model behavior than approaches that ignore them.
comment: 20 pages, 6 figures, 1 table
☆ TopoFR: A Closer Look at Topology Alignment on Face Recognition NeurIPS 2024
The field of face recognition (FR) has undergone significant advancements with the rise of deep learning. Recently, the success of unsupervised learning and graph neural networks has demonstrated the effectiveness of data structure information. Considering that the FR task can leverage large-scale training data, which intrinsically contains significant structure information, we aim to investigate how to encode such critical structure information into the latent space. As revealed from our observations, directly aligning the structure information between the input and latent spaces inevitably suffers from an overfitting problem, leading to a structure collapse phenomenon in the latent space. To address this problem, we propose TopoFR, a novel FR model that leverages a topological structure alignment strategy called PTSA and a hard sample mining strategy named SDE. Concretely, PTSA uses persistent homology to align the topological structures of the input and latent spaces, effectively preserving the structure information and improving the generalization performance of FR model. To mitigate the impact of hard samples on the latent space structure, SDE accurately identifies hard samples by automatically computing structure damage score (SDS) for each sample, and directs the model to prioritize optimizing these samples. Experimental results on popular face benchmarks demonstrate the superiority of our TopoFR over the state-of-the-art methods. Code and models are available at: https://github.com/modelscope/facechain/tree/main/face_module/TopoFR.
comment: Accepted by NeurIPS 2024
☆ STACKFEED: Structured Textual Actor-Critic Knowledge Base Editing with FeedBack
Large Language Models (LLMs) often generate incorrect or outdated information, especially in low-resource settings or when dealing with private data. To address this, Retrieval-Augmented Generation (RAG) uses external knowledge bases (KBs), but these can also suffer from inaccuracies. We introduce STACKFEED, a novel Structured Textual Actor-Critic Knowledge base editing with FEEDback approach that iteratively refines the KB based on expert feedback using a multi-actor, centralized critic reinforcement learning framework. Each document is assigned to an actor, modeled as a ReACT agent, which performs structured edits based on document-specific targeted instructions from a centralized critic. Experimental results show that STACKFEED significantly improves KB quality and RAG system performance, enhancing accuracy by up to 8% over baselines.
☆ Burning RED: Unlocking Subtask-Driven Reinforcement Learning and Risk-Awareness in Average-Reward Markov Decision Processes
Average-reward Markov decision processes (MDPs) provide a foundational framework for sequential decision-making under uncertainty. However, average-reward MDPs have remained largely unexplored in reinforcement learning (RL) settings, with the majority of RL-based efforts having been allocated to episodic and discounted MDPs. In this work, we study a unique structural property of average-reward MDPs and utilize it to introduce Reward-Extended Differential (or RED) reinforcement learning: a novel RL framework that can be used to effectively and efficiently solve various subtasks simultaneously in the average-reward setting. We introduce a family of RED learning algorithms for prediction and control, including proven-convergent algorithms for the tabular case. We then showcase the power of these algorithms by demonstrating how they can be used to learn a policy that optimizes, for the first time, the well-known conditional value-at-risk (CVaR) risk measure in a fully-online manner, without the use of an explicit bi-level optimization scheme or an augmented state-space.
comment: arXiv admin note: substantial text overlap with arXiv:2006.16318, arXiv:2110.13855 by other authors
☆ Regularized Robustly Reliable Learners and Instance Targeted Attacks
Instance-targeted data poisoning attacks, where an adversary corrupts a training set to induce errors on specific test points, have raised significant concerns. Balcan et al (2022) proposed an approach to addressing this challenge by defining a notion of robustly-reliable learners that provide per-instance guarantees of correctness under well-defined assumptions, even in the presence of data poisoning attacks. They then give a generic optimal (but computationally inefficient) robustly reliable learner as well as a computationally efficient algorithm for the case of linear separators over log-concave distributions. In this work, we address two challenges left open by Balcan et al (2022). The first is that the definition of robustly-reliable learners in Balcan et al (2022) becomes vacuous for highly-flexible hypothesis classes: if there are two classifiers h_0, h_1 \in H both with zero error on the training set such that h_0(x) \neq h_1(x), then a robustly-reliable learner must abstain on x. We address this problem by defining a modified notion of regularized robustly-reliable learners that allows for nontrivial statements in this case. The second is that the generic algorithm of Balcan et al (2022) requires re-running an ERM oracle (essentially, retraining the classifier) on each test point x, which is generally impractical even if ERM can be implemented efficiently. To tackle this problem, we show that at least in certain interesting cases we can design algorithms that can produce their outputs in time sublinear in training time, by using techniques from dynamic algorithm design.
☆ ROSAR: An Adversarial Re-Training Framework for Robust Side-Scan Sonar Object Detection
This paper introduces ROSAR, a novel framework enhancing the robustness of deep learning object detection models tailored for side-scan sonar (SSS) images, generated by autonomous underwater vehicles using sonar sensors. By extending our prior work on knowledge distillation (KD), this framework integrates KD with adversarial retraining to address the dual challenges of model efficiency and robustness against SSS noises. We introduce three novel, publicly available SSS datasets, capturing different sonar setups and noise conditions. We propose and formalize two SSS safety properties and utilize them to generate adversarial datasets for retraining. Through a comparative analysis of projected gradient descent (PGD) and patch-based adversarial attacks, ROSAR demonstrates significant improvements in model robustness and detection accuracy under SSS-specific conditions, enhancing the model's robustness by up to 1.85%. ROSAR is available at https://github.com/remaro-network/ROSAR-framework.
☆ SLaNC: Static LayerNorm Calibration NeurIPS 2024
The ever increasing sizes of Large Language Models (LLMs) beyond hundreds of billions of parameters have generated enormous pressure on the manufacturers of dedicated hardware accelerators and made the innovative design of the latter one of the most rapidly expanding fields of the AI industry. Various approaches have been explored to enable efficient and accurate processing of LLMs on the available accelerators given their computational and storage limitations. Among these, various quantization techniques have become the main focus of the community as a means of reducing the compute, communication and storage requirements. Quantization to lower precision formats naturally poses a number of challenges caused by the limited range of the available value representations. When it comes to processing the popular Transformer models on hardware, one of the main issues becomes calculation of the LayerNorm simply because accumulation of the variance requires a much wider dynamic range than the hardware enables. In this article, we address this matter and propose a computationally-efficient scaling technique that can be easily applied to Transformer models during inference. Our method suggests a straightforward way of scaling the LayerNorm inputs based on the static weights of the immediately preceding linear layers. The scaling factors are computed offline, based solely on the linear layer weights, hence no latency or computational overhead is added during inference. Most importantly, our technique ensures that no numerical issues such as overflow or underflow could happen during the compute. This approach offers smooth, accurate and resource-effective inference across a wide range of hardware architectures. The article provides theoretical justification as well as supporting numerical simulations.
comment: 9 pages, 3 figures, NeurIPS 2024 MLNCP Workshop
☆ Graph Classification Gaussian Processes via Hodgelet Spectral Features NeurIPS 2024
The problem of classifying graphs is ubiquitous in machine learning. While it is standard to apply graph neural networks for such tasks, Gaussian processes can also be used, by transforming graph features into the spectral domain, and using the resulting spectral features as input points. However, this approach only takes into account features on vertices, whereas some graph data also support features on edges. In this work, we present a Gaussian process-based classification algorithm that can utilise vertex and/or edges features to help classify graphs. Furthermore, we take advantage of the Hodge decomposition of vertex and edge features to increase the flexibility of the model, which can be beneficial on some tasks.
comment: NeurIPS 2024 Workshop on Bayesian Decision-Making and Uncertainty (oral presentation)
☆ Rethinking Legal Judgement Prediction in a Realistic Scenario in the Era of Large Language Models EMNLP 2024
This study investigates judgment prediction in a realistic scenario within the context of Indian judgments, utilizing a range of transformer-based models, including InLegalBERT, BERT, and XLNet, alongside LLMs such as Llama-2 and GPT-3.5 Turbo. In this realistic scenario, we simulate how judgments are predicted at the point when a case is presented for a decision in court, using only the information available at that time, such as the facts of the case, statutes, precedents, and arguments. This approach mimics real-world conditions, where decisions must be made without the benefit of hindsight, unlike retrospective analyses often found in previous studies. For transformer models, we experiment with hierarchical transformers and the summarization of judgment facts to optimize input for these models. Our experiments with LLMs reveal that GPT-3.5 Turbo excels in realistic scenarios, demonstrating robust performance in judgment prediction. Furthermore, incorporating additional legal information, such as statutes and precedents, significantly improves the outcome of the prediction task. The LLMs also provide explanations for their predictions. To evaluate the quality of these predictions and explanations, we introduce two human evaluation metrics: Clarity and Linking. Our findings from both automatic and human evaluations indicate that, despite advancements in LLMs, they are yet to achieve expert-level performance in judgment prediction and explanation tasks.
comment: Accepted on NLLP at EMNLP 2024
☆ Data-Driven Approaches for Modelling Target Behaviour
The performance of tracking algorithms strongly depends on the chosen model assumptions regarding the target dynamics. If there is a strong mismatch between the chosen model and the true object motion, the track quality may be poor or the track is easily lost. Still, the true dynamics might not be known a priori or it is too complex to be expressed in a tractable mathematical formulation. This paper provides a comparative study between three different methods that use machine learning to describe the underlying object motion based on training data. The first method builds on Gaussian Processes (GPs) for predicting the object motion, the second learns the parameters of an Interacting Multiple Model (IMM) filter and the third uses a Long Short-Term Memory (LSTM) network as a motion model. All methods are compared against an Extended Kalman Filter (EKF) with an analytic motion model as a benchmark and their respective strengths are highlighted in one simulated and two real-world scenarios.
comment: 11 pages, 9 figures. Submitted to IEEE Transactions on Signal Processing on October 14, 2024
☆ Reproducible Machine Learning-based Voice Pathology Detection: Introducing the Pitch Difference Feature
In this study, we propose a robust set of features derived from a thorough research of contemporary practices in voice pathology detection. The feature set is based on the combination of acoustic handcrafted features. Additionally, we introduce pitch difference as a novel feature. We combine this feature set, containing data from the publicly available Saarbr\"ucken Voice Database (SVD), with preprocessing using the K-Means Synthetic Minority Over-Sampling Technique algorithm to address class imbalance. Moreover, we applied multiple ML models as binary classifiers. We utilized support vector machine, k-nearest neighbors, naive Bayes, decision tree, random forest and AdaBoost classifiers. To determine the best classification approach, we performed grid search on feasible hyperparameters of respective classifiers and subsections of features. Our approach has achieved the state-of-the-art performance, measured by unweighted average recall in voice pathology detection on SVD database. We intentionally omit accuracy as it is highly biased metric in case of unbalanced data compared to aforementioned metrics. The results are further enhanced by eliminating the potential overestimation of the results with repeated stratified cross-validation. This advancement demonstrates significant potential for the clinical deployment of ML methods, offering a valuable tool for an objective examination of voice pathologies. To support our claims, we provide a publicly available GitHub repository with DOI 10.5281/zenodo.13771573. Finally, we provide REFORMS checklist.
comment: 33 pages, 8 figures, code repository: https://github.com/aailab-uct/Automated-Robust-and-Reproducible-Voice-Pathology-Detection
☆ Transparent Networks for Multivariate Time Series
Transparent models, which are machine learning models that produce inherently interpretable predictions, are receiving significant attention in high-stakes domains. However, despite much real-world data being collected as time series, there is a lack of studies on transparent time series models. To address this gap, we propose a novel transparent neural network model for time series called Generalized Additive Time Series Model (GATSM). GATSM consists of two parts: 1) independent feature networks to learn feature representations, and 2) a transparent temporal module to learn temporal patterns across different time steps using the feature representations. This structure allows GATSM to effectively capture temporal patterns and handle dynamic-length time series while preserving transparency. Empirical experiments show that GATSM significantly outperforms existing generalized additive models and achieves comparable performance to black-box time series models, such as recurrent neural networks and Transformer. In addition, we demonstrate that GATSM finds interesting patterns in time series. The source code is available at https://github.com/gim4855744/GATSM.
☆ Non-convergence to global minimizers in data driven supervised deep learning: Adam and stochastic gradient descent optimization provably fail to converge to global minimizers in the training of deep neural networks with ReLU activation
Deep learning methods - consisting of a class of deep neural networks (DNNs) trained by a stochastic gradient descent (SGD) optimization method - are nowadays key tools to solve data driven supervised learning problems. Despite the great success of SGD methods in the training of DNNs, it remains a fundamental open problem of research to explain the success and the limitations of such methods in rigorous theoretical terms. In particular, even in the standard setup of data driven supervised learning problems, it remained an open research problem to prove (or disprove) that SGD methods converge in the training of DNNs with the popular rectified linear unit (ReLU) activation function with high probability to global minimizers in the optimization landscape. In this work we answer this question negatively. Specifically, in this work we prove for a large class of SGD methods that the considered optimizer does with high probability not converge to global minimizers of the optimization problem. It turns out that the probability to not converge to a global minimizer converges at least exponentially quickly to one as the width of the first hidden layer of the ANN and the depth of the ANN, respectively, increase. The general non-convergence results of this work do not only apply to the plain vanilla standard SGD method but also to a large class of accelerated and adaptive SGD methods such as the momentum SGD, the Nesterov accelerated SGD, the Adagrad, the RMSProp, the Adam, the Adamax, the AMSGrad, and the Nadam optimizers.
comment: 91 pages
☆ Adaptive Probabilistic ODE Solvers Without Adaptive Memory Requirements
Despite substantial progress in recent years, probabilistic solvers with adaptive step sizes can still not solve memory-demanding differential equations -- unless we care only about a single point in time (which is far too restrictive; we want the whole time series). Counterintuitively, the culprit is the adaptivity itself: Its unpredictable memory demands easily exceed our machine's capabilities, making our simulations fail unexpectedly and without warning. Still, dropping adaptivity would abandon years of progress, which can't be the answer. In this work, we solve this conundrum. We develop an adaptive probabilistic solver with fixed memory demands building on recent developments in robust state estimation. Switching to our method (i) eliminates memory issues for long time series, (ii) accelerates simulations by orders of magnitude through unlocking just-in-time compilation, and (iii) makes adaptive probabilistic solvers compatible with scientific computing in JAX.
☆ Get Rid of Task Isolation: A Continuous Multi-task Spatio-Temporal Learning Framework NeurIPS 2024
Spatiotemporal learning has become a pivotal technique to enable urban intelligence. Traditional spatiotemporal models mostly focus on a specific task by assuming a same distribution between training and testing sets. However, given that urban systems are usually dynamic, multi-sourced with imbalanced data distributions, current specific task-specific models fail to generalize to new urban conditions and adapt to new domains without explicitly modeling interdependencies across various dimensions and types of urban data. To this end, we argue that there is an essential to propose a Continuous Multi-task Spatio-Temporal learning framework (CMuST) to empower collective urban intelligence, which reforms the urban spatiotemporal learning from single-domain to cooperatively multi-dimensional and multi-task learning. Specifically, CMuST proposes a new multi-dimensional spatiotemporal interaction network (MSTI) to allow cross-interactions between context and main observations as well as self-interactions within spatial and temporal aspects to be exposed, which is also the core for capturing task-level commonality and personalization. To ensure continuous task learning, a novel Rolling Adaptation training scheme (RoAda) is devised, which not only preserves task uniqueness by constructing data summarization-driven task prompts, but also harnesses correlated patterns among tasks by iterative model behavior modeling. We further establish a benchmark of three cities for multi-task spatiotemporal learning, and empirically demonstrate the superiority of CMuST via extensive evaluations on these datasets. The impressive improvements on both few-shot streaming data and new domain tasks against existing SOAT methods are achieved. Code is available at https://github.com/DILab-USTCSZ/CMuST.
comment: Accepted by NeurIPS 2024
☆ Inverse Problems and Data Assimilation: A Machine Learning Approach
The aim of these notes is to demonstrate the potential for ideas in machine learning to impact on the fields of inverse problems and data assimilation. The perspective is one that is primarily aimed at researchers from inverse problems and/or data assimilation who wish to see a mathematical presentation of machine learning as it pertains to their fields. As a by-product, we include a succinct mathematical treatment of various topics in machine learning.
comment: 254 pages
☆ Continual Deep Reinforcement Learning to Prevent Catastrophic Forgetting in Jamming Mitigation
Deep Reinforcement Learning (DRL) has been highly effective in learning from and adapting to RF environments and thus detecting and mitigating jamming effects to facilitate reliable wireless communications. However, traditional DRL methods are susceptible to catastrophic forgetting (namely forgetting old tasks when learning new ones), especially in dynamic wireless environments where jammer patterns change over time. This paper considers an anti-jamming system and addresses the challenge of catastrophic forgetting in DRL applied to jammer detection and mitigation. First, we demonstrate the impact of catastrophic forgetting in DRL when applied to jammer detection and mitigation tasks, where the network forgets previously learned jammer patterns while adapting to new ones. This catastrophic interference undermines the effectiveness of the system, particularly in scenarios where the environment is non-stationary. We present a method that enables the network to retain knowledge of old jammer patterns while learning to handle new ones. Our approach substantially reduces catastrophic forgetting, allowing the anti-jamming system to learn new tasks without compromising its ability to perform previously learned tasks effectively. Furthermore, we introduce a systematic methodology for sequentially learning tasks in the anti-jamming framework. By leveraging continual DRL techniques based on PackNet, we achieve superior anti-jamming performance compared to standard DRL methods. Our proposed approach not only addresses catastrophic forgetting but also enhances the adaptability and robustness of the system in dynamic jamming environments. We demonstrate the efficacy of our method in preserving knowledge of past jammer patterns, learning new tasks efficiently, and achieving superior anti-jamming performance compared to traditional DRL approaches.
comment: IEEE MILCOM 2024
☆ AI-based particle track identification in scintillating fibres read out with imaging sensors
This paper presents the development and application of an AI-based method for particle track identification using scintillating fibres read out with imaging sensors. We propose a variational autoencoder (VAE) to efficiently filter and identify frames containing signal from the substantial data generated by SPAD array sensors. Our VAE model, trained on purely background frames, demonstrated a high capability to distinguish frames containing particle tracks from background noise. The performance of the VAE-based anomaly detection was validated with experimental data, demonstrating the method's ability to efficiently identify relevant events with rapid processing time, suggesting a solid prospect for deployment as a fast inference tool on hardware for real-time anomaly detection. This work highlights the potential of combining advanced sensor technology with machine learning techniques to enhance particle detection and tracking.
comment: 19 pages, 9 figures
☆ UniGEM: A Unified Approach to Generation and Property Prediction for Molecules
Molecular generation and molecular property prediction are both crucial for drug discovery, but they are often developed independently. Inspired by recent studies, which demonstrate that diffusion model, a prominent generative approach, can learn meaningful data representations that enhance predictive tasks, we explore the potential for developing a unified generative model in the molecular domain that effectively addresses both molecular generation and property prediction tasks. However, the integration of these tasks is challenging due to inherent inconsistencies, making simple multi-task learning ineffective. To address this, we propose UniGEM, the first unified model to successfully integrate molecular generation and property prediction, delivering superior performance in both tasks. Our key innovation lies in a novel two-phase generative process, where predictive tasks are activated in the later stages, after the molecular scaffold is formed. We further enhance task balance through innovative training strategies. Rigorous theoretical analysis and comprehensive experiments demonstrate our significant improvements in both tasks. The principles behind UniGEM hold promise for broader applications, including natural language processing and computer vision.
comment: 11 pages, 5 figures
☆ Do we need more complex representations for structure? A comparison of note duration representation for Music Transformers ECML
In recent years, deep learning has achieved formidable results in creative computing. When it comes to music, one viable model for music generation are Transformer based models. However, while transformers models are popular for music generation, they often rely on annotated structural information. In this work, we inquire if the off-the-shelf Music Transformer models perform just as well on structural similarity metrics using only unannotated MIDI information. We show that a slight tweak to the most common representation yields small but significant improvements. We also advocate that searching for better unannotated musical representations is more cost-effective than producing large amounts of curated and annotated data.
comment: Presented at the Music for Machine Learning Workshop with ECML/PKDD. To be published by Springer
☆ Artificial Intelligence-Based Triaging of Cutaneous Melanocytic Lesions
Pathologists are facing an increasing workload due to a growing volume of cases and the need for more comprehensive diagnoses. Aiming to facilitate workload reduction and faster turnaround times, we developed an artificial intelligence (AI) model for triaging cutaneous melanocytic lesions based on whole slide images. The AI model was developed and validated using a retrospective cohort from the UMC Utrecht. The dataset consisted of 52,202 whole slide images from 27,167 unique specimens, acquired from 20,707 patients. Specimens with only common nevi were assigned to the low complexity category (86.6%). In contrast, specimens with any other melanocytic lesion subtype, including non-common nevi, melanocytomas, and melanomas, were assigned to the high complexity category (13.4%). The dataset was split on patient level into a development set (80%) and test sets (20%) for independent evaluation. Predictive performance was primarily measured using the area under the receiver operating characteristic curve (AUROC) and the area under the precision-recall curve (AUPRC). A simulation experiment was performed to study the effect of implementing AI-based triaging in the clinic. The AI model reached an AUROC of 0.966 (95% CI, 0.960-0.972) and an AUPRC of 0.857 (95% CI, 0.836-0.877) on the in-distribution test set, and an AUROC of 0.899 (95% CI, 0.860-0.934) and an AUPRC of 0.498 (95% CI, 0.360-0.639) on the out-of-distribution test set. In the simulation experiment, using random case assignment as baseline, AI-based triaging prevented an average of 43.9 (95% CI, 36-55) initial examinations of high complexity cases by general pathologists for every 500 cases. In conclusion, the AI model achieved a strong predictive performance in differentiating between cutaneous melanocytic lesions of high and low complexity. The improvement in workflow efficiency due to AI-based triaging could be substantial.
comment: 14 pages, 6 figures
☆ Comparison of deep learning and conventional methods for disease onset prediction
Background: Conventional prediction methods such as logistic regression and gradient boosting have been widely utilized for disease onset prediction for their reliability and interpretability. Deep learning methods promise enhanced prediction performance by extracting complex patterns from clinical data, but face challenges like data sparsity and high dimensionality. Methods: This study compares conventional and deep learning approaches to predict lung cancer, dementia, and bipolar disorder using observational data from eleven databases from North America, Europe, and Asia. Models were developed using logistic regression, gradient boosting, ResNet, and Transformer, and validated both internally and externally across the data sources. Discrimination performance was assessed using AUROC, and calibration was evaluated using Eavg. Findings: Across 11 datasets, conventional methods generally outperformed deep learning methods in terms of discrimination performance, particularly during external validation, highlighting their better transportability. Learning curves suggest that deep learning models require substantially larger datasets to reach the same performance levels as conventional methods. Calibration performance was also better for conventional methods, with ResNet showing the poorest calibration. Interpretation: Despite the potential of deep learning models to capture complex patterns in structured observational healthcare data, conventional models remain highly competitive for disease onset prediction, especially in scenarios involving smaller datasets and if lengthy training times need to be avoided. The study underscores the need for future research focused on optimizing deep learning models to handle the sparsity, high dimensionality, and heterogeneity inherent in healthcare datasets, and find new strategies to exploit the full capabilities of deep learning methods.
☆ A Kernelizable Primal-Dual Formulation of the Multilinear Singular Value Decomposition
The ability to express a learning task in terms of a primal and a dual optimization problem lies at the core of a plethora of machine learning methods. For example, Support Vector Machine (SVM), Least-Squares Support Vector Machine (LS-SVM), Ridge Regression (RR), Lasso Regression (LR), Principal Component Analysis (PCA), and more recently Singular Value Decomposition (SVD) have all been defined either in terms of primal weights or in terms of dual Lagrange multipliers. The primal formulation is computationally advantageous in the case of large sample size while the dual is preferred for high-dimensional data. Crucially, said learning problems can be made nonlinear through the introduction of a feature map in the primal problem, which corresponds to applying the kernel trick in the dual. In this paper we derive a primal-dual formulation of the Multilinear Singular Value Decomposition (MLSVD), which recovers as special cases both PCA and SVD. Besides enabling computational gains through the derived primal formulation, we propose a nonlinear extension of the MLSVD using feature maps, which results in a dual problem where a kernel tensor arises. We discuss potential applications in the context of signal analysis and deep learning.
☆ A Practical Approach to Causal Inference over Time
In this paper, we focus on estimating the causal effect of an intervention over time on a dynamical system. To that end, we formally define causal interventions and their effects over time on discrete-time stochastic processes (DSPs). Then, we show under which conditions the equilibrium states of a DSP, both before and after a causal intervention, can be captured by a structural causal model (SCM). With such an equivalence at hand, we provide an explicit mapping from vector autoregressive models (VARs), broadly applied in econometrics, to linear, but potentially cyclic and/or affected by unmeasured confounders, SCMs. The resulting causal VAR framework allows us to perform causal inference over time from observational time series data. Our experiments on synthetic and real-world datasets show that the proposed framework achieves strong performance in terms of observational forecasting while enabling accurate estimation of the causal effect of interventions on dynamical systems. We demonstrate, through a case study, the potential practical questions that can be addressed using the proposed causal VAR framework.
☆ Model-Based Differentially Private Knowledge Transfer for Large Language Models
As large language models (LLMs) become increasingly prevalent in web services, effectively leveraging domain-specific knowledge while ensuring privacy has become critical. Existing methods, such as retrieval-augmented generation (RAG) and differentially private data synthesis, often compromise either the utility of domain knowledge or the privacy of sensitive data, limiting their applicability in specialized domains. To address these challenges, we propose \textit{Llamdex}, a novel framework that integrates privacy-preserving, domain-specific models into LLMs. Our approach significantly enhances the accuracy of domain-specific tasks, achieving up to a 26\% improvement compared to existing methods under the same differential privacy constraints. Experimental results show that Llamdex not only improves the accuracy of LLM responses but also maintains comparable inference efficiency to the original LLM, highlighting its potential for real-world applications.
☆ The Implicit Bias of Structured State Space Models Can Be Poisoned With Clean Labels
Neural networks are powered by an implicit bias: a tendency of gradient descent to fit training data in a way that generalizes to unseen data. A recent class of neural network models gaining increasing popularity is structured state space models (SSMs), regarded as an efficient alternative to transformers. Prior work argued that the implicit bias of SSMs leads to generalization in a setting where data is generated by a low dimensional teacher. In this paper, we revisit the latter setting, and formally establish a phenomenon entirely undetected by prior work on the implicit bias of SSMs. Namely, we prove that while implicit bias leads to generalization under many choices of training data, there exist special examples whose inclusion in training completely distorts the implicit bias, to a point where generalization fails. This failure occurs despite the special training examples being labeled by the teacher, i.e. having clean labels! We empirically demonstrate the phenomenon, with SSMs trained independently and as part of non-linear neural networks. In the area of adversarial machine learning, disrupting generalization with cleanly labeled training examples is known as clean-label poisoning. Given the proliferation of SSMs, particularly in large language models, we believe significant efforts should be invested in further delineating their susceptibility to clean-label poisoning, and in developing methods for overcoming this susceptibility.
☆ Moirai-MoE: Empowering Time Series Foundation Models with Sparse Mixture of Experts
Time series foundation models have demonstrated impressive performance as zero-shot forecasters. However, achieving effectively unified training on time series remains an open challenge. Existing approaches introduce some level of model specialization to account for the highly heterogeneous nature of time series data. For instance, Moirai pursues unified training by employing multiple input/output projection layers, each tailored to handle time series at a specific frequency. Similarly, TimesFM maintains a frequency embedding dictionary for this purpose. We identify two major drawbacks to this human-imposed frequency-level model specialization: (1) Frequency is not a reliable indicator of the underlying patterns in time series. For example, time series with different frequencies can display similar patterns, while those with the same frequency may exhibit varied patterns. (2) Non-stationarity is an inherent property of real-world time series, leading to varied distributions even within a short context window of a single time series. Frequency-level specialization is too coarse-grained to capture this level of diversity. To address these limitations, this paper introduces Moirai-MoE, using a single input/output projection layer while delegating the modeling of diverse time series patterns to the sparse mixture of experts (MoE) within Transformers. With these designs, Moirai-MoE reduces reliance on human-defined heuristics and enables automatic token-level specialization. Extensive experiments on 39 datasets demonstrate the superiority of Moirai-MoE over existing foundation models in both in-distribution and zero-shot scenarios. Furthermore, this study conducts comprehensive model analyses to explore the inner workings of time series MoE foundation models and provides valuable insights for future research.
☆ Information propagation dynamics in Deep Graph Networks
Graphs are a highly expressive abstraction for modeling entities and their relations, such as molecular structures, social networks, and traffic networks. Deep Graph Networks (DGNs) have emerged as a family of deep learning models that can effectively process and learn such structured information. However, learning effective information propagation patterns within DGNs remains a critical challenge that heavily influences the model capabilities, both in the static domain and in the temporal domain (where features and/or topology evolve). Given this challenge, this thesis investigates the dynamics of information propagation within DGNs for static and dynamic graphs, focusing on their design as dynamical systems. Throughout this work, we provide theoretical and empirical evidence to demonstrate the effectiveness of our proposed architectures in propagating and preserving long-term dependencies between nodes, and in learning complex spatio-temporal patterns from irregular and sparsely sampled dynamic graphs. In summary, this thesis provides a comprehensive exploration of the intersection between graphs, deep learning, and dynamical systems, offering insights and advancements for the field of graph representation learning and paving the way for more effective and versatile graph-based learning models.
comment: PhD thesis
☆ TABCF: Counterfactual Explanations for Tabular Data Using a Transformer-Based VAE
In the field of Explainable AI (XAI), counterfactual (CF) explanations are one prominent method to interpret a black-box model by suggesting changes to the input that would alter a prediction. In real-world applications, the input is predominantly in tabular form and comprised of mixed data types and complex feature interdependencies. These unique data characteristics are difficult to model, and we empirically show that they lead to bias towards specific feature types when generating CFs. To overcome this issue, we introduce TABCF, a CF explanation method that leverages a transformer-based Variational Autoencoder (VAE) tailored for modeling tabular data. Our approach uses transformers to learn a continuous latent space and a novel Gumbel-Softmax detokenizer that enables precise categorical reconstruction while preserving end-to-end differentiability. Extensive quantitative evaluation on five financial datasets demonstrates that TABCF does not exhibit bias toward specific feature types, and outperforms existing methods in producing effective CFs that align with common CF desiderata.
comment: Paper accepted at ICAIF '24: 5th ACM International Conference on AI in Finance, Brooklyn, NY, USA, November 2024
☆ Compositional Shielding and Reinforcement Learning for Multi-Agent Systems
Deep reinforcement learning has emerged as a powerful tool for obtaining high-performance policies. However, the safety of these policies has been a long-standing issue. One promising paradigm to guarantee safety is a shield, which shields a policy from making unsafe actions. However, computing a shield scales exponentially in the number of state variables. This is a particular concern in multi-agent systems with many agents. In this work, we propose a novel approach for multi-agent shielding. We address scalability by computing individual shields for each agent. The challenge is that typical safety specifications are global properties, but the shields of individual agents only ensure local properties. Our key to overcome this challenge is to apply assume-guarantee reasoning. Specifically, we present a sound proof rule that decomposes a (global, complex) safety specification into (local, simple) obligations for the shields of the individual agents. Moreover, we show that applying the shields during reinforcement learning significantly improves the quality of the policies obtained for a given training budget. We demonstrate the effectiveness and scalability of our multi-agent shielding framework in two case studies, reducing the computation time from hours to seconds and achieving fast learning convergence.
☆ Advancing Academic Knowledge Retrieval via LLM-enhanced Representation Similarity Fusion KDD
In an era marked by robust technological growth and swift information renewal, furnishing researchers and the populace with top-tier, avant-garde academic insights spanning various domains has become an urgent necessity. The KDD Cup 2024 AQA Challenge is geared towards advancing retrieval models to identify pertinent academic terminologies from suitable papers for scientific inquiries. This paper introduces the LLM-KnowSimFuser proposed by Robo Space, which wins the 2nd place in the competition. With inspirations drawed from the superior performance of LLMs on multiple tasks, after careful analysis of the provided datasets, we firstly perform fine-tuning and inference using LLM-enhanced pre-trained retrieval models to introduce the tremendous language understanding and open-domain knowledge of LLMs into this task, followed by a weighted fusion based on the similarity matrix derived from the inference results. Finally, experiments conducted on the competition datasets show the superiority of our proposal, which achieved a score of 0.20726 on the final leaderboard.
comment: The 2nd Place of KDD Cup 2024 OAG-Challenge AQA
☆ Principled Bayesian Optimisation in Collaboration with Human Experts NeurIPS 2024
Bayesian optimisation for real-world problems is often performed interactively with human experts, and integrating their domain knowledge is key to accelerate the optimisation process. We consider a setup where experts provide advice on the next query point through binary accept/reject recommendations (labels). Experts' labels are often costly, requiring efficient use of their efforts, and can at the same time be unreliable, requiring careful adjustment of the degree to which any expert is trusted. We introduce the first principled approach that provides two key guarantees. (1) Handover guarantee: similar to a no-regret property, we establish a sublinear bound on the cumulative number of experts' binary labels. Initially, multiple labels per query are needed, but the number of expert labels required asymptotically converges to zero, saving both expert effort and computation time. (2) No-harm guarantee with data-driven trust level adjustment: our adaptive trust level ensures that the convergence rate will not be worse than the one without using advice, even if the advice from experts is adversarial. Unlike existing methods that employ a user-defined function that hand-tunes the trust level adjustment, our approach enables data-driven adjustments. Real-world applications empirically demonstrate that our method not only outperforms existing baselines, but also maintains robustness despite varying labelling accuracy, in tasks of battery design with human experts.
comment: Accepted to NeurIPS 2024 as a spotlight
☆ Mobility-Aware Federated Learning: Multi-Armed Bandit Based Selection in Vehicular Network
In this paper,we study a vehicle selection problem for federated learning (FL) over vehicular networks. Specifically, we design a mobility-aware vehicular federated learning (MAVFL) scheme in which vehicles drive through a road segment to perform FL. Some vehicles may drive out of the segment which leads to unsuccessful training.In the proposed scheme, the real-time successful training participation ratio is utilized to implement vehicle selection. We conduct the convergence analysis to indicate the influence of vehicle mobility on training loss. Furthermore, we propose a multi-armed bandit-based vehicle selection algorithm to minimize the utility function considering training loss and delay. The simulation results show that compared with baselines, the proposed algorithm can achieve better training performance with approximately 28\% faster convergence.
comment: Accepted by 2024 IEEE Globecom Workshops (GC Wkshps)
☆ Diversity-Aware Reinforcement Learning for de novo Drug Design
Fine-tuning a pre-trained generative model has demonstrated good performance in generating promising drug molecules. The fine-tuning task is often formulated as a reinforcement learning problem, where previous methods efficiently learn to optimize a reward function to generate potential drug molecules. Nevertheless, in the absence of an adaptive update mechanism for the reward function, the optimization process can become stuck in local optima. The efficacy of the optimal molecule in a local optimization may not translate to usefulness in the subsequent drug optimization process or as a potential standalone clinical candidate. Therefore, it is important to generate a diverse set of promising molecules. Prior work has modified the reward function by penalizing structurally similar molecules, primarily focusing on finding molecules with higher rewards. To date, no study has comprehensively examined how different adaptive update mechanisms for the reward function influence the diversity of generated molecules. In this work, we investigate a wide range of intrinsic motivation methods and strategies to penalize the extrinsic reward, and how they affect the diversity of the set of generated molecules. Our experiments reveal that combining structure- and prediction-based methods generally yields better results in terms of molecular diversity.
☆ A Stochastic Approach to Bi-Level Optimization for Hyperparameter Optimization and Meta Learning
We tackle the general differentiable meta learning problem that is ubiquitous in modern deep learning, including hyperparameter optimization, loss function learning, few-shot learning, invariance learning and more. These problems are often formalized as Bi-Level optimizations (BLO). We introduce a novel perspective by turning a given BLO problem into a stochastic optimization, where the inner loss function becomes a smooth probability distribution, and the outer loss becomes an expected loss over the inner distribution. To solve this stochastic optimization, we adopt Stochastic Gradient Langevin Dynamics (SGLD) MCMC to sample inner distribution, and propose a recurrent algorithm to compute the MC-estimated hypergradient. Our derivation is similar to forward-mode differentiation, but we introduce a new first-order approximation that makes it feasible for large models without needing to store huge Jacobian matrices. The main benefits are two-fold: i) Our stochastic formulation takes into account uncertainty, which makes the method robust to suboptimal inner optimization or non-unique multiple inner minima due to overparametrization; ii) Compared to existing methods that often exhibit unstable behavior and hyperparameter sensitivity in practice, our method leads to considerably more reliable solutions. We demonstrate that the new approach achieves promising results on diverse meta learning problems and easily scales to learning 87M hyperparameters in the case of Vision Transformers.
☆ Coupled autoregressive active inference agents for control of multi-joint dynamical systems
We propose an active inference agent to identify and control a mechanical system with multiple bodies connected by joints. This agent is constructed from multiple scalar autoregressive model-based agents, coupled together by virtue of sharing memories. Each subagent infers parameters through Bayesian filtering and controls by minimizing expected free energy over a finite time horizon. We demonstrate that a coupled agent of this kind is able to learn the dynamics of a double mass-spring-damper system, and drive it to a desired position through a balance of explorative and exploitative actions. It outperforms the uncoupled subagents in terms of surprise and goal alignment.
comment: 14 pages, 3 figures, accepted to the International Workshop on Active Inference 2024
☆ On Calibration of LLM-based Guard Models for Reliable Content Moderation
Large language models (LLMs) pose significant risks due to the potential for generating harmful content or users attempting to evade guardrails. Existing studies have developed LLM-based guard models designed to moderate the input and output of threat LLMs, ensuring adherence to safety policies by blocking content that violates these protocols upon deployment. However, limited attention has been given to the reliability and calibration of such guard models. In this work, we empirically conduct comprehensive investigations of confidence calibration for 9 existing LLM-based guard models on 12 benchmarks in both user input and model output classification. Our findings reveal that current LLM-based guard models tend to 1) produce overconfident predictions, 2) exhibit significant miscalibration when subjected to jailbreak attacks, and 3) demonstrate limited robustness to the outputs generated by different types of response models. Additionally, we assess the effectiveness of post-hoc calibration methods to mitigate miscalibration. We demonstrate the efficacy of temperature scaling and, for the first time, highlight the benefits of contextual calibration for confidence calibration of guard models, particularly in the absence of validation sets. Our analysis and experiments underscore the limitations of current LLM-based guard models and provide valuable insights for the future development of well-calibrated guard models toward more reliable content moderation. We also advocate for incorporating reliability evaluation of confidence calibration when releasing future LLM-based guard models.
comment: 19 pages, 9 figures
☆ Deterministic Apple Tasting
In binary ($0/1$) online classification with apple tasting feedback, the learner receives feedback only when predicting $1$. Besides some degenerate learning tasks, all previously known learning algorithms for this model are randomized. Consequently, prior to this work it was unknown whether deterministic apple tasting is generally feasible. In this work, we provide the first widely-applicable deterministic apple tasting learner, and show that in the realizable case, a hypothesis class is learnable if and only if it is deterministically learnable, confirming a conjecture of [Raman, Subedi, Raman, Tewari-24]. Quantitatively, we show that every class $\mathcal{H}$ is learnable with mistake bound $O \left(\sqrt{\mathtt{L}(\mathcal{H}) T \log T} \right)$ (where $\mathtt{L}(\mathcal{H})$ is the Littlestone dimension of $\mathcal{H}$), and that this is tight for some classes. We further study the agnostic case, in which the best hypothesis makes at most $k$ many mistakes, and prove a trichotomy stating that every class $\mathcal{H}$ must be either easy, hard, or unlearnable. Easy classes have (both randomized and deterministic) mistake bound $\Theta_{\mathcal{H}}(k)$. Hard classes have randomized mistake bound $\tilde{\Theta}_{\mathcal{H}} \left(k + \sqrt{T} \right)$, and deterministic mistake bound $\tilde{\Theta}_{\mathcal{H}} \left(\sqrt{k \cdot T} \right)$, where $T$ is the time horizon. Unlearnable classes have (both randomized and deterministic) mistake bound $\Theta(T)$. Our upper bound is based on a deterministic algorithm for learning from expert advice with apple tasting feedback, a problem interesting in its own right. For this problem, we show that the optimal deterministic mistake bound is $\Theta \left(\sqrt{T (k + \log n)} \right)$ for all $k$ and $T \leq n \leq 2^T$, where $n$ is the number of experts.
☆ Tighter Risk Bounds for Mixtures of Experts
In this work, we provide upper bounds on the risk of mixtures of experts by imposing local differential privacy (LDP) on their gating mechanism. These theoretical guarantees are tailored to mixtures of experts that utilize the one-out-of-$n$ gating mechanism, as opposed to the conventional $n$-out-of-$n$ mechanism. The bounds exhibit logarithmic dependence on the number of experts, and encapsulate the dependence on the gating mechanism in the LDP parameter, making them significantly tighter than existing bounds, under reasonable conditions. Experimental results support our theory, demonstrating that our approach enhances the generalization ability of mixtures of experts and validating the feasibility of imposing LDP on the gating mechanism.
☆ Improved Depth Estimation of Bayesian Neural Networks NeurIPS 2024
This paper proposes improvements over earlier work by Nazareth and Blei (2022) for estimating the depth of Bayesian neural networks. Here, we propose a discrete truncated normal distribution over the network depth to independently learn its mean and variance. Posterior distributions are inferred by minimizing the variational free energy, which balances the model complexity and accuracy. Our method improves test accuracy in the spiral data set and reduces the variance in posterior depth estimates.
comment: NeurIPS 2024 Workshop on Bayesian Decision-making and Uncertainty https://openreview.net/forum?id=6TLRVdWGzI
☆ PIVOT-R: Primitive-Driven Waypoint-Aware World Model for Robotic Manipulation NeurIPS 2024
Language-guided robotic manipulation is a challenging task that requires an embodied agent to follow abstract user instructions to accomplish various complex manipulation tasks. Previous work trivially fitting the data without revealing the relation between instruction and low-level executable actions, these models are prone to memorizing the surficial pattern of the data instead of acquiring the transferable knowledge, and thus are fragile to dynamic environment changes. To address this issue, we propose a PrIrmitive-driVen waypOinT-aware world model for Robotic manipulation (PIVOT-R) that focuses solely on the prediction of task-relevant waypoints. Specifically, PIVOT-R consists of a Waypoint-aware World Model (WAWM) and a lightweight action prediction module. The former performs primitive action parsing and primitive-driven waypoint prediction, while the latter focuses on decoding low-level actions. Additionally, we also design an asynchronous hierarchical executor (AHE), which can use different execution frequencies for different modules of the model, thereby helping the model reduce computational redundancy and improve model execution efficiency. Our PIVOT-R outperforms state-of-the-art (SoTA) open-source models on the SeaWave benchmark, achieving an average relative improvement of 19.45% across four levels of instruction tasks. Moreover, compared to the synchronously executed PIVOT-R, the execution efficiency of PIVOT-R with AHE is increased by 28-fold, with only a 2.9% drop in performance. These results provide compelling evidence that our PIVOT-R can significantly improve both the performance and efficiency of robotic manipulation.
comment: Accepted to NeurIPS 2024
☆ GIFT-Eval: A Benchmark For General Time Series Forecasting Model Evaluation
Time series foundation models excel in zero-shot forecasting, handling diverse tasks without explicit training. However, the advancement of these models has been hindered by the lack of comprehensive benchmarks. To address this gap, we introduce the General Time Series Forecasting Model Evaluation, GIFT-Eval, a pioneering benchmark aimed at promoting evaluation across diverse datasets. GIFT-Eval encompasses 28 datasets over 144,000 time series and 177 million data points, spanning seven domains, 10 frequencies, multivariate inputs, and prediction lengths ranging from short to long-term forecasts. To facilitate the effective pretraining and evaluation of foundation models, we also provide a non-leaking pretraining dataset containing approximately 230 billion data points. Additionally, we provide a comprehensive analysis of 17 baselines, which includes statistical models, deep learning models, and foundation models. We discuss each model in the context of various benchmark characteristics and offer a qualitative analysis that spans both deep learning and foundation models. We believe the insights from this analysis, along with access to this new standard zero-shot time series forecasting benchmark, will guide future developments in time series foundation models. The codebase, datasets, and a leaderboard showing all the results in detail will be available soon.
☆ Stein Variational Evolution Strategies
Stein Variational Gradient Descent (SVGD) is a highly efficient method to sample from an unnormalized probability distribution. However, the SVGD update relies on gradients of the log-density, which may not always be available. Existing gradient-free versions of SVGD make use of simple Monte Carlo approximations or gradients from surrogate distributions, both with limitations. To improve gradient-free Stein variational inference, we combine SVGD steps with evolution strategy (ES) updates. Our results demonstrate that the resulting algorithm generates high-quality samples from unnormalized target densities without requiring gradient information. Compared to prior gradient-free SVGD methods, we find that the integration of the ES update in SVGD significantly improves the performance on multiple challenging benchmark problems.
☆ Bayesian Optimisation with Unknown Hyperparameters: Regret Bounds Logarithmically Closer to Optimal
Bayesian Optimization (BO) is widely used for optimising black-box functions but requires us to specify the length scale hyperparameter, which defines the smoothness of the functions the optimizer will consider. Most current BO algorithms choose this hyperparameter by maximizing the marginal likelihood of the observed data, albeit risking misspecification if the objective function is less smooth in regions we have not yet explored. The only prior solution addressing this problem with theoretical guarantees was A-GP-UCB, proposed by Berkenkamp et al. (2019). This algorithm progressively decreases the length scale, expanding the class of functions considered by the optimizer. However, A-GP-UCB lacks a stopping mechanism, leading to over-exploration and slow convergence. To overcome this, we introduce Length scale Balancing (LB) - a novel approach, aggregating multiple base surrogate models with varying length scales. LB intermittently adds smaller length scale candidate values while retaining longer scales, balancing exploration and exploitation. We formally derive a cumulative regret bound of LB and compare it with the regret of an oracle BO algorithm using the optimal length scale. Denoting the factor by which the regret bound of A-GP-UCB was away from oracle as $g(T)$, we show that LB is only $\log g(T)$ away from oracle regret. We also empirically evaluate our algorithm on synthetic and real-world benchmarks and show it outperforms A-GP-UCB, maximum likelihood estimation and MCMC.
☆ Collaborative filtering based on nonnegative/binary matrix factorization
Collaborative filtering generates recommendations based on user-item similarities through rating data, which may involve numerous unrated items. To predict scores for unrated items, matrix factorization techniques, such as nonnegative matrix factorization (NMF), are often employed to predict scores for unrated items. Nonnegative/binary matrix factorization (NBMF), which is an extension of NMF, approximates a nonnegative matrix as the product of nonnegative and binary matrices. Previous studies have employed NBMF for image analysis where the data were dense. In this paper, we propose a modified NBMF algorithm that can be applied to collaborative filtering where data are sparse. In the modified method, unrated elements in a rating matrix are masked, which improves the collaborative filtering performance. Utilizing a low-latency Ising machine in NBMF is advantageous in terms of the computation time, making the proposed method beneficial.
comment: 12 pages, 7 figures
☆ Learning Sub-Second Routing Optimization in Computer Networks requires Packet-Level Dynamics
Finding efficient routes for data packets is an essential task in computer networking. The optimal routes depend greatly on the current network topology, state and traffic demand, and they can change within milliseconds. Reinforcement Learning can help to learn network representations that provide routing decisions for possibly novel situations. So far, this has commonly been done using fluid network models. We investigate their suitability for millisecond-scale adaptations with a range of traffic mixes and find that packet-level network models are necessary to capture true dynamics, in particular in the presence of TCP traffic. To this end, we present $\textit{PackeRL}$, the first packet-level Reinforcement Learning environment for routing in generic network topologies. Our experiments confirm that learning-based strategies that have been trained in fluid environments do not generalize well to this more realistic, but more challenging setup. Hence, we also introduce two new algorithms for learning sub-second Routing Optimization. We present $\textit{M-Slim}$, a dynamic shortest-path algorithm that excels at high traffic volumes but is computationally hard to scale to large network topologies, and $\textit{FieldLines}$, a novel next-hop policy design that re-optimizes routing for any network topology within milliseconds without requiring any re-training. Both algorithms outperform current learning-based approaches as well as commonly used static baseline protocols in scenarios with high-traffic volumes. All findings are backed by extensive experiments in realistic network conditions in our fast and versatile training and evaluation framework.
comment: Accepted at Transactions of Machine Learning Research (TMLR) 2024
☆ Sharpness-Aware Minimization Efficiently Selects Flatter Minima Late in Training
Sharpness-Aware Minimization (SAM) has substantially improved the generalization of neural networks under various settings. Despite the success, its effectiveness remains poorly understood. In this work, we discover an intriguing phenomenon in the training dynamics of SAM, shedding lights on understanding its implicit bias towards flatter minima over Stochastic Gradient Descent (SGD). Specifically, we find that SAM efficiently selects flatter minima late in training. Remarkably, even a few epochs of SAM applied at the end of training yield nearly the same generalization and solution sharpness as full SAM training. Subsequently, we delve deeper into the underlying mechanism behind this phenomenon. Theoretically, we identify two phases in the learning dynamics after applying SAM late in training: i) SAM first escapes the minimum found by SGD exponentially fast; and ii) then rapidly converges to a flatter minimum within the same valley. Furthermore, we empirically investigate the role of SAM during the early training phase. We conjecture that the optimization method chosen in the late phase is more crucial in shaping the final solution's properties. Based on this viewpoint, we extend our findings from SAM to Adversarial Training.
comment: 24 pages, 9 figures
☆ BookWorm: A Dataset for Character Description and Analysis EMNLP 2024
Characters are at the heart of every story, driving the plot and engaging readers. In this study, we explore the understanding of characters in full-length books, which contain complex narratives and numerous interacting characters. We define two tasks: character description, which generates a brief factual profile, and character analysis, which offers an in-depth interpretation, including character development, personality, and social context. We introduce the BookWorm dataset, pairing books from the Gutenberg Project with human-written descriptions and analyses. Using this dataset, we evaluate state-of-the-art long-context models in zero-shot and fine-tuning settings, utilizing both retrieval-based and hierarchical processing for book-length inputs. Our findings show that retrieval-based approaches outperform hierarchical ones in both tasks. Additionally, fine-tuned models using coreference-based retrieval produce the most factual descriptions, as measured by fact- and entailment-based metrics. We hope our dataset, experiments, and analysis will inspire further research in character-based narrative understanding.
comment: 30 pages, 2 figures, EMNLP 2024 Findings
☆ Groningen: Spatial Prediction of Rock Gas Saturation by Leveraging Selected and Augmented Well and Seismic Data with Classifier Ensembles
This paper presents a proof of concept for spatial prediction of rock saturation probability using classifier ensemble methods on the example of the giant Groningen gas field. The stages of generating 1481 seismic field attributes and selecting 63 significant attributes are described. The effectiveness of the proposed method of augmentation of well and seismic data is shown, which increased the training sample by 9 times. On a test sample of 42 wells (blind well test), the results demonstrate good accuracy in predicting the ensemble of classifiers: the Matthews correlation coefficient is 0.7689, and the F1-score for the "gas reservoir" class is 0.7949. Prediction of gas reservoir thicknesses within the field and adjacent areas is made.
comment: 19 pages, 9 figures, 7 tables
☆ Optimal Time Complexity Algorithms for Computing General Random Walk Graph Kernels on Sparse Graphs
We present the first linear time complexity randomized algorithms for unbiased approximation of the celebrated family of general random walk kernels (RWKs) for sparse graphs. This includes both labelled and unlabelled instances. The previous fastest methods for general RWKs were of cubic time complexity and not applicable to labelled graphs. Our method samples dependent random walks to compute novel graph embeddings in $\mathbb{R}^d$ whose dot product is equal to the true RWK in expectation. It does so without instantiating the direct product graph in memory, meaning we can scale to massive datasets that cannot be stored on a single machine. We derive exponential concentration bounds to prove that our estimator is sharp, and show that the ability to approximate general RWKs (rather than just special cases) unlocks efficient implicit graph kernel learning. Our method is up to $\mathbf{27\times}$ faster than its counterparts for efficient computation on large graphs and scales to graphs $\mathbf{128 \times}$ bigger than largest examples amenable to brute-force computation.
☆ SpeGCL: Self-supervised Graph Spectrum Contrastive Learning without Positive Samples
Graph Contrastive Learning (GCL) excels at managing noise and fluctuations in input data, making it popular in various fields (e.g., social networks, and knowledge graphs). Our study finds that the difference in high-frequency information between augmented graphs is greater than that in low-frequency information. However, most existing GCL methods focus mainly on the time domain (low-frequency information) for node feature representations and cannot make good use of high-frequency information to speed up model convergence. Furthermore, existing GCL paradigms optimize graph embedding representations by pulling the distance between positive sample pairs closer and pushing the distance between positive and negative sample pairs farther away, but our theoretical analysis shows that graph contrastive learning benefits from pushing negative pairs farther away rather than pulling positive pairs closer. To solve the above-mentioned problems, we propose a novel spectral GCL framework without positive samples, named SpeGCL. Specifically, to solve the problem that existing GCL methods cannot utilize high-frequency information, SpeGCL uses a Fourier transform to extract high-frequency and low-frequency information of node features, and constructs a contrastive learning mechanism in a Fourier space to obtain better node feature representation. Furthermore, SpeGCL relies entirely on negative samples to refine the graph embedding. We also provide a theoretical justification for the efficacy of using only negative samples in SpeGCL. Extensive experiments on un-supervised learning, transfer learning, and semi-supervised learning have validated the superiority of our SpeGCL framework over the state-of-the-art GCL methods.
comment: 13 pages, 3 figures
☆ Replay-and-Forget-Free Graph Class-Incremental Learning: A Task Profiling and Prompting Approach NeurIPS 2024
Class-incremental learning (CIL) aims to continually learn a sequence of tasks, with each task consisting of a set of unique classes. Graph CIL (GCIL) follows the same setting but needs to deal with graph tasks (e.g., node classification in a graph). The key characteristic of CIL lies in the absence of task identifiers (IDs) during inference, which causes a significant challenge in separating classes from different tasks (i.e., inter-task class separation). Being able to accurately predict the task IDs can help address this issue, but it is a challenging problem. In this paper, we show theoretically that accurate task ID prediction on graph data can be achieved by a Laplacian smoothing-based graph task profiling approach, in which each graph task is modeled by a task prototype based on Laplacian smoothing over the graph. It guarantees that the task prototypes of the same graph task are nearly the same with a large smoothing step, while those of different tasks are distinct due to differences in graph structure and node attributes. Further, to avoid the catastrophic forgetting of the knowledge learned in previous graph tasks, we propose a novel graph prompting approach for GCIL which learns a small discriminative graph prompt for each task, essentially resulting in a separate classification model for each task. The prompt learning requires the training of a single graph neural network (GNN) only once on the first task, and no data replay is required thereafter, thereby obtaining a GCIL model being both replay-free and forget-free. Extensive experiments on four GCIL benchmarks show that i) our task prototype-based method can achieve 100% task ID prediction accuracy on all four datasets, ii) our GCIL model significantly outperforms state-of-the-art competing methods by at least 18% in average CIL accuracy, and iii) our model is fully free of forgetting on the four datasets.
comment: Accepted by NeurIPS 2024
☆ CoMAT: Chain of Mathematically Annotated Thought Improves Mathematical Reasoning
Mathematical reasoning remains a significant challenge for large language models (LLMs), despite progress in prompting techniques such as Chain-of-Thought (CoT). We present Chain of Mathematically Annotated Thought (CoMAT), which enhances reasoning through two stages: Symbolic Conversion (converting natural language queries into symbolic form) and Reasoning Execution (deriving answers from symbolic representations). CoMAT operates entirely with a single LLM and without external solvers. Across four LLMs, CoMAT outperforms traditional CoT on six out of seven benchmarks, achieving gains of 4.48% on MMLU-Redux (MATH) and 4.58% on GaoKao MCQ. In addition to improved performance, CoMAT ensures faithfulness and verifiability, offering a transparent reasoning process for complex mathematical tasks
comment: 8 pages, 12 figures
☆ GraphCLIP: Enhancing Transferability in Graph Foundation Models for Text-Attributed Graphs
Recently, research on Text-Attributed Graphs (TAGs) has gained significant attention due to the prevalence of free-text node features in real-world applications and the advancements in Large Language Models (LLMs) that bolster TAG methodologies. However, current TAG approaches face two primary challenges: (i) Heavy reliance on label information and (ii) Limited cross-domain zero/few-shot transferability. These issues constrain the scaling of both data and model size, owing to high labor costs and scaling laws, complicating the development of graph foundation models with strong transferability. In this work, we propose the GraphCLIP framework to address these challenges by learning graph foundation models with strong cross-domain zero/few-shot transferability through a self-supervised contrastive graph-summary pretraining method. Specifically, we generate and curate large-scale graph-summary pair data with the assistance of LLMs, and introduce a novel graph-summary pretraining method, combined with invariant learning, to enhance graph foundation models with strong cross-domain zero-shot transferability. For few-shot learning, we propose a novel graph prompt tuning technique aligned with our pretraining objective to mitigate catastrophic forgetting and minimize learning costs. Extensive experiments show the superiority of GraphCLIP in both zero-shot and few-shot settings, while evaluations across various downstream tasks confirm the versatility of GraphCLIP. Our code is available at: https://github.com/ZhuYun97/GraphCLIP
comment: Under Review
☆ Feature Averaging: An Implicit Bias of Gradient Descent Leading to Non-Robustness in Neural Networks
In this work, we investigate a particular implicit bias in the gradient descent training process, which we term "Feature Averaging", and argue that it is one of the principal factors contributing to non-robustness of deep neural networks. Despite the existence of multiple discriminative features capable of classifying data, neural networks trained by gradient descent exhibit a tendency to learn the average (or certain combination) of these features, rather than distinguishing and leveraging each feature individually. In particular, we provide a detailed theoretical analysis of the training dynamics of gradient descent in a two-layer ReLU network for a binary classification task, where the data distribution consists of multiple clusters with orthogonal cluster center vectors. We rigorously prove that gradient descent converges to the regime of feature averaging, wherein the weights associated with each hidden-layer neuron represent an average of the cluster centers (each center corresponding to a distinct feature). It leads the network classifier to be non-robust due to an attack that aligns with the negative direction of the averaged features. Furthermore, we prove that, with the provision of more granular supervised information, a two-layer multi-class neural network is capable of learning individual features, from which one can derive a binary classifier with the optimal robustness under our setting. Besides, we also conduct extensive experiments using synthetic datasets, MNIST and CIFAR-10 to substantiate the phenomenon of feature averaging and its role in adversarial robustness of neural networks. We hope the theoretical and empirical insights can provide a deeper understanding of the impact of the gradient descent training on feature learning process, which in turn influences the robustness of the network, and how more detailed supervision may enhance model robustness.
comment: 78 pages, 10 figures
☆ DiRW: Path-Aware Digraph Learning for Heterophily
Recently, graph neural network (GNN) has emerged as a powerful representation learning tool for graph-structured data. However, most approaches are tailored for undirected graphs, neglecting the abundant information embedded in the edges of directed graphs (digraphs). In fact, digraphs are widely applied in the real world (e.g., social networks and recommendations) and are also confirmed to offer a new perspective for addressing topological heterophily challenges (i.e., connected nodes have complex patterns of feature distribution or labels). Despite recent significant advancements in DiGNNs, existing spatial- and spectral-based methods have inherent limitations due to the complex learning mechanisms and reliance on high-quality topology, leading to low efficiency and unstable performance. To address these issues, we propose Directed Random Walk (DiRW), which can be viewed as a plug-and-play strategy or an innovative neural architecture that provides a guidance or new learning paradigm for most spatial-based methods or digraphs. Specifically, DiRW incorporates a direction-aware path sampler optimized from the perspectives of walk probability, length, and number in a weight-free manner by considering node profiles and topological structure. Building upon this, DiRW utilizes a node-wise learnable path aggregator for generalized messages obtained by our proposed adaptive walkers to represent the current node. Extensive experiments on 9 datasets demonstrate that DiRW: (1) enhances most spatial-based methods as a plug-and-play strategy; (2) achieves SOTA performance as a new digraph learning paradigm.
comment: Under Review
☆ QIANets: Quantum-Integrated Adaptive Networks for Reduced Latency and Improved Inference Times in CNN Models NeurIPS 2024
Convolutional neural networks (CNNs) have made significant advances in computer vision tasks, yet their high inference times and latency often limit real-world applicability. While model compression techniques have gained popularity as solutions, they often overlook the critical balance between low latency and uncompromised accuracy. By harnessing quantum-inspired pruning, tensor decomposition, and annealing-based matrix factorization - three quantum-inspired concepts - we introduce QIANets: a novel approach of redesigning the traditional GoogLeNet, DenseNet, and ResNet-18 model architectures to process more parameters and computations whilst maintaining low inference times. Despite experimental limitations, the method was tested and evaluated, demonstrating reductions in inference times, along with effective accuracy preservations.
comment: Accepted to NeurIPS 2024 workshop on Neural Compression
☆ A Multi-Task Text Classification Pipeline with Natural Language Explanations: A User-Centric Evaluation in Sentiment Analysis and Offensive Language Identification in Greek Tweets
Interpretability is a topic that has been in the spotlight for the past few years. Most existing interpretability techniques produce interpretations in the form of rules or feature importance. These interpretations, while informative, may be harder to understand for non-expert users and therefore, cannot always be considered as adequate explanations. To that end, explanations in natural language are often preferred, as they are easier to comprehend and also more presentable to end-users. This work introduces an early concept for a novel pipeline that can be used in text classification tasks, offering predictions and explanations in natural language. It comprises of two models: a classifier for labelling the text and an explanation generator which provides the explanation. The proposed pipeline can be adopted by any text classification task, given that ground truth rationales are available to train the explanation generator. Our experiments are centred around the tasks of sentiment analysis and offensive language identification in Greek tweets, using a Greek Large Language Model (LLM) to obtain the necessary explanations that can act as rationales. The experimental evaluation was performed through a user study based on three different metrics and achieved promising results for both datasets.
comment: Work In Progress
☆ ABBA-VSM: Time Series Classification using Symbolic Representation on the Edge
In recent years, Edge AI has become more prevalent with applications across various industries, from environmental monitoring to smart city management. Edge AI facilitates the processing of Internet of Things (IoT) data and provides privacy-enabled and latency-sensitive services to application users using Machine Learning (ML) algorithms, e.g., Time Series Classification (TSC). However, existing TSC algorithms require access to full raw data and demand substantial computing resources to train and use them effectively in runtime. This makes them impractical for deployment in resource-constrained Edge environments. To address this, in this paper, we propose an Adaptive Brownian Bridge-based Symbolic Aggregation Vector Space Model (ABBA-VSM). It is a new TSC model designed for classification services on Edge. Here, we first adaptively compress the raw time series into symbolic representations, thus capturing the changing trends of data. Subsequently, we train the classification model directly on these symbols. ABBA-VSM reduces communication data between IoT and Edge devices, as well as computation cycles, in the development of resource-efficient TSC services on Edge. We evaluate our solution with extensive experiments using datasets from the UCR time series classification archive. The results demonstrate that the ABBA-VSM achieves up to 80% compression ratio and 90-100% accuracy for binary classification. Whereas, for non-binary classification, it achieves an average compression ratio of 60% and accuracy ranging from 60-80%.
comment: 15 pages with references, 5 figures
☆ QUIS: Question-guided Insights Generation for Automated Exploratory Data Analysis
Discovering meaningful insights from a large dataset, known as Exploratory Data Analysis (EDA), is a challenging task that requires thorough exploration and analysis of the data. Automated Data Exploration (ADE) systems use goal-oriented methods with Large Language Models and Reinforcement Learning towards full automation. However, these methods require human involvement to anticipate goals that may limit insight extraction, while fully automated systems demand significant computational resources and retraining for new datasets. We introduce QUIS, a fully automated EDA system that operates in two stages: insight generation (ISGen) driven by question generation (QUGen). The QUGen module generates questions in iterations, refining them from previous iterations to enhance coverage without human intervention or manually curated examples. The ISGen module analyzes data to produce multiple relevant insights in response to each question, requiring no prior training and enabling QUIS to adapt to new datasets.
comment: 6 pages
☆ Matrix Sketching in Bandits: Current Pitfalls and New Framework
The utilization of sketching techniques has progressively emerged as a pivotal method for enhancing the efficiency of online learning. In linear bandit settings, current sketch-based approaches leverage matrix sketching to reduce the per-round time complexity from \(\Omega\left(d^2\right)\) to \(O(d)\), where \(d\) is the input dimension. Despite this improved efficiency, these approaches encounter critical pitfalls: if the spectral tail of the covariance matrix does not decrease rapidly, it can lead to linear regret. In this paper, we revisit the regret analysis and algorithm design concerning approximating the covariance matrix using matrix sketching in linear bandits. We illustrate how inappropriate sketch sizes can result in unbounded spectral loss, thereby causing linear regret. To prevent this issue, we propose Dyadic Block Sketching, an innovative streaming matrix sketching approach that adaptively manages sketch size to constrain global spectral loss. This approach effectively tracks the best rank-\( k \) approximation in an online manner, ensuring efficiency when the geometry of the covariance matrix is favorable. Then, we apply the proposed Dyadic Block Sketching to linear bandits and demonstrate that the resulting bandit algorithm can achieve sublinear regret without prior knowledge of the covariance matrix, even under the worst case. Our method is a general framework for efficient sketch-based linear bandits, applicable to all existing sketch-based approaches, and offers improved regret bounds accordingly. Additionally, we conduct comprehensive empirical studies using both synthetic and real-world data to validate the accuracy of our theoretical findings and to highlight the effectiveness of our algorithm.
☆ LoLCATs: On Low-Rank Linearizing of Large Language Models
Recent works show we can linearize large language models (LLMs) -- swapping the quadratic attentions of popular Transformer-based LLMs with subquadratic analogs, such as linear attention -- avoiding the expensive pretraining costs. However, linearizing LLMs often significantly degrades model quality, still requires training over billions of tokens, and remains limited to smaller 1.3B to 7B LLMs. We thus propose Low-rank Linear Conversion via Attention Transfer (LoLCATs), a simple two-step method that improves LLM linearizing quality with orders of magnitudes less memory and compute. We base these steps on two findings. First, we can replace an LLM's softmax attentions with closely-approximating linear attentions, simply by training the linear attentions to match their softmax counterparts with an output MSE loss ("attention transfer"). Then, this enables adjusting for approximation errors and recovering LLM quality simply with low-rank adaptation (LoRA). LoLCATs significantly improves linearizing quality, training efficiency, and scalability. We significantly reduce the linearizing quality gap and produce state-of-the-art subquadratic LLMs from Llama 3 8B and Mistral 7B v0.1, leading to 20+ points of improvement on 5-shot MMLU. Furthermore, LoLCATs does so with only 0.2% of past methods' model parameters and 0.4% of their training tokens. Finally, we apply LoLCATs to create the first linearized 70B and 405B LLMs (50x larger than prior work). When compared with prior approaches under the same compute budgets, LoLCATs significantly improves linearizing quality, closing the gap between linearized and original Llama 3.1 70B and 405B LLMs by 77.8% and 78.1% on 5-shot MMLU.
comment: 47 pages, 20 figures, 18 tables, preprint
☆ Feedback Favors the Generalization of Neural ODEs
The well-known generalization problem hinders the application of artificial neural networks in continuous-time prediction tasks with varying latent dynamics. In sharp contrast, biological systems can neatly adapt to evolving environments benefiting from real-time feedback mechanisms. Inspired by the feedback philosophy, we present feedback neural networks, showing that a feedback loop can flexibly correct the learned latent dynamics of neural ordinary differential equations (neural ODEs), leading to a prominent generalization improvement. The feedback neural network is a novel two-DOF neural network, which possesses robust performance in unseen scenarios with no loss of accuracy performance on previous tasks. A linear feedback form is presented to correct the learned latent dynamics firstly, with a convergence guarantee. Then, domain randomization is utilized to learn a nonlinear neural feedback form. Finally, extensive tests including trajectory prediction of a real irregular object and model predictive control of a quadrotor with various uncertainties, are implemented, indicating significant improvements over state-of-the-art model-based and learning-based methods.
comment: 22 pages, 17 figures
☆ Measurability in the Fundamental Theorem of Statistical Learning
The Fundamental Theorem of Statistical Learning states that a hypothesis space is PAC learnable if and only if its VC dimension is finite. For the agnostic model of PAC learning, the literature so far presents proofs of this theorem that often tacitly impose several measurability assumptions on the involved sets and functions. We scrutinize these proofs from a measure-theoretic perspective in order to extract the assumptions needed for a rigorous argument. This leads to a sound statement as well as a detailed and self-contained proof of the Fundamental Theorem of Statistical Learning in the agnostic setting, showcasing the minimal measurability requirements needed. We then discuss applications in Model Theory, considering NIP and o-minimal structures. Our main theorem presents sufficient conditions for the PAC learnability of hypothesis spaces defined over o-minimal expansions of the reals.
comment: 41 pages plus appendix
☆ Revisiting and Benchmarking Graph Autoencoders: A Contrastive Learning Perspective
Graph autoencoders (GAEs) are self-supervised learning models that can learn meaningful representations of graph-structured data by reconstructing the input graph from a low-dimensional latent space. Over the past few years, GAEs have gained significant attention in academia and industry. In particular, the recent advent of GAEs with masked autoencoding schemes marks a significant advancement in graph self-supervised learning research. While numerous GAEs have been proposed, the underlying mechanisms of GAEs are not well understood, and a comprehensive benchmark for GAEs is still lacking. In this work, we bridge the gap between GAEs and contrastive learning by establishing conceptual and methodological connections. We revisit the GAEs studied in previous works and demonstrate how contrastive learning principles can be applied to GAEs. Motivated by these insights, we introduce lrGAE (left-right GAE), a general and powerful GAE framework that leverages contrastive learning principles to learn meaningful representations. Our proposed lrGAE not only facilitates a deeper understanding of GAEs but also sets a new benchmark for GAEs across diverse graph-based learning tasks. The source code for lrGAE, including the baselines and all the code for reproducing the results, is publicly available at https://github.com/EdisonLeeeee/lrGAE.
comment: Preprint, under review
☆ Learning via Surrogate PAC-Bayes
PAC-Bayes learning is a comprehensive setting for (i) studying the generalisation ability of learning algorithms and (ii) deriving new learning algorithms by optimising a generalisation bound. However, optimising generalisation bounds might not always be viable for tractable or computational reasons, or both. For example, iteratively querying the empirical risk might prove computationally expensive. In response, we introduce a novel principled strategy for building an iterative learning algorithm via the optimisation of a sequence of surrogate training objectives, inherited from PAC-Bayes generalisation bounds. The key argument is to replace the empirical risk (seen as a function of hypotheses) in the generalisation bound by its projection onto a constructible low dimensional functional space: these projections can be queried much more efficiently than the initial risk. On top of providing that generic recipe for learning via surrogate PAC-Bayes bounds, we (i) contribute theoretical results establishing that iteratively optimising our surrogates implies the optimisation of the original generalisation bounds, (ii) instantiate this strategy to the framework of meta-learning, introducing a meta-objective offering a closed form expression for meta-gradient, (iii) illustrate our approach with numerical experiments inspired by an industrial biochemical problem.
☆ Neural Quasiprobabilistic Likelihood Ratio Estimation with Negatively Weighted Data
Motivated by real-world situations found in high energy particle physics, we consider a generalisation of the likelihood-ratio estimation task to a quasiprobabilistic setting where probability densities can be negative. By extension, this framing also applies to importance sampling in a setting where the importance weights can be negative. The presence of negative densities and negative weights, pose an array of challenges to traditional neural likelihood ratio estimation methods. We address these challenges by introducing a novel loss function. In addition, we introduce a new model architecture based on the decomposition of a likelihood ratio using signed mixture models, providing a second strategy for overcoming these challenges. Finally, we demonstrate our approach on a pedagogical example and a real-world example from particle physics.
comment: 59 pages, 29 figures
☆ SkillAggregation: Reference-free LLM-Dependent Aggregation
Large Language Models (LLMs) are increasingly used to assess NLP tasks due to their ability to generate human-like judgments. Single LLMs were used initially, however, recent work suggests using multiple LLMs as judges yields improved performance. An important step in exploiting multiple judgements is the combination stage, aggregation. Existing methods in NLP either assign equal weight to all LLM judgments or are designed for specific tasks such as hallucination detection. This work focuses on aggregating predictions from multiple systems where no reference labels are available. A new method called SkillAggregation is proposed, which learns to combine estimates from LLM judges without needing additional data or ground truth. It extends the Crowdlayer aggregation method, developed for image classification, to exploit the judge estimates during inference. The approach is compared to a range of standard aggregation methods on HaluEval-Dialogue, TruthfulQA and Chatbot Arena tasks. SkillAggregation outperforms Crowdlayer on all tasks, and yields the best performance over all approaches on the majority of tasks.
☆ Large Language Model-Enhanced Reinforcement Learning for Generic Bus Holding Control Strategies
Bus holding control is a widely-adopted strategy for maintaining stability and improving the operational efficiency of bus systems. Traditional model-based methods often face challenges with the low accuracy of bus state prediction and passenger demand estimation. In contrast, Reinforcement Learning (RL), as a data-driven approach, has demonstrated great potential in formulating bus holding strategies. RL determines the optimal control strategies in order to maximize the cumulative reward, which reflects the overall control goals. However, translating sparse and delayed control goals in real-world tasks into dense and real-time rewards for RL is challenging, normally requiring extensive manual trial-and-error. In view of this, this study introduces an automatic reward generation paradigm by leveraging the in-context learning and reasoning capabilities of Large Language Models (LLMs). This new paradigm, termed the LLM-enhanced RL, comprises several LLM-based modules: reward initializer, reward modifier, performance analyzer, and reward refiner. These modules cooperate to initialize and iteratively improve the reward function according to the feedback from training and test results for the specified RL-based task. Ineffective reward functions generated by the LLM are filtered out to ensure the stable evolution of the RL agents' performance over iterations. To evaluate the feasibility of the proposed LLM-enhanced RL paradigm, it is applied to various bus holding control scenarios, including a synthetic single-line system and a real-world multi-line system. The results demonstrate the superiority and robustness of the proposed paradigm compared to vanilla RL strategies, the LLM-based controller, and conventional space headway-based feedback control. This study sheds light on the great potential of utilizing LLMs in various smart mobility applications.
comment: 41 pages, 15 figures
☆ Fed-piLot: Optimizing LoRA Assignment for Efficient Federated Foundation Model Fine-Tuning
Foundation models (FMs) have shown remarkable advancements in enhancing the performance of intelligent applications. To address the need for data privacy in FM fine-tuning, federated learning has emerged as the de facto framework. Specifically, Federated FMs (FedFMs) fine-tuning using low-rank adaptation (LoRA) modules instead of the full model over multiple clients can achieve both parameter efficiency and data privacy. However, recent studies rarely address the challenges posed by clients with heterogeneous resources, particularly in GPU memory capacity. In this paper, we introduce Fed-piLot, an efficient FedFM fine-tuning framework with optimized local LoRA assignments for heterogeneous clients. By emphasizing the different memory consumption for training different LoRA layers, as well as the varying contributions of different layers to model performance, we formulate the LoRA assignment as a Knapsack Optimization Problem. We design a Local-Global Information Gain Score (IG-Score) based value function to optimize LoRA assignment under clients' memory constraints. To further mitigate the impact of heterogeneity in model updates, we propose a novel Spatial-Temporal model aggregation (STAgg) rule using the Dynamic Weight Adjustment (DWA) strategy. Experimental results on three datasets under both IID and non-IID conditions demonstrate the effectiveness and efficiency of Fed-piLot. The code will be publicly available.
☆ Predicting from Strings: Language Model Embeddings for Bayesian Optimization
Bayesian Optimization is ubiquitous in the field of experimental design and blackbox optimization for improving search efficiency, but has been traditionally restricted to regression models which are only applicable to fixed search spaces and tabular input features. We propose Embed-then-Regress, a paradigm for applying in-context regression over string inputs, through the use of string embedding capabilities of pretrained language models. By expressing all inputs as strings, we are able to perform general-purpose regression for Bayesian Optimization over various domains including synthetic, combinatorial, and hyperparameter optimization, obtaining comparable results to state-of-the-art Gaussian Process-based algorithms. Code can be found at github.com/google-research/optformer/embed_then_regress.
☆ Hamiltonian Neural Networks for Robust Out-of-Time Credit Scoring
This paper introduces a novel Hamiltonian-inspired neural network approach to credit scoring, designed to address the challenges of class imbalance and out-of-time (OOT) prediction in financial risk management. Drawing from concepts in Hamiltonian mechanics, we develop a symplectic optimizer and a new loss function to capture the complex dynamics of credit risk evolution. Using the Freddie Mac Single-Family Loan-Level Dataset, we evaluate our model's performance against other machine learning approaches. Our method shows superior discriminative power in OOT scenarios, as measured by the Area Under the Curve (AUC), indicating better ranking ability and robustness to class imbalance. The Hamiltonian-inspired approach shows particular strength in maintaining consistent performance between in-sample and OOT test sets, suggesting improved generalization to future, unseen data. These findings suggest that physics-inspired techniques offer a promising direction for developing more robust and reliable credit scoring models, particularly in uncertain economic situations.
☆ Gaussian Mixture Vector Quantization with Aggregated Categorical Posterior
The vector quantization is a widely used method to map continuous representation to discrete space and has important application in tokenization for generative mode, bottlenecking information and many other tasks in machine learning. Vector Quantized Variational Autoencoder (VQ-VAE) is a type of variational autoencoder using discrete embedding as latent. We generalize the technique further, enriching the probabilistic framework with a Gaussian mixture as the underlying generative model. This framework leverages a codebook of latent means and adaptive variances to capture complex data distributions. This principled framework avoids various heuristics and strong assumptions that are needed with the VQ-VAE to address training instability and to improve codebook utilization. This approach integrates the benefits of both discrete and continuous representations within a variational Bayesian framework. Furthermore, by introducing the \textit{Aggregated Categorical Posterior Evidence Lower Bound} (ALBO), we offer a principled alternative optimization objective that aligns variational distributions with the generative model. Our experiments demonstrate that GM-VQ improves codebook utilization and reduces information loss without relying on handcrafted heuristics.
☆ Is Parameter Collision Hindering Continual Learning in LLMs?
Large Language Models (LLMs) often suffer from catastrophic forgetting when learning multiple tasks sequentially, making continual learning (CL) essential for their dynamic deployment. Existing state-of-the-art (SOTA) methods, such as O-LoRA, typically focus on constructing orthogonality tasks to decouple parameter interdependence from various domains.In this paper, we reveal that building non-collision parameters is a more critical factor in addressing CL challenges. Our theoretical and experimental analyses demonstrate that non-collision parameters can provide better task orthogonality, which is a sufficient but unnecessary condition. Furthermore, knowledge from multiple domains will be preserved in non-collision parameter subspaces, making it more difficult to forget previously seen data. Leveraging this insight, we propose Non-collision Low-Rank Adaptation (N-LoRA), a simple yet effective approach leveraging low collision rates to enhance CL in LLMs. Experimental results on multiple CL benchmarks indicate that N-LoRA achieves superior performance (+2.9), higher task orthogonality (*4.1 times), and lower parameter collision (*58.1 times) than SOTA methods.
☆ GUISE: Graph GaUssIan Shading watErmark
In the expanding field of generative artificial intelligence, integrating robust watermarking technologies is essential to protect intellectual property and maintain content authenticity. Traditionally, watermarking techniques have been developed primarily for rich information media such as images and audio. However, these methods have not been adequately adapted for graph-based data, particularly molecular graphs. Latent 3D graph diffusion(LDM-3DG) is an ascendant approach in the molecular graph generation field. This model effectively manages the complexities of molecular structures, preserving essential symmetries and topological features. We adapt the Gaussian Shading, a proven performance lossless watermarking technique, to the latent graph diffusion domain to protect this sophisticated new technology. Our adaptation simplifies the watermark diffusion process through duplication and padding, making it adaptable and suitable for various message types. We conduct several experiments using the LDM-3DG model on publicly available datasets QM9 and Drugs, to assess the robustness and effectiveness of our technique. Our results demonstrate that the watermarked molecules maintain statistical parity in 9 out of 10 performance metrics compared to the original. Moreover, they exhibit a 100% detection rate and a 99% extraction rate in a 2D decoded pipeline, while also showing robustness against post-editing attacks.
☆ Identity-Focused Inference and Extraction Attacks on Diffusion Models
The increasing reliance on diffusion models for generating synthetic images has amplified concerns about the unauthorized use of personal data, particularly facial images, in model training. In this paper, we introduce a novel identity inference framework to hold model owners accountable for including individuals' identities in their training data. Our approach moves beyond traditional membership inference attacks by focusing on identity-level inference, providing a new perspective on data privacy violations. Through comprehensive evaluations on two facial image datasets, Labeled Faces in the Wild (LFW) and CelebA, our experiments demonstrate that the proposed membership inference attack surpasses baseline methods, achieving an attack success rate of up to 89% and an AUC-ROC of 0.91, while the identity inference attack attains 92% on LDM models trained on LFW, and the data extraction attack achieves 91.6% accuracy on DDPMs, validating the effectiveness of our approach across diffusion models.
comment: 5 figures, 3 tables,12 pages main body content
☆ Balanced Neural ODEs: nonlinear model order reduction and Koopman operator approxmations
Variational Autoencoders (VAEs) are a powerful framework for learning compact latent representations, while NeuralODEs excel in learning transient system dynamics. This work combines the strengths of both to create fast surrogate models with adjustable complexity. By leveraging the VAE's dimensionality reduction using a non-hierarchical prior, our method adaptively assigns stochastic noise, naturally complementing known NeuralODE training enhancements and enabling probabilistic time series modeling. We show that standard Latent ODEs struggle with dimensionality reduction in systems with time-varying inputs. Our approach mitigates this by continuously propagating variational parameters through time, establishing fixed information channels in latent space. This results in a flexible and robust method that can learn different system complexities, e.g. deep neural networks or linear matrices. Hereby, it enables efficient approximation of the Koopman operator without the need for predefining its dimensionality. As our method balances dimensionality reduction and reconstruction accuracy, we call it Balanced Neural ODE (B-NODE). We demonstrate the effectiveness of this method on academic test cases and apply it to a real-world example of a thermal power plant.
comment: Conference paper under review
☆ Automated Filtering of Human Feedback Data for Aligning Text-to-Image Diffusion Models
Fine-tuning text-to-image diffusion models with human feedback is an effective method for aligning model behavior with human intentions. However, this alignment process often suffers from slow convergence due to the large size and noise present in human feedback datasets. In this work, we propose FiFA, a novel automated data filtering algorithm designed to enhance the fine-tuning of diffusion models using human feedback datasets with direct preference optimization (DPO). Specifically, our approach selects data by solving an optimization problem to maximize three components: preference margin, text quality, and text diversity. The concept of preference margin is used to identify samples that contain high informational value to address the noisy nature of feedback dataset, which is calculated using a proxy reward model. Additionally, we incorporate text quality, assessed by large language models to prevent harmful contents, and consider text diversity through a k-nearest neighbor entropy estimator to improve generalization. Finally, we integrate all these components into an optimization process, with approximating the solution by assigning importance score to each data pair and selecting the most important ones. As a result, our method efficiently filters data automatically, without the need for manual intervention, and can be applied to any large-scale dataset. Experimental results show that FiFA significantly enhances training stability and achieves better performance, being preferred by humans 17% more, while using less than 0.5% of the full data and thus 1% of the GPU hours compared to utilizing full human feedback datasets.
☆ HSR-Enhanced Sparse Attention Acceleration
Large Language Models (LLMs) have demonstrated remarkable capabilities across various applications, but their performance on long-context tasks is often limited by the computational complexity of attention mechanisms. This paper introduces a novel approach to accelerate attention computation in LLMs, particularly for long-context scenarios. We leverage the inherent sparsity within attention mechanisms, both in conventional Softmax attention and ReLU attention (with $\mathsf{ReLU}^\alpha$ activation, $\alpha \in \mathbb{N}_+$), to significantly reduce the running time complexity. Our method employs a Half-Space Reporting (HSR) data structure to rapidly identify non-zero or "massively activated" entries in the attention matrix. We present theoretical analyses for two key scenarios: attention generation and full attention computation with long input context. Our approach achieves a running time of $O(mn^{4/5})$ significantly faster than the naive approach $O(mn)$ for attention generation, where $n$ is the context length, $m$ is the query length, and $d$ is the hidden dimension. We can also reduce the running time of full attention computation from $O(mn)$ to $O(mn^{1 - 1 / \lfloor d/2\rfloor} + mn^{4/5})$. Importantly, our method introduces no error for ReLU attention and only provably negligible error for Softmax attention, where the latter is supported by our empirical validation. This work represents a significant step towards enabling efficient long-context processing in LLMs, potentially broadening their applicability across various domains.
☆ Improved Regret Bound for Safe Reinforcement Learning via Tighter Cost Pessimism and Reward Optimism
This paper studies the safe reinforcement learning problem formulated as an episodic finite-horizon tabular constrained Markov decision process with an unknown transition kernel and stochastic reward and cost functions. We propose a model-based algorithm based on novel cost and reward function estimators that provide tighter cost pessimism and reward optimism. While guaranteeing no constraint violation in every episode, our algorithm achieves a regret upper bound of $\widetilde{\mathcal{O}}((\bar C - \bar C_b)^{-1}H^{2.5} S\sqrt{AK})$ where $\bar C$ is the cost budget for an episode, $\bar C_b$ is the expected cost under a safe baseline policy over an episode, $H$ is the horizon, and $S$, $A$ and $K$ are the number of states, actions, and episodes, respectively. This improves upon the best-known regret upper bound, and when $\bar C- \bar C_b=\Omega(H)$, it nearly matches the regret lower bound of $\Omega(H^{1.5}\sqrt{SAK})$. We deduce our cost and reward function estimators via a Bellman-type law of total variance to obtain tight bounds on the expected sum of the variances of value function estimates. This leads to a tighter dependence on the horizon in the function estimators. We also present numerical results to demonstrate the computational effectiveness of our proposed framework.
☆ Tracing Human Stress from Physiological Signals using UWB Radar
Stress tracing is an important research domain that supports many applications, such as health care and stress management; and its closest related works are derived from stress detection. However, these existing works cannot well address two important challenges facing stress detection. First, most of these studies involve asking users to wear physiological sensors to detect their stress states, which has a negative impact on the user experience. Second, these studies have failed to effectively utilize multimodal physiological signals, which results in less satisfactory detection results. This paper formally defines the stress tracing problem, which emphasizes the continuous detection of human stress states. A novel deep stress tracing method, named DST, is presented. Note that DST proposes tracing human stress based on physiological signals collected by a noncontact ultrawideband radar, which is more friendly to users when collecting their physiological signals. In DST, a signal extraction module is carefully designed at first to robustly extract multimodal physiological signals from the raw RF data of the radar, even in the presence of body movement. Afterward, a multimodal fusion module is proposed in DST to ensure that the extracted multimodal physiological signals can be effectively fused and utilized. Extensive experiments are conducted on three real-world datasets, including one self-collected dataset and two publicity datasets. Experimental results show that the proposed DST method significantly outperforms all the baselines in terms of tracing human stress states. On average, DST averagely provides a 6.31% increase in detection accuracy on all datasets, compared with the best baselines.
comment: 19 pages, 11 figures
☆ Diagnosing Hate Speech Classification: Where Do Humans and Machines Disagree, and Why?
This study uses the cosine similarity ratio, embedding regression, and manual re-annotation to diagnose hate speech classification. We begin by computing cosine similarity ratio on a dataset "Measuring Hate Speech" that contains 135,556 annotated comments on social media. This way, we show a basic use of cosine similarity as a description of hate speech content. We then diagnose hate speech classification starting from understanding the inconsistency of human annotation from the dataset. Using embedding regression as a basic diagnostic, we found that female annotators are more sensitive to racial slurs that target the black population. We perform with a more complicated diagnostic by training a hate speech classifier using a SoTA pre-trained large language model, NV-Embed-v2, to convert texts to embeddings and run a logistic regression. This classifier achieves a testing accuracy of 94%. In diagnosing where machines disagree with human annotators, we found that machines make fewer mistakes than humans despite the fact that human annotations are treated as ground truth in the training set. Machines perform better in correctly labeling long statements of facts, but perform worse in labeling short instances of swear words. We hypothesize that this is due to model alignment - while curating models at their creation prevents the models from producing obvious hate speech, it also reduces the model's ability to detect such content.
♻ ☆ Tighter Privacy Auditing of DP-SGD in the Hidden State Threat Model
Machine learning models can be trained with formal privacy guarantees via differentially private optimizers such as DP-SGD. In this work, we focus on a threat model where the adversary has access only to the final model, with no visibility into intermediate updates. In the literature, this hidden state threat model exhibits a significant gap between the lower bound from empirical privacy auditing and the theoretical upper bound provided by privacy accounting. To challenge this gap, we propose to audit this threat model with adversaries that \emph{craft a gradient sequence} designed to maximize the privacy loss of the final model without relying on intermediate updates. Our experiments show that this approach consistently outperforms previous attempts at auditing the hidden state model. Furthermore, our results advance the understanding of achievable privacy guarantees within this threat model. Specifically, when the crafted gradient is inserted at every optimization step, we show that concealing the intermediate model updates in DP-SGD does not amplify privacy. The situation is more complex when the crafted gradient is not inserted at every step: our auditing lower bound matches the privacy upper bound only for an adversarially-chosen loss landscape and a sufficiently large batch size. This suggests that existing privacy upper bounds can be improved in certain regimes.
♻ ☆ Towards Generalist Robot Learning from Internet Video: A Survey
Scaling deep learning to huge internet-scraped datasets has yielded remarkably general capabilities in natural language processing and visual understanding and generation. In contrast, data is scarce and expensive to collect in robotics. This has seen robot learning struggle to match the generality of capabilities observed in other domains. Learning from Videos (LfV) methods seek to address this data bottleneck by augmenting traditional robot data with large internet-scraped video datasets. Such video data may provide the model with foundational information regarding physical behaviours and the physics of the world. This holds great promise for improving the generality of our robots. In this survey, we present an overview of the emerging field of LfV. We outline fundamental concepts, including the benefits and challenges of LfV. We provide a comprehensive review of current methods for: extracting knowledge from large-scale internet video; tackling key LfV challenges; and boosting downstream reinforcement and robot learning via the use of video data. LfV datasets and benchmarks are also reviewed. The survey closes with a critical discussion of challenges and opportunities. Here, we advocate for scalable foundation model approaches that can leverage the full range of available internet video to aid the learning of robot policies and dynamics models. We hope this survey can inform and catalyse further LfV research, facilitating progress towards the development of general-purpose robots.
comment: Refactored paper structure, significantly reduced paper length, rewritten abstract and introduction. Other minor improvements
♻ ☆ AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents
The robustness of LLMs to jailbreak attacks, where users design prompts to circumvent safety measures and misuse model capabilities, has been studied primarily for LLMs acting as simple chatbots. Meanwhile, LLM agents -- which use external tools and can execute multi-stage tasks -- may pose a greater risk if misused, but their robustness remains underexplored. To facilitate research on LLM agent misuse, we propose a new benchmark called AgentHarm. The benchmark includes a diverse set of 110 explicitly malicious agent tasks (440 with augmentations), covering 11 harm categories including fraud, cybercrime, and harassment. In addition to measuring whether models refuse harmful agentic requests, scoring well on AgentHarm requires jailbroken agents to maintain their capabilities following an attack to complete a multi-step task. We evaluate a range of leading LLMs, and find (1) leading LLMs are surprisingly compliant with malicious agent requests without jailbreaking, (2) simple universal jailbreak templates can be adapted to effectively jailbreak agents, and (3) these jailbreaks enable coherent and malicious multi-step agent behavior and retain model capabilities. To enable simple and reliable evaluation of attacks and defenses for LLM-based agents, we publicly release AgentHarm at https://huggingface.co/datasets/ai-safety-institute/AgentHarm.
♻ ☆ Reducing the Barriers to Entry for Foundation Model Training
The world has recently witnessed an unprecedented acceleration in demands for Machine Learning and Artificial Intelligence applications. This spike in demand has imposed tremendous strain on the underlying technology stack in supply chain, GPU-accelerated hardware, software, datacenter power density, and energy consumption. If left on the current technological trajectory, future demands show insurmountable spending trends, further limiting market players, stifling innovation, and widening the technology gap. To address these challenges, we propose a fundamental change in the AI training infrastructure throughout the technology ecosystem. The changes require advancements in supercomputing and novel AI training approaches, from high-end software to low-level hardware, microprocessor, and chip design, while advancing the energy efficiency required by a sustainable infrastructure. This paper presents the analytical framework that quantitatively highlights the challenges and points to the opportunities to reduce the barriers to entry for training large language models.
♻ ☆ The Future of Large Language Model Pre-training is Federated
Generative pre-trained large language models (LLMs) have demonstrated impressive performance over a wide range of tasks, thanks to the unprecedented amount of data they have been trained on. As established scaling laws indicate, LLMs' future performance improvement depends on the amount of computing and data sources they can leverage for pre-training. Federated learning (FL) has the potential to unleash the majority of the planet's data and computational resources, which are underutilized by the data-center-focused training methodology of current LLM practice. Our work presents a robust, flexible, reproducible FL approach that enables large-scale collaboration across institutions to train LLMs. We propose a scalable deployment system called Photon to enable the investigation and development of this new training paradigm for LLM pre-training. We show that Photon can be used by organizations interested in collaborating with their private data sources and computational resources for pre-training LLMs with billions of parameters. This paradigm would mobilize more computational and data resources while matching or potentially exceeding centralized performance. We further show the effectiveness of the federated training scales with model size and present our approach for training billion-scale federated LLMs using limited resources. Thus far, we have used Photon to train LLM models to the size of 7B parameters and anticipate larger models being completed in the near future. Finally, we show that LLM training is highly resilient to the classical challenges of federated statistical and hardware heterogeneity. Furthermore, we show that convergence is robust to partial participation, opening the avenue for compute-efficient collaborative training. Photon will help data-rich actors to become the protagonists of LLMs pre-training instead of leaving the stage to compute-rich actors alone.
comment: 24 pages, 15 figures, pre-print
♻ ☆ C-Learner: Constrained Learning for Causal Inference and Semiparametric Statistics
Popular debiased causal estimation methods, e.g. for the average treatment effect -- such as one-step estimation (e.g., augmented inverse propensity weighting) and targeted maximum likelihood estimation -- enjoy desirable asymptotic properties such as statistical efficiency and double robustness. However, they often produce unstable estimates when there is limited overlap between treatment and control, and require ad hoc adjustments in practice (e.g., truncating propensity scores). In contrast, simple plug-in estimators are stable but lack good asymptotic properties. We propose a novel debiased estimator that achieves the best of both worlds, producing stable plug-in estimates with desirable asymptotic properties. Our constrained learning framework solves for the best plug-in estimator under the constraint that the first-order error with respect to the plugged-in quantity is zero, and can leverage flexible model classes including neural networks and tree ensembles. In several experimental settings, including ones in which we handle text-based covariates by fine-tuning language models, our constrained learning-based estimator outperforms one-step estimation and targeting in challenging settings with limited overlap between treatment and control, and performs comparably otherwise.
♻ ☆ Shadowcast: Stealthy Data Poisoning Attacks Against Vision-Language Models
Vision-Language Models (VLMs) excel in generating textual responses from visual inputs, but their versatility raises security concerns. This study takes the first step in exposing VLMs' susceptibility to data poisoning attacks that can manipulate responses to innocuous, everyday prompts. We introduce Shadowcast, a stealthy data poisoning attack where poison samples are visually indistinguishable from benign images with matching texts. Shadowcast demonstrates effectiveness in two attack types. The first is a traditional Label Attack, tricking VLMs into misidentifying class labels, such as confusing Donald Trump for Joe Biden. The second is a novel Persuasion Attack, leveraging VLMs' text generation capabilities to craft persuasive and seemingly rational narratives for misinformation, such as portraying junk food as healthy. We show that Shadowcast effectively achieves the attacker's intentions using as few as 50 poison samples. Crucially, the poisoned samples demonstrate transferability across different VLM architectures, posing a significant concern in black-box settings. Moreover, Shadowcast remains potent under realistic conditions involving various text prompts, training data augmentation, and image compression techniques. This work reveals how poisoned VLMs can disseminate convincing yet deceptive misinformation to everyday, benign users, emphasizing the importance of data integrity for responsible VLM deployments. Our code is available at: https://github.com/umd-huang-lab/VLM-Poisoning.
comment: Accepted by Thirty-Eighth Annual Conference on Neural Information Processing Systems (Neurips 2024)
♻ ☆ Compressed Federated Reinforcement Learning with a Generative Model ECML-PKDD 2024
Reinforcement learning has recently gained unprecedented popularity, yet it still grapples with sample inefficiency. Addressing this challenge, federated reinforcement learning (FedRL) has emerged, wherein agents collaboratively learn a single policy by aggregating local estimations. However, this aggregation step incurs significant communication costs. In this paper, we propose CompFedRL, a communication-efficient FedRL approach incorporating both \textit{periodic aggregation} and (direct/error-feedback) compression mechanisms. Specifically, we consider compressed federated $Q$-learning with a generative model setup, where a central server learns an optimal $Q$-function by periodically aggregating compressed $Q$-estimates from local agents. For the first time, we characterize the impact of these two mechanisms (which have remained elusive) by providing a finite-time analysis of our algorithm, demonstrating strong convergence behaviors when utilizing either direct or error-feedback compression. Our bounds indicate improved solution accuracy concerning the number of agents and other federated hyperparameters while simultaneously reducing communication costs. To corroborate our theory, we also conduct in-depth numerical experiments to verify our findings, considering Top-$K$ and Sparsified-$K$ sparsification operators.
comment: European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML-PKDD 2024)
♻ ☆ Extracting Training Data from Unconditional Diffusion Models
As diffusion probabilistic models (DPMs) are being employed as mainstream models for generative artificial intelligence (AI), the study of their memorization of the raw training data has attracted growing attention. Existing works in this direction aim to establish an understanding of whether or to what extent DPMs learn by memorization. Such an understanding is crucial for identifying potential risks of data leakage and copyright infringement in diffusion models and, more importantly, for more controllable generation and trustworthy application of Artificial Intelligence Generated Content (AIGC). While previous works have made important observations of when DPMs are prone to memorization, these findings are mostly empirical, and the developed data extraction methods only work for conditional diffusion models. In this work, we aim to establish a theoretical understanding of memorization in DPMs with 1) a memorization metric for theoretical analysis, 2) an analysis of conditional memorization with informative and random labels, and 3) two better evaluation metrics for measuring memorization. Based on the theoretical analysis, we further propose a novel data extraction method called \textbf{Surrogate condItional Data Extraction (SIDE)} that leverages a classifier trained on generated data as a surrogate condition to extract training data directly from unconditional diffusion models. Our empirical results demonstrate that SIDE can extract training data from diffusion models where previous methods fail, and it is on average over 50\% more effective across different scales of the CelebA dataset.
comment: This is an old version of arXiv:2410.02467. Please refer to the new version
♻ ☆ Feudal Graph Reinforcement Learning
Graph-based representations and message-passing modular policies constitute prominent approaches to tackling composable control problems in reinforcement learning (RL). However, as shown by recent graph deep learning literature, such local message-passing operators can create information bottlenecks and hinder global coordination. The issue becomes more serious in tasks requiring high-level planning. In this work, we propose a novel methodology, named Feudal Graph Reinforcement Learning (FGRL), that addresses such challenges by relying on hierarchical RL and a pyramidal message-passing architecture. In particular, FGRL defines a hierarchy of policies where high-level commands are propagated from the top of the hierarchy down through a layered graph structure. The bottom layers mimic the morphology of the physical system, while the upper layers correspond to higher-order sub-modules. The resulting agents are then characterized by a committee of policies where actions at a certain level set goals for the level below, thus implementing a hierarchical decision-making structure that can naturally implement task decomposition. We evaluate the proposed framework on a graph clustering problem and MuJoCo locomotion tasks; simulation results show that FGRL compares favorably against relevant baselines. Furthermore, an in-depth analysis of the command propagation mechanism provides evidence that the introduced message-passing scheme favors learning hierarchical decision-making policies.
♻ ☆ Neural Networks with LSTM and GRU in Modeling Active Fires in the Amazon
This study presents a comprehensive methodology for modeling and forecasting the historical time series of active fire spots detected by the AQUA\_M-T satellite in the Amazon, Brazil. The approach employs a mixed Recurrent Neural Network (RNN) model, combining Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) architectures to predict the monthly accumulations of daily detected active fire spots. Data analysis revealed a consistent seasonality over time, with annual maximum and minimum values tending to repeat at the same periods each year. The primary objective is to verify whether the forecasts capture this inherent seasonality through machine learning techniques. The methodology involved careful data preparation, model configuration, and training using cross-validation with two seeds, ensuring that the data generalizes well to both the test and validation sets for both seeds. The results indicate that the combined LSTM and GRU model delivers excellent forecasting performance, demonstrating its effectiveness in capturing complex temporal patterns and modeling the observed time series. This research significantly contributes to the application of deep learning techniques in environmental monitoring, specifically in forecasting active fire spots. The proposed approach highlights the potential for adaptation to other time series forecasting challenges, opening new opportunities for research and development in machine learning and prediction of natural phenomena. Keywords: Time Series Forecasting; Recurrent Neural Networks; Deep Learning.
comment: 16 pages and 24 figures, in Portuguese language
♻ ☆ Differentially Private Attention Computation
Large language models (LLMs), especially those based on the Transformer architecture, have had a profound impact on various aspects of daily life, such as natural language processing, content generation, research methodologies, and more. Nevertheless, a crucial concern regarding the inference results of large language models is the issue of security and privacy. Given that large language models can generate results that may leak sensitive confidential or copyright information in many scenarios, it is crucial to compute the attention matrix with provable privacy guarantees, as attention is all you need. In this work, we propose a novel and efficient algorithm for approximating the attention matrix while providing differential privacy (DP) guarantees. To achieve this, we build on recent advancements in fast attention computation and differentially private matrix publishing.
♻ ☆ Fair-OBNC: Correcting Label Noise for Fairer Datasets
Data used by automated decision-making systems, such as Machine Learning models, often reflects discriminatory behavior that occurred in the past. These biases in the training data are sometimes related to label noise, such as in COMPAS, where more African-American offenders are wrongly labeled as having a higher risk of recidivism when compared to their White counterparts. Models trained on such biased data may perpetuate or even aggravate the biases with respect to sensitive information, such as gender, race, or age. However, while multiple label noise correction approaches are available in the literature, these focus on model performance exclusively. In this work, we propose Fair-OBNC, a label noise correction method with fairness considerations, to produce training datasets with measurable demographic parity. The presented method adapts Ordering-Based Noise Correction, with an adjusted criterion of ordering, based both on the margin of error of an ensemble, and the potential increase in the observed demographic parity of the dataset. We evaluate Fair-OBNC against other different pre-processing techniques, under different scenarios of controlled label noise. Our results show that the proposed method is the overall better alternative within the pool of label correction methods, being capable of attaining better reconstructions of the original labels. Models trained in the corrected data have an increase, on average, of 150% in demographic parity, when compared to models trained in data with noisy labels, across the considered levels of label noise.
♻ ☆ AR-TTA: A Simple Method for Real-World Continual Test-Time Adaptation BMVC2024
Test-time adaptation is a promising research direction that allows the source model to adapt itself to changes in data distribution without any supervision. Yet, current methods are usually evaluated on benchmarks that are only a simplification of real-world scenarios. Hence, we propose to validate test-time adaptation methods using the recently introduced datasets for autonomous driving, namely CLAD-C and SHIFT. We observe that current test-time adaptation methods struggle to effectively handle varying degrees of domain shift, often resulting in degraded performance that falls below that of the source model. We noticed that the root of the problem lies in the inability to preserve the knowledge of the source model and adapt to dynamically changing, temporally correlated data streams. Therefore, we enhance the well-established self-training framework by incorporating a small memory buffer to increase model stability and at the same time perform dynamic adaptation based on the intensity of domain shift. The proposed method, named AR-TTA, outperforms existing approaches on both synthetic and more real-world benchmarks and shows robustness across a variety of TTA scenarios. The code is available at https://github.com/dmn-sjk/AR-TTA.
comment: Accepted for BMVC2024
♻ ☆ Flexible Heteroscedastic Count Regression with Deep Double Poisson Networks
Neural networks that can produce accurate, input-conditional uncertainty representations are critical for real-world applications. Recent progress on heteroscedastic continuous regression has shown great promise for calibrated uncertainty quantification on complex tasks, like image regression. However, when these methods are applied to discrete regression tasks, such as crowd counting, ratings prediction, or inventory estimation, they tend to produce predictive distributions with numerous pathologies. Moreover, discrete models based on the Generalized Linear Model (GLM) framework either cannot process complex input or are not fully heterosedastic. To address these issues we propose the Deep Double Poisson Network (DDPN). In contrast to networks trained to minimize Gaussian negative log likelihood (NLL), discrete network parameterizations (i.e., Poisson, Negative binomial), and GLMs, DDPN can produce discrete predictive distributions of arbitrary flexibility. Additionally, we propose a technique to tune the prioritization of mean fit and probabilistic calibration during training. We show DDPN 1) vastly outperforms existing discrete models; 2) meets or exceeds the accuracy and flexibility of networks trained with Gaussian NLL; 3) produces proper predictive distributions over discrete counts; and 4) exhibits superior out-of-distribution detection. DDPN can easily be applied to a variety of count regression datasets including tabular, image, point cloud, and text data.
♻ ☆ Fairness in KI-Systemen
The more AI-assisted decisions affect people's lives, the more important the fairness of such decisions becomes. In this chapter, we provide an introduction to research on fairness in machine learning. We explain the main fairness definitions and strategies for achieving fairness using concrete examples and place fairness research in the European context. Our contribution is aimed at an interdisciplinary audience and therefore avoids mathematical formulation but emphasizes visualizations and examples. -- Je mehr KI-gest\"utzte Entscheidungen das Leben von Menschen betreffen, desto wichtiger ist die Fairness solcher Entscheidungen. In diesem Kapitel geben wir eine Einf\"uhrung in die Forschung zu Fairness im maschinellen Lernen. Wir erkl\"aren die wesentlichen Fairness-Definitionen und Strategien zur Erreichung von Fairness anhand konkreter Beispiele und ordnen die Fairness-Forschung in den europ\"aischen Kontext ein. Unser Beitrag richtet sich dabei an ein interdisziplin\"ares Publikum und verzichtet daher auf die mathematische Formulierung sondern betont Visualisierungen und Beispiele.
comment: in German language, originally published in: Sabrina Schork (editor): "Vertrauen in k\"unstliche Intelligenz", Springer Fachmedien Wiesbaden GmbH
♻ ☆ Synthetic Potential Outcomes and Causal Mixture Identifiability
A mixture model consists of a latent class that exerts a discrete signal on the observed data. Uncovering these latent classes is fundamental to unsupervised learning. In this paper, we consider the problem of recovering latent classes defined with respect to causal responses. We allow overlapping support in the distributions of these classes, meaning individuals cannot be clustered into groups with a similar response. Instead, we build on a setting from proximal causal inference to develop a method of moments approach to synthetically sample potential outcome distributions. This approach is the first known identifiability result for what we call Mixtures of Treatment Effects (MTEs). More broadly, we show how MTEs fit into a hierarchy of causal identifiability that unifies a number of perspectives on latent class confounding.
♻ ☆ Recurrent Action Transformer with Memory
Recently, the use of transformers in offline reinforcement learning has become a rapidly developing area. This is due to their ability to treat the agent's trajectory in the environment as a sequence, thereby reducing the policy learning problem to sequence modeling. In environments where the agent's decisions depend on past events (POMDPs), it is essential to capture both the event itself and the decision point in the context of the model. However, the quadratic complexity of the attention mechanism limits the potential for context expansion. One solution to this problem is to extend transformers with memory mechanisms. This paper proposes a Recurrent Action Transformer with Memory (RATE), a novel model architecture that incorporates a recurrent memory mechanism designed to regulate information retention. To evaluate our model, we conducted extensive experiments on memory-intensive environments (ViZDoom-Two-Colors, T-Maze, Memory Maze, Minigrid-Memory), classic Atari games, and MuJoCo control environments. The results show that using memory can significantly improve performance in memory-intensive environments, while maintaining or improving results in classic environments. We believe that our results will stimulate research on memory mechanisms for transformers applicable to offline reinforcement learning.
comment: 24 pages, 14 figures
♻ ☆ BA-LoRA: Bias-Alleviating Low-Rank Adaptation to Mitigate Catastrophic Inheritance in Large Language Models
Large language models (LLMs) have demonstrated remarkable proficiency across various natural language processing (NLP) tasks. However, adapting LLMs to downstream applications requires computationally intensive and memory-demanding fine-tuning procedures. To alleviate these burdens, parameter-efficient fine-tuning (PEFT) techniques have emerged as a promising approach to tailor LLMs with minimal computational overhead. While PEFT methods offer substantial advantages, they do not fully address the pervasive issue of bias propagation from pre-training data. This work introduces Bias-Alleviating Low-Rank Adaptation (BA-LoRA), a novel PEFT method designed to counteract bias inheritance. BA-LoRA incorporates three distinct regularization terms: (1) a consistency regularizer, (2) a diversity regularizer, and (3) a singular value decomposition regularizer. These regularizers aim to enhance the models' consistency, diversity, and generalization capabilities during fine-tuning. We conduct extensive experiments on natural language understanding (NLU) and natural language generation (NLG) tasks using prominent LLMs such as LLaMA, Mistral, and Gemma. The results demonstrate that BA-LoRA outperforms LoRA and its state-of-the-art variants. Moreover, our method effectively mitigates the adverse effects of pre-training bias, leading to more reliable and robust model outputs. The code is available at https://github.com/cyp-jlu-ai/BA-LoRA.
comment: 23 pages
♻ ☆ A Synthesis of Green Architectural Tactics for ML-Enabled Systems ICSE
The rapid adoption of artificial intelligence (AI) and machine learning (ML) has generated growing interest in understanding their environmental impact and the challenges associated with designing environmentally friendly ML-enabled systems. While Green AI research, i.e., research that tries to minimize the energy footprint of AI, is receiving increasing attention, very few concrete guidelines are available on how ML-enabled systems can be designed to be more environmentally sustainable. In this paper, we provide a catalog of 30 green architectural tactics for ML-enabled systems to fill this gap. An architectural tactic is a high-level design technique to improve software quality, in our case environmental sustainability. We derived the tactics from the analysis of 51 peer-reviewed publications that primarily explore Green AI, and validated them using a focus group approach with three experts. The 30 tactics we identified are aimed to serve as an initial reference guide for further exploration into Green AI from a software engineering perspective, and assist in designing sustainable ML-enabled systems. To enhance transparency and facilitate their widespread use and extension, we make the tactics available online in easily consumable formats. Wide-spread adoption of these tactics has the potential to substantially reduce the societal impact of ML-enabled systems regarding their energy and carbon footprint.
comment: Accepted for publication at the 2024 International Conference on Software Engineering - Software Engineering in Society (ICSE-SEIS'2024); *Caveat:* the published version (https://doi.org/10.1145/3639475.3640111) contains a mistake in Section 4.2 for tactic T6 (incorrect comparison of energy consumption of kNN and Random Forest); this mistake is corrected in this postprint
♻ ☆ Geometry-Informed Neural Networks
Geometry is a ubiquitous tool in computer graphics, design, and engineering. However, the lack of large shape datasets limits the application of state-of-the-art supervised learning methods and motivates the exploration of alternative learning strategies. To this end, we introduce geometry-informed neural networks (GINNs) -- a framework for training shape-generative neural fields without data by leveraging user-specified design requirements in the form of objectives and constraints. By adding diversity as an explicit constraint, GINNs avoid mode-collapse and can generate multiple diverse solutions, often required in geometry tasks. Experimentally, we apply GINNs to several validation problems and a realistic 3D engineering design problem, showing control over geometrical and topological properties, such as surface smoothness or the number of holes. These results demonstrate the potential of training shape-generative models without data, paving the way for new generative design approaches without large datasets.
♻ ☆ Barlow Twins Deep Neural Network for Advanced 1D Drug-Target Interaction Prediction
Accurate prediction of drug-target interactions is critical for advancing drug discovery. By reducing time and cost, machine learning and deep learning can accelerate this laborious discovery process. In a novel approach, BarlowDTI, we utilise the powerful Barlow Twins architecture for feature-extraction while considering the structure of the target protein. Our method achieves state-of-the-art predictive performance against multiple established benchmarks using only one-dimensional input. The use of gradient boosting machine as the underlying predictor ensures fast and efficient predictions without the need for substantial computational resources. We also investigate how the model reaches its decision based on individual training samples. By comparing co-crystal structures, we find that BarlowDTI effectively exploits catalytically active and stabilising residues, highlighting the model's ability to generalise from one-dimensional input data. In addition, we further benchmark new baselines against existing methods. Together, these innovations improve the efficiency and effectiveness of drug-target interaction predictions, providing robust tools for accelerating drug development and deepening the understanding of molecular interactions. Therefore, we provide an easy-to-use web interface that can be freely accessed at https://www.bio.nat.tum.de/oc2/barlowdti .
comment: Refined model architecture, additional results added
♻ ☆ Pure Message Passing Can Estimate Common Neighbor for Link Prediction
Message Passing Neural Networks (MPNNs) have emerged as the {\em de facto} standard in graph representation learning. However, when it comes to link prediction, they often struggle, surpassed by simple heuristics such as Common Neighbor (CN). This discrepancy stems from a fundamental limitation: while MPNNs excel in node-level representation, they stumble with encoding the joint structural features essential to link prediction, like CN. To bridge this gap, we posit that, by harnessing the orthogonality of input vectors, pure message-passing can indeed capture joint structural features. Specifically, we study the proficiency of MPNNs in approximating CN heuristics. Based on our findings, we introduce the Message Passing Link Predictor (MPLP), a novel link prediction model. MPLP taps into quasi-orthogonal vectors to estimate link-level structural features, all while preserving the node-level complexities. Moreover, our approach demonstrates that leveraging message-passing to capture structural features could offset MPNNs' expressiveness limitations at the expense of estimation variance. We conduct experiments on benchmark datasets from various domains, where our method consistently outperforms the baseline methods.
comment: Accepted to Neurips'24
♻ ☆ eXponential FAmily Dynamical Systems (XFADS): Large-scale nonlinear Gaussian state-space modeling
State-space graphical models and the variational autoencoder framework provide a principled apparatus for learning dynamical systems from data. State-of-the-art probabilistic approaches are often able to scale to large problems at the cost of flexibility of the variational posterior or expressivity of the dynamics model. However, those consolidations can be detrimental if the ultimate goal is to learn a generative model capable of explaining the spatiotemporal structure of the data and making accurate forecasts. We introduce a low-rank structured variational autoencoding framework for nonlinear Gaussian state-space graphical models capable of capturing dense covariance structures that are important for learning dynamical systems with predictive capabilities. Our inference algorithm exploits the covariance structures that arise naturally from sample based approximate Gaussian message passing and low-rank amortized posterior updates -- effectively performing approximate variational smoothing with time complexity scaling linearly in the state dimensionality. In comparisons with other deep state-space model architectures our approach consistently demonstrates the ability to learn a more predictive generative model. Furthermore, when applied to neural physiological recordings, our approach is able to learn a dynamical system capable of forecasting population spiking and behavioral correlates from a small portion of single trials.
♻ ☆ A Closer Look at Time Steps is Worthy of Triple Speed-Up for Diffusion Model Training
Training diffusion models is always a computation-intensive task. In this paper, we introduce a novel speed-up method for diffusion model training, called, which is based on a closer look at time steps. Our key findings are: i) Time steps can be empirically divided into acceleration, deceleration, and convergence areas based on the process increment. ii) These time steps are imbalanced, with many concentrated in the convergence area. iii) The concentrated steps provide limited benefits for diffusion training. To address this, we design an asymmetric sampling strategy that reduces the frequency of steps from the convergence area while increasing the sampling probability for steps from other areas. Additionally, we propose a weighting strategy to emphasize the importance of time steps with rapid-change process increments. As a plug-and-play and architecture-agnostic approach, SpeeD consistently achieves 3-times acceleration across various diffusion architectures, datasets, and tasks. Notably, due to its simple design, our approach significantly reduces the cost of diffusion model training with minimal overhead. Our research enables more researchers to train diffusion models at a lower cost.
♻ ☆ Deep Reinforcement Learning for Traveling Purchaser Problems
The traveling purchaser problem (TPP) is an important combinatorial optimization problem with broad applications. Due to the coupling between routing and purchasing, existing works on TPPs commonly address route construction and purchase planning simultaneously, which, however, leads to exact methods with high computational cost and heuristics with sophisticated design but limited performance. In sharp contrast, we propose a novel approach based on deep reinforcement learning (DRL), which addresses route construction and purchase planning separately, while evaluating and optimizing the solution from a global perspective. The key components of our approach include a bipartite graph representation for TPPs to capture the market-product relations, and a policy network that extracts information from the bipartite graph and uses it to sequentially construct the route. One significant benefit of our framework is that we can efficiently construct the route using the policy network, and once the route is determined, the associated purchasing plan can be easily derived through linear programming, while, leveraging DRL, we can train the policy network to optimize the global solution objective. Furthermore, by introducing a meta-learning strategy, the policy network can be trained stably on large-sized TPP instances, and generalize well across instances of varying sizes and distributions, even to much larger instances that are never seen during training. Experiments on various synthetic TPP instances and the TPPLIB benchmark demonstrate that our DRL-based approach can significantly outperform well-established TPP heuristics, reducing the optimality gap by 40%-90%, and also showing an advantage in runtime, especially on large-sized instances.
♻ ☆ TotalVibeSegmentator: Full Torso Segmentation for the NAKO and UK Biobank in Volumetric Interpolated Breath-hold Examination Body Images
Objectives: To present a publicly available torso segmentation network for large epidemiology datasets on volumetric interpolated breath-hold examination (VIBE) images. Materials & Methods: We extracted preliminary segmentations from TotalSegmentator, spine, and body composition networks for VIBE images, then improved them iteratively and retrained a nnUNet network. Using subsets of NAKO (85 subjects) and UK Biobank (16 subjects), we evaluated with Dice-score on a holdout set (12 subjects) and existing organ segmentation approach (1000 subjects), generating 71 semantic segmentation types for VIBE images. We provide an additional network for the vertebra segments 22 individual vertebra types. Results: We achieved an average Dice score of 0.89 +- 0.07 overall 71 segmentation labels. We scored > 0.90 Dice-score on the abdominal organs except for the pancreas with a Dice of 0.70. Conclusion: Our work offers a detailed and refined publicly available full torso segmentation on VIBE images.
comment: https://github.com/robert-graf/TotalVibeSegmentator
♻ ☆ Can we hop in general? A discussion of benchmark selection and design using the Hopper environment
Empirical, benchmark-driven testing is a fundamental paradigm in the current RL community. While using off-the-shelf benchmarks in reinforcement learning (RL) research is a common practice, this choice is rarely discussed. Benchmark choices are often done based on intuitive ideas like "legged robots" or "visual observations". In this paper, we argue that benchmarking in RL needs to be treated as a scientific discipline itself. To illustrate our point, we present a case study on different variants of the Hopper environment to show that the selection of standard benchmarking suites can drastically change how we judge performance of algorithms. The field does not have a cohesive notion of what the different Hopper environments are representative - they do not even seem to be representative of each other. Our experimental results suggests a larger issue in the deep RL literature: benchmark choices are neither commonly justified, nor does there exist a language that could be used to justify the selection of certain environments. This paper concludes with a discussion of the requirements for proper discussion and evaluations of benchmarks and recommends steps to start a dialogue towards this goal.
♻ ☆ Provable Robustness of (Graph) Neural Networks Against Data Poisoning and Backdoor Attacks NeurIPS 2024
Generalization of machine learning models can be severely compromised by data poisoning, where adversarial changes are applied to the training data. This vulnerability has led to interest in certifying (i.e., proving) that such changes up to a certain magnitude do not affect test predictions. We, for the first time, certify Graph Neural Networks (GNNs) against poisoning attacks, including backdoors, targeting the node features of a given graph. Our certificates are white-box and based upon $(i)$ the neural tangent kernel, which characterizes the training dynamics of sufficiently wide networks; and $(ii)$ a novel reformulation of the bilevel optimization problem describing poisoning as a mixed-integer linear program. Consequently, we leverage our framework to provide fundamental insights into the role of graph structure and its connectivity on the worst-case robustness behavior of convolution-based and PageRank-based GNNs. We note that our framework is more general and constitutes the first approach to derive white-box poisoning certificates for NNs, which can be of independent interest beyond graph-related tasks.
comment: A preliminary version of this work appeared at the AdvML-Frontiers @ NeurIPS 2024 workshop
♻ ☆ A policy gradient approach for Finite Horizon Constrained Markov Decision Processes
The infinite horizon setting is widely adopted for problems of reinforcement learning (RL). These invariably result in stationary policies that are optimal. In many situations, finite horizon control problems are of interest and for such problems, the optimal policies are time-varying in general. Another setting that has become popular in recent times is of Constrained Reinforcement Learning, where the agent maximizes its rewards while it also aims to satisfy some given constraint criteria. However, this setting has only been studied in the context of infinite horizon MDPs where stationary policies are optimal. We present an algorithm for constrained RL in the Finite Horizon Setting where the horizon terminates after a fixed (finite) time. We use function approximation in our algorithm which is essential when the state and action spaces are large or continuous and use the policy gradient method to find the optimal policy. The optimal policy that we obtain depends on the stage and so is non-stationary in general. To the best of our knowledge, our paper presents the first policy gradient algorithm for the finite horizon setting with constraints. We show the convergence of our algorithm to a constrained optimal policy. We also compare and analyze the performance of our algorithm through experiments and show that our algorithm performs better than some other well known algorithms.
♻ ☆ HAMMR: HierArchical MultiModal React agents for generic VQA
Combining Large Language Models (LLMs) with external specialized tools (LLMs+tools) is a recent paradigm to solve multimodal tasks such as Visual Question Answering (VQA). While this approach was demonstrated to work well when optimized and evaluated for each individual benchmark, in practice it is crucial for the next generation of real-world AI systems to handle a broad range of multimodal problems. Therefore we pose the VQA problem from a unified perspective and evaluate a single system on a varied suite of VQA tasks including counting, spatial reasoning, OCR-based reasoning, visual pointing, external knowledge, and more. In this setting, we demonstrate that naively applying the LLM+tools approach using the combined set of all tools leads to poor results. This motivates us to introduce HAMMR: HierArchical MultiModal React. We start from a multimodal ReAct-based system and make it hierarchical by enabling our HAMMR agents to call upon other specialized agents. This enhances the compositionality of the LLM+tools approach, which we show to be critical for obtaining high accuracy on generic VQA. Concretely, on our generic VQA suite, HAMMR outperforms the naive LLM+tools approach by 19.5%. Additionally, HAMMR achieves state-of-the-art results on this task, outperforming the generic standalone PaLI-X VQA model by 5.0%.
♻ ☆ A general machine learning model of aluminosilicate melt viscosity and its application to the surface properties of dry lava planets
Ultra-short-period exoplanets like K2-141 b likely have magma oceans on their dayside, which play a critical role in redistributing heat within the planet. This could lead to a warm nightside surface, measurable by the James Webb Space Telescope, offering insights into the planet's structure. Accurate models of properties like viscosity, which can vary by orders of magnitude, are essential for such studies. We present a new model for predicting molten magma viscosity, applicable in diverse scenarios, including magma oceans on lava planets. Using a database of 28,898 viscosity measurements on phospho-alumino-silicate melts, spanning superliquidus to undercooled temperatures and pressures up to 30 GPa, we trained a greybox artificial neural network, refined by a Gaussian process. This model achieves high predictive accuracy (RMSE $\approx 0.4 \log_{10}$ Pa$\cdot$s) and can handle compositions from SiO$_2$ to multicomponent magmatic and industrial glasses, accounting for pressure effects up to 30 GPa for compositions such as peridotite. Applying this model, we calculated the viscosity of K2-141 b's magma ocean under different compositions. Phase diagram calculations suggest that the dayside is fully molten, with extreme temperatures primarily controlling viscosity. A tenuous atmosphere (0.1 bar) might exist around a 40{\deg} radius from the substellar point. At higher longitudes, atmospheric pressure drops, and by 90{\deg}, magma viscosity rapidly increases as solidification occurs. The nightside surface is likely solid, but previously estimated surface temperatures above 400 K imply a partly molten mantle, feeding geothermal flux through vertical convection.
comment: 19 pages, 9 figures, 2 tables
♻ ☆ QeMFi: A Multifidelity Dataset of Quantum Chemical Properties of Diverse Molecules
Progress in both Machine Learning (ML) and Quantum Chemistry (QC) methods have resulted in high accuracy ML models for QC properties. Datasets such as MD17 and WS22 have been used to benchmark these models at some level of QC method, or fidelity, which refers to the accuracy of the chosen QC method. Multifidelity ML (MFML) methods, where models are trained on data from more than one fidelity, have shown to be effective over single fidelity methods. Much research is progressing in this direction for diverse applications ranging from energy band gaps to excitation energies. One hurdle for effective research here is the lack of a diverse multifidelity dataset for benchmarking. We provide the Quantum chemistry MultiFidelity (QeMFi) dataset consisting of five fidelities calculated with the TD-DFT formalism. The fidelities differ in their basis set choice: STO-3G, 3-21G, 6-31G, def2-SVP, and def2-TZVP. QeMFi offers to the community a variety of QC properties such as vertical excitation properties and molecular dipole moments, further including QC computation times allowing for a time benefit benchmark of multifidelity models for ML-QC.
comment: Updated dataset name, corrected certain figures, and added more complete literature review. Results remain unchanged
♻ ☆ Temporal Action Localization for Inertial-based Human Activity Recognition
As of today, state-of-the-art activity recognition from wearable sensors relies on algorithms being trained to classify fixed windows of data. In contrast, video-based Human Activity Recognition, known as Temporal Action Localization (TAL), has followed a segment-based prediction approach, localizing activity segments in a timeline of arbitrary length. This paper is the first to systematically demonstrate the applicability of state-of-the-art TAL models for both offline and near-online Human Activity Recognition (HAR) using raw inertial data as well as pre-extracted latent features as input. Offline prediction results show that TAL models are able to outperform popular inertial models on a multitude of HAR benchmark datasets, with improvements reaching as much as 26% in F1-score. We show that by analyzing timelines as a whole, TAL models can produce more coherent segments and achieve higher NULL-class accuracy across all datasets. We demonstrate that TAL is less suited for the immediate classification of small-sized windows of data, yet offers an interesting perspective on inertial-based HAR -- alleviating the need for fixed-size windows and enabling algorithms to recognize activities of arbitrary length. With design choices and training concepts yet to be explored, we argue that TAL architectures could be of significant value to the inertial-based HAR community. The code and data download to reproduce experiments is publicly available via github.com/mariusbock/tal_for_har.
comment: 19 pages, 7 figures, 4 tables
♻ ☆ ConML: A Universal Meta-Learning Framework with Task-Level Contrastive Learning
Meta-learning enables learning systems to adapt quickly to new tasks, similar to humans. To emulate this human-like rapid learning and enhance alignment and discrimination abilities, we propose ConML, a universal meta-learning framework that can be applied to various meta-learning algorithms without relying on specific model architectures nor target models. The core of ConML is task-level contrastive learning, which extends contrastive learning from the representation space in unsupervised learning to the model space in meta-learning. By leveraging task identity as an additional supervision signal during meta-training, we contrast the outputs of the meta-learner in the model space, minimizing inner-task distance (between models trained on different subsets of the same task) and maximizing inter-task distance (between models from different tasks). We demonstrate that ConML integrates seamlessly with optimization-based, metric-based, and amortization-based meta-learning algorithms, as well as in-context learning, resulting in performance improvements across diverse few-shot learning tasks.
♻ ☆ An Adaptive Placement and Parallelism Framework for Accelerating RLHF Training
Recently, ChatGPT or InstructGPT like large language models (LLM) has made a significant impact in the AI world. Many works have attempted to reproduce the complex InstructGPT's training pipeline, namely Reinforcement Learning with Human Feedback (RLHF). However, the mainstream distributed RLHF training methods typically adopt a fixed model placement strategy, referred to as the Co-located strategy. This strategy treats all four interdependent models involved in RLHF as a single entity, distributing them across all devices and applying parallelism techniques designed for a single model, regardless of the workload heterogeneity inherent to each model. As a result, this strategy exacerbates the generation bottlenecks in the RLHF training and degrades the overall training efficiency. To address these issues, we propose a flexible model placement framework that offers two general and agile model placement strategies. The Interleaving strategy helps reduce memory redundancy and communication costs of RLHF training by placing models without dependencies on exclusive devices with careful orchestration. On the other hand, the Disaggregated strategy improves the throughput of model training by separating the training and inference runtime of the RLHF pipeline with additional shadow models. Furthermore, our framework provides a simple user interface and guidelines to easily and flexibly configure these strategies in various training scenarios. Our experiments have shown that our strategy can achieve notable improvements up to 11x, compared to the current state-of-the-art (SOTA) approaches. The results highlight the effectiveness and adaptability of our methods in accelerating the training of distributed RLHF.
♻ ☆ Interpolation, Extrapolation, Hyperpolation: Generalising into new dimensions
This paper introduces the concept of hyperpolation: a way of generalising from a limited set of data points that is a peer to the more familiar concepts of interpolation and extrapolation. Hyperpolation is the task of estimating the value of a function at new locations that lie outside the subspace (or manifold) of the existing data. We shall see that hyperpolation is possible and explore its links to creativity in the arts and sciences. We will also examine the role of hyperpolation in machine learning and suggest that the lack of fundamental creativity in current AI systems is deeply connected to their limited ability to hyperpolate.
comment: 22 pages, 8 figures -- added references to 2 papers that introduce related concepts
♻ ☆ Localized Schrödinger Bridge Sampler
We consider the generative problem of sampling from an unknown distribution for which only a sufficiently large number of training samples are available. In this paper, we build on previous work combining Schr\"odinger bridges and Langevin dynamics. A key bottleneck of this approach is the exponential dependence of the required training samples on the dimension, $d$, of the ambient state space. We propose a localization strategy which exploits conditional independence of conditional expectation values. Localization thus replaces a single high-dimensional Schr\"odinger bridge problem by $d$ low-dimensional Schr\"odinger bridge problems over the available training samples. In this context, a connection to multi-head self attention transformer architectures is established. As for the original Schr\"odinger bridge sampling approach, the localized sampler is stable and geometric ergodic. The sampler also naturally extends to conditional sampling and to Bayesian inference. We demonstrate the performance of our proposed scheme through experiments on a Gaussian problem with increasing dimensions, on a temporal stochastic process, and on a stochastic subgrid-scale parametrization conditional sampling problem.
♻ ☆ Automatic Classification of White Blood Cell Images using Convolutional Neural Network
Human immune system contains white blood cells (WBC) that are good indicator of many diseases like bacterial infections, AIDS, cancer, spleen, etc. White blood cells have been sub classified into four types: monocytes, lymphocytes, eosinophils and neutrophils on the basis of their nucleus, shape and cytoplasm. Traditionally in laboratories, pathologists and hematologists analyze these blood cells through microscope and then classify them manually. This manual process takes more time and increases the chance of human error. Hence, there is a need to automate this process. In this paper, first we have used different CNN pre-train models such as ResNet-50, InceptionV3, VGG16 and MobileNetV2 to automatically classify the white blood cells. These pre-train models are applied on Kaggle dataset of microscopic images. Although we achieved reasonable accuracy ranging between 92 to 95%, still there is need to enhance the performance. Hence, inspired by these architectures, a framework has been proposed to automatically categorize the four kinds of white blood cells with increased accuracy. The aim is to develop a convolution neural network (CNN) based classification system with decent generalization ability. The proposed CNN model has been tested on white blood cells images from Kaggle and LISC datasets. Accuracy achieved is 99.57% and 98.67% for both datasets respectively. Our proposed convolutional neural network-based model provides competitive performance as compared to previous results reported in literature.
♻ ☆ On Giant's Shoulders: Effortless Weak to Strong by Dynamic Logits Fusion NeurIPS 2024
Efficient fine-tuning of large language models for task-specific applications is imperative, yet the vast number of parameters in these models makes their training increasingly challenging. Despite numerous proposals for effective methods, a substantial memory overhead remains for gradient computations during updates. \thm{Can we fine-tune a series of task-specific small models and transfer their knowledge directly to a much larger model without additional training?} In this paper, we explore weak-to-strong specialization using logit arithmetic, facilitating a direct answer to this question. Existing weak-to-strong methods often employ a static knowledge transfer ratio and a single small model for transferring complex knowledge, which leads to suboptimal performance. % To address this, To surmount these limitations, we propose a dynamic logit fusion approach that works with a series of task-specific small models, each specialized in a different task. This method adaptively allocates weights among these models at each decoding step, learning the weights through Kullback-Leibler divergence constrained optimization problems. We conduct extensive experiments across various benchmarks in both single-task and multi-task settings, achieving leading results. By transferring expertise from the 7B model to the 13B model, our method closes the performance gap by 96.4\% in single-task scenarios and by 86.3\% in multi-task scenarios compared to full fine-tuning of the 13B model. Notably, we achieve surpassing performance on unseen tasks. Moreover, we further demonstrate that our method can effortlessly integrate in-context learning for single tasks and task arithmetic for multi-task scenarios.
comment: Accepted by NeurIPS 2024
♻ ☆ The Minimax Rate of HSIC Estimation for Translation-Invariant Kernels NeurIPS 2024
Kernel techniques are among the most influential approaches in data science and statistics. Under mild conditions, the reproducing kernel Hilbert space associated to a kernel is capable of encoding the independence of $M\ge 2$ random variables. Probably the most widespread independence measure relying on kernels is the so-called Hilbert-Schmidt independence criterion (HSIC; also referred to as distance covariance in the statistics literature). Despite various existing HSIC estimators designed since its introduction close to two decades ago, the fundamental question of the rate at which HSIC can be estimated is still open. In this work, we prove that the minimax optimal rate of HSIC estimation on $\mathbb R^d$ for Borel measures containing the Gaussians with continuous bounded translation-invariant characteristic kernels is $\mathcal O\!\left(n^{-1/2}\right)$. Specifically, our result implies the optimality in the minimax sense of many of the most-frequently used estimators (including the U-statistic, the V-statistic, and the Nystr\"om-based one) on $\mathbb R^d$.
comment: Accepted for publication at NeurIPS 2024
♻ ☆ Large Language Model Inference Acceleration: A Comprehensive Hardware Perspective
Large Language Models (LLMs) have demonstrated remarkable capabilities across various fields, from natural language understanding to text generation. Compared to non-generative LLMs like BERT and DeBERTa, generative LLMs like GPT series and Llama series are currently the main focus due to their superior algorithmic performance. The advancements in generative LLMs are closely intertwined with the development of hardware capabilities. Various hardware platforms exhibit distinct hardware characteristics, which can help improve LLM inference performance. Therefore, this paper comprehensively surveys efficient generative LLM inference on different hardware platforms. First, we provide an overview of the algorithm architecture of mainstream generative LLMs and delve into the inference process. Then, we summarize different optimization methods for different platforms such as CPU, GPU, FPGA, ASIC, and PIM/NDP, and provide inference results for generative LLMs. Furthermore, we perform a qualitative and quantitative comparison of inference performance with batch sizes 1 and 8 on different hardware platforms by considering hardware power consumption, absolute inference speed (tokens/s), and energy efficiency (tokens/J). We compare the performance of the same optimization methods across different hardware platforms, the performance across different hardware platforms, and the performance of different methods on the same hardware platform. This provides a systematic and comprehensive summary of existing inference acceleration work by integrating software optimization methods and hardware platforms, which can point to the future trends and potential developments of generative LLMs and hardware technology for edge-side scenarios.
comment: 43 pages, 15 figures
♻ ☆ Forecasting with Hyper-Trees
We introduce the concept of Hyper-Trees and offer a new direction in applying tree-based models to time series data. Unlike conventional applications of decision trees that forecast time series directly, Hyper-Trees are designed to learn the parameters of time series models. Our framework combines the effectiveness of gradient boosted trees on tabular data with the advantages of established time series models, thereby naturally inducing a time series inductive bias to tree models. By relating the parameters of a target time series model to features, Hyper-Trees also address the issue of parameter non-stationarity. To resolve the inherent scaling issue of boosted trees when estimating a large number of target model parameters, we combine decision trees and neural networks within a unified framework. In this novel approach, the trees first generate informative representations from the input features, which a shallow network then maps to the target model parameters. With our research, we aim to explore the effectiveness of Hyper-Trees across various forecasting scenarios and to extend the application of gradient boosted trees outside their conventional use in time series modeling.
comment: Forecasting, Gradient Boosting Machines, Hyper-Networks, Parameter Non-Stationarity, Time Series
♻ ☆ NoiseBench: Benchmarking the Impact of Real Label Noise on Named Entity Recognition EMNLP2024
Available training data for named entity recognition (NER) often contains a significant percentage of incorrect labels for entity types and entity boundaries. Such label noise poses challenges for supervised learning and may significantly deteriorate model quality. To address this, prior work proposed various noise-robust learning approaches capable of learning from data with partially incorrect labels. These approaches are typically evaluated using simulated noise where the labels in a clean dataset are automatically corrupted. However, as we show in this paper, this leads to unrealistic noise that is far easier to handle than real noise caused by human error or semi-automatic annotation. To enable the study of the impact of various types of real noise, we introduce NoiseBench, an NER benchmark consisting of clean training data corrupted with 6 types of real noise, including expert errors, crowdsourcing errors, automatic annotation errors and LLM errors. We present an analysis that shows that real noise is significantly more challenging than simulated noise, and show that current state-of-the-art models for noise-robust learning fall far short of their theoretically achievable upper bound. We release NoiseBench to the research community.
comment: data available at https://github.com/elenamer/NoiseBench; to appear at EMNLP2024 main conference
♻ ☆ Geometry-aware training of factorized layers in tensor Tucker format
Reducing parameter redundancies in neural network architectures is crucial for achieving feasible computational and memory requirements during training and inference phases. Given its easy implementation and flexibility, one promising approach is layer factorization, which reshapes weight tensors into a matrix format and parameterizes them as the product of two small rank matrices. However, this approach typically requires an initial full-model warm-up phase, prior knowledge of a feasible rank, and it is sensitive to parameter initialization. In this work, we introduce a novel approach to train the factors of a Tucker decomposition of the weight tensors. Our training proposal proves to be optimal in locally approximating the original unfactorized dynamics independently of the initialization. Furthermore, the rank of each mode is dynamically updated during training. We provide a theoretical analysis of the algorithm, showing convergence, approximation and local descent guarantees. The method's performance is further illustrated through a variety of experiments, showing remarkable training compression rates and comparable or even better performance than the full baseline and alternative layer factorization strategies.
♻ ☆ JurEE not Judges: safeguarding llm interactions with small, specialised Encoder Ensembles
We introduce JurEE, an ensemble of efficient, encoder-only transformer models designed to strengthen safeguards in AI-User interactions within LLM-based systems. Unlike existing LLM-as-Judge methods, which often struggle with generalization across risk taxonomies and only provide textual outputs, JurEE offers probabilistic risk estimates across a wide range of prevalent risks. Our approach leverages diverse data sources and employs progressive synthetic data generation techniques, including LLM-assisted augmentation, to enhance model robustness and performance. We create an in-house benchmark comprising of other reputable benchmarks such as the OpenAI Moderation Dataset and ToxicChat, where we find JurEE significantly outperforms baseline models, demonstrating superior accuracy, speed, and cost-efficiency. This makes it particularly suitable for applications requiring stringent content moderation, such as customer-facing chatbots. The encoder-ensemble's modular design allows users to set tailored risk thresholds, enhancing its versatility across various safety-related applications. JurEE's collective decision-making process, where each specialized encoder model contributes to the final output, not only improves predictive accuracy but also enhances interpretability. This approach provides a more efficient, performant, and economical alternative to traditional LLMs for large-scale implementations requiring robust content moderation.
♻ ☆ Accelerating Diffusion Transformers with Token-wise Feature Caching
Diffusion transformers have shown significant effectiveness in both image and video synthesis at the expense of huge computation costs. To address this problem, feature caching methods have been introduced to accelerate diffusion transformers by caching the features in previous timesteps and reusing them in the following timesteps. However, previous caching methods ignore that different tokens exhibit different sensitivities to feature caching, and feature caching on some tokens may lead to 10$\times$ more destruction to the overall generation quality compared with other tokens. In this paper, we introduce token-wise feature caching, allowing us to adaptively select the most suitable tokens for caching, and further enable us to apply different caching ratios to neural layers in different types and depths. Extensive experiments on PixArt-$\alpha$, OpenSora, and DiT demonstrate our effectiveness in both image and video generation with no requirements for training. For instance, 2.36$\times$ and 1.93$\times$ acceleration are achieved on OpenSora and PixArt-$\alpha$ with almost no drop in generation quality.
♻ ☆ MARS: A neurosymbolic approach for interpretable drug discovery
Neurosymbolic (NeSy) artificial intelligence describes the combination of logic or rule-based techniques with neural networks. Compared to neural approaches, NeSy methods often possess enhanced interpretability, which is particularly promising for biomedical applications like drug discovery. However, since interpretability is broadly defined, there are no clear guidelines for assessing the biological plausibility of model interpretations. To assess interpretability in the context of drug discovery, we devise a novel prediction task, called drug mechanism-of-action (MoA) deconvolution, with an associated, tailored knowledge graph (KG), MoA-net. We then develop the MoA Retrieval System (MARS), a NeSy approach for drug discovery which leverages logical rules with learned rule weights. Using this interpretable feature alongside domain knowledge, we find that MARS and other NeSy approaches on KGs are susceptible to reasoning shortcuts, in which the prediction of true labels is driven by "degree-bias" rather than the domain-based rules. Subsequently, we demonstrate ways to identify and mitigate this. Thereafter, MARS achieves performance on par with current state-of-the-art models while producing model interpretations aligned with known MoAs.
comment: Under review. 10 pages, 5 supplementary pages. Corresponding code is here: https://github.com/laurendelong21/MARS and here: https://github.com/laurendelong21/MoA-Net
♻ ☆ Bridging the Sim-to-Real Gap from the Information Bottleneck Perspective
Reinforcement Learning (RL) has recently achieved remarkable success in robotic control. However, most works in RL operate in simulated environments where privileged knowledge (e.g., dynamics, surroundings, terrains) is readily available. Conversely, in real-world scenarios, robot agents usually rely solely on local states (e.g., proprioceptive feedback of robot joints) to select actions, leading to a significant sim-to-real gap. Existing methods address this gap by either gradually reducing the reliance on privileged knowledge or performing a two-stage policy imitation. However, we argue that these methods are limited in their ability to fully leverage the available privileged knowledge, resulting in suboptimal performance. In this paper, we formulate the sim-to-real gap as an information bottleneck problem and therefore propose a novel privileged knowledge distillation method called the Historical Information Bottleneck (HIB). In particular, HIB learns a privileged knowledge representation from historical trajectories by capturing the underlying changeable dynamic information. Theoretical analysis shows that the learned privileged knowledge representation helps reduce the value discrepancy between the oracle and learned policies. Empirical experiments on both simulated and real-world tasks demonstrate that HIB yields improved generalizability compared to previous methods. Videos of real-world experiments are available at https://sites.google.com/view/history-ib .
comment: Accepted by CoRL 2024
♻ ☆ Improving Consistency Models with Generator-Induced Flows
Consistency models imitate the multi-step sampling of score-based diffusion in a single forward pass of a neural network. They can be learned in two ways: consistency distillation and consistency training. The former relies on the true velocity field of the corresponding differential equation, approximated by a pre-trained neural network. In contrast, the latter uses a single-sample Monte Carlo estimate of this velocity field. The related estimation error induces a discrepancy between consistency distillation and training that, we show, still holds in the continuous-time limit. To alleviate this issue, we propose a novel flow that transports noisy data towards their corresponding outputs derived from the currently trained model --~as a proxy of the true flow. Our empirical findings demonstrate that this approach mitigates the previously identified discrepancy. Furthermore, we present theoretical and empirical evidence indicating that our generator-induced flow surpasses dedicated optimal transport-based consistency models in effectively reducing the noise-data transport cost. Consequently, our method not only accelerates consistency training convergence but also enhances its overall performance. The code is available at: https://github.com/thibautissenhuth/consistency_GC.
♻ ☆ On the Tradeoff between Privacy Preservation and Byzantine-Robustness in Decentralized Learning
This paper jointly considers privacy preservation and Byzantine-robustness in decentralized learning. In a decentralized network, honest-but-curious agents faithfully follow the prescribed algorithm, but expect to infer their neighbors' private data from messages received during the learning process, while dishonest-and-Byzantine agents disobey the prescribed algorithm, and deliberately disseminate wrong messages to their neighbors so as to bias the learning process. For this novel setting, we investigate a generic privacy-preserving and Byzantine-robust decentralized stochastic gradient descent (SGD) framework, in which Gaussian noise is injected to preserve privacy and robust aggregation rules are adopted to counteract Byzantine attacks. We analyze its learning error and privacy guarantee, discovering an essential tradeoff between privacy preservation and Byzantine-robustness in decentralized learning -- the learning error caused by defending against Byzantine attacks is exacerbated by the Gaussian noise added to preserve privacy. For a class of state-of-the-art robust aggregation rules, we give unified analysis of the "mixing abilities". Building upon this analysis, we reveal how the "mixing abilities" affect the tradeoff between privacy preservation and Byzantine-robustness. The theoretical results provide guidelines for achieving a favorable tradeoff with proper design of robust aggregation rules. Numerical experiments are conducted and corroborate our theoretical findings.
♻ ☆ Learning to Balance: Diverse Normalization for Cloth-Changing Person Re-Identification
Cloth-Changing Person Re-Identification (CC-ReID) involves recognizing individuals in images regardless of clothing status. In this paper, we empirically and experimentally demonstrate that completely eliminating or fully retaining clothing features is detrimental to the task. Existing work, either relying on clothing labels, silhouettes, or other auxiliary data, fundamentally aim to balance the learning of clothing and identity features. However, we practically find that achieving this balance is challenging and nuanced. In this study, we introduce a novel module called Diverse Norm, which expands personal features into orthogonal spaces and employs channel attention to separate clothing and identity features. A sample re-weighting optimization strategy is also introduced to guarantee the opposite optimization direction. Diverse Norm presents a simple yet effective approach that does not require additional data. Furthermore, Diverse Norm can be seamlessly integrated ResNet50 and significantly outperforms the state-of-the-art methods.
♻ ☆ Anomaly Detection by Context Contrasting
Anomaly detection focuses on identifying samples that deviate from the norm. When working with high-dimensional data such as images, a crucial requirement for detecting anomalous patterns is learning lower-dimensional representations that capture concepts of normality. Recent advances in self-supervised learning have shown great promise in this regard. However, many successful self-supervised anomaly detection methods assume prior knowledge about anomalies to create synthetic outliers during training. Yet, in real-world applications, we often do not know what to expect from unseen data, and we can solely leverage knowledge about normal data. In this work, we propose Con$_2$, which learns representations through context augmentations that allow us to observe samples from two distinct perspectives while keeping the invariances of normal data. Con$_2$ learns rich representations of context-augmented samples by clustering them according to their context while simultaneously aligning their positions across clusters. At test time, representations of anomalies that do not adhere to the invariances of normal data then deviate from their respective context cluster. Learning representations in such a way thus allows us to detect anomalies without making assumptions about anomalous data.
♻ ☆ Improving Multimodal Learning with Multi-Loss Gradient Modulation
Learning from multiple modalities, such as audio and video, offers opportunities for leveraging complementary information, enhancing robustness, and improving contextual understanding and performance. However, combining such modalities presents challenges, especially when modalities differ in data structure, predictive contribution, and the complexity of their learning processes. It has been observed that one modality can potentially dominate the learning process, hindering the effective utilization of information from other modalities and leading to sub-optimal model performance. To address this issue the vast majority of previous works suggest to assess the unimodal contributions and dynamically adjust the training to equalize them. We improve upon previous work by introducing a multi-loss objective and further refining the balancing process, allowing it to dynamically adjust the learning pace of each modality in both directions, acceleration and deceleration, with the ability to phase out balancing effects upon convergence. We achieve superior results across three audio-video datasets: on CREMA-D, models with ResNet backbone encoders surpass the previous best by 1.9% to 12.4%, and Conformer backbone models deliver improvements ranging from 2.8% to 14.1% across different fusion methods. On AVE, improvements range from 2.7% to 7.7%, while on UCF101, gains reach up to 6.1%.
♻ ☆ How to Build the Virtual Cell with Artificial Intelligence: Priorities and Opportunities
The cell is arguably the most fundamental unit of life and is central to understanding biology. Accurate modeling of cells is important for this understanding as well as for determining the root causes of disease. Recent advances in artificial intelligence (AI), combined with the ability to generate large-scale experimental data, present novel opportunities to model cells. Here we propose a vision of leveraging advances in AI to construct virtual cells, high-fidelity simulations of cells and cellular systems under different conditions that are directly learned from biological data across measurements and scales. We discuss desired capabilities of such AI Virtual Cells, including generating universal representations of biological entities across scales, and facilitating interpretable in silico experiments to predict and understand their behavior using virtual instruments. We further address the challenges, opportunities and requirements to realize this vision including data needs, evaluation strategies, and community standards and engagement to ensure biological accuracy and broad utility. We envision a future where AI Virtual Cells help identify new drug targets, predict cellular responses to perturbations, as well as scale hypothesis exploration. With open science collaborations across the biomedical ecosystem that includes academia, philanthropy, and the biopharma and AI industries, a comprehensive predictive understanding of cell mechanisms and interactions has come into reach.
♻ ☆ Learning Representations of Instruments for Partial Identification of Treatment Effects
Reliable estimation of treatment effects from observational data is important in many disciplines such as medicine. However, estimation is challenging when unconfoundedness as a standard assumption in the causal inference literature is violated. In this work, we leverage arbitrary (potentially high-dimensional) instruments to estimate bounds on the conditional average treatment effect (CATE). Our contributions are three-fold: (1) We propose a novel approach for partial identification through a mapping of instruments to a discrete representation space so that we yield valid bounds on the CATE. This is crucial for reliable decision-making in real-world applications. (2) We derive a two-step procedure that learns tight bounds using a tailored neural partitioning of the latent instrument space. As a result, we avoid instability issues due to numerical approximations or adversarial training. Furthermore, our procedure aims to reduce the estimation variance in finite-sample settings to yield more reliable estimates. (3) We show theoretically that our procedure obtains valid bounds while reducing estimation variance. We further perform extensive experiments to demonstrate the effectiveness across various settings. Overall, our procedure offers a novel path for practitioners to make use of potentially high-dimensional instruments (e.g., as in Mendelian randomization).
♻ ☆ FedCache 2.0: Federated Edge Learning with Knowledge Caching and Dataset Distillation
Federated Edge Learning (FEL) has emerged as a promising approach for enabling edge devices to collaboratively train machine learning models while preserving data privacy. Despite its advantages, practical FEL deployment faces significant challenges related to device constraints and device-server interactions, necessitating heterogeneous, user-adaptive model training with limited and uncertain communication. In this paper, we introduce FedCache 2.0, a novel personalized FEL architecture that simultaneously addresses these challenges. FedCache 2.0 incorporates the benefits of both dataset distillation and knowledge cache-driven federated learning by storing and organizing distilled data as knowledge in the server-side knowledge cache. Moreover, a device-centric cache sampling strategy is introduced to tailor transferred knowledge for individual devices within controlled communication bandwidth. Extensive experiments on five datasets covering image recognition, audio understanding, and mobile sensor data mining tasks demonstrate that (1) FedCache 2.0 significantly outperforms state-of-the-art methods regardless of model structures, data distributions, and modalities. (2) FedCache 2.0 can train splendid personalized on-device models with at least $\times$28.6 improvement in communication efficiency.
comment: 17 pages, 7 figures, 14 tables
♻ ☆ Personalized Item Representations in Federated Multimodal Recommendation
Federated recommendation systems are essential for providing personalized recommendations while protecting user privacy. However, current methods mainly rely on ID-based item embeddings, neglecting the rich multimodal information of items. To address this, we propose a Federated Multimodal Recommendation System, called FedMR. FedMR uses a foundation model on the server to encode multimodal item data, such as images and text. To handle data heterogeneity caused by user preference differences, FedMR introduces a Mixing Feature Fusion Module on each client, which adjusts fusion strategy weights based on user interaction history to generate personalized item representations that capture users' fine-grained preferences. FedMR is compatible with existing ID-based federated recommendation systems, improving performance without modifying the original framework. Experiments on four real-world multimodal datasets demonstrate FedMR's effectiveness. The code is available at https://anonymous.4open.science/r/FedMR.
comment: 12 pages, 4 figures, 5 tables, conference
♻ ☆ RealCompo: Balancing Realism and Compositionality Improves Text-to-Image Diffusion Models NeurIPS 2024
Diffusion models have achieved remarkable advancements in text-to-image generation. However, existing models still have many difficulties when faced with multiple-object compositional generation. In this paper, we propose RealCompo, a new training-free and transferred-friendly text-to-image generation framework, which aims to leverage the respective advantages of text-to-image models and spatial-aware image diffusion models (e.g., layout, keypoints and segmentation maps) to enhance both realism and compositionality of the generated images. An intuitive and novel balancer is proposed to dynamically balance the strengths of the two models in denoising process, allowing plug-and-play use of any model without extra training. Extensive experiments show that our RealCompo consistently outperforms state-of-the-art text-to-image models and spatial-aware image diffusion models in multiple-object compositional generation while keeping satisfactory realism and compositionality of the generated images. Notably, our RealCompo can be seamlessly extended with a wide range of spatial-aware image diffusion models and stylized diffusion models. Our code is available at: https://github.com/YangLing0818/RealCompo
comment: NeurIPS 2024. Project: https://github.com/YangLing0818/RealCompo
♻ ☆ Learning Video Temporal Dynamics with Cross-Modal Attention for Robust Audio-Visual Speech Recognition
Audio-visual speech recognition (AVSR) aims to transcribe human speech using both audio and video modalities. In practical environments with noise-corrupted audio, the role of video information becomes crucial. However, prior works have primarily focused on enhancing audio features in AVSR, overlooking the importance of video features. In this study, we strengthen the video features by learning three temporal dynamics in video data: context order, playback direction, and the speed of video frames. Cross-modal attention modules are introduced to enrich video features with audio information so that speech variability can be taken into account when training on the video temporal dynamics. Based on our approach, we achieve the state-of-the-art performance on the LRS2 and LRS3 AVSR benchmarks for the noise-dominant settings. Our approach excels in scenarios especially for babble and speech noise, indicating the ability to distinguish the speech signal that should be recognized from lip movements in the video modality. We support the validity of our methodology by offering the ablation experiments for the temporal dynamics losses and the cross-modal attention architecture design.
comment: Accepted at SLT 2024 Main Conference; Code is available at https://github.com/sungnyun/avsr-temporal-dynamics
♻ ☆ Finite-difference-informed graph network for solving steady-state incompressible flows on block-structured grids
Advances in deep learning have enabled physics-informed neural networks to solve partial differential equations. Numerical differentiation using the finite-difference (FD) method is efficient in physics-constrained designs, even in parameterized settings. In traditional computational fluid dynamics(CFD), body-fitted block-structured grids are often employed for complex flow cases when obtaining FD solutions. However, convolution operators in convolutional neural networks for FD are typically limited to single-block grids. To address this issue, \blueText{graphs and graph networks are used} to learn flow representations across multi-block-structured grids. \blueText{A graph convolution-based FD method (GC-FDM) is proposed} to train graph networks in a label-free physics-constrained manner, enabling differentiable FD operations on unstructured graph outputs. To demonstrate model performance from single- to multi-block-structured grids, \blueText{the parameterized steady incompressible Navier-Stokes equations are solved} for a lid-driven cavity flow and the flows around single and double circular cylinder configurations. When compared to a CFD solver under various boundary conditions, the proposed method achieves a relative error in velocity field predictions on the order of $10^{-3}$. Furthermore, the proposed method reduces training costs by approximately 20\% compared to a physics-informed neural network. \blueText{To} further verify the effectiveness of GC-FDM in multi-block processing, \blueText{a 30P30N airfoil geometry is considered} and the \blueText{predicted} results are reasonable compared with those given by CFD. \blueText{Finally, the applicability of GC-FDM to three-dimensional (3D) case is tested using a 3D cavity geometry.
♻ ☆ ICU-Sepsis: A Benchmark MDP Built from Real Medical Data
We present ICU-Sepsis, an environment that can be used in benchmarks for evaluating reinforcement learning (RL) algorithms. Sepsis management is a complex task that has been an important topic in applied RL research in recent years. Therefore, MDPs that model sepsis management can serve as part of a benchmark to evaluate RL algorithms on a challenging real-world problem. However, creating usable MDPs that simulate sepsis care in the ICU remains a challenge due to the complexities involved in acquiring and processing patient data. ICU-Sepsis is a lightweight environment that models personalized care of sepsis patients in the ICU. The environment is a tabular MDP that is widely compatible and is challenging even for state-of-the-art RL algorithms, making it a valuable tool for benchmarking their performance. However, we emphasize that while ICU-Sepsis provides a standardized environment for evaluating RL algorithms, it should not be used to draw conclusions that guide medical practice.
comment: Reinforcement Learning Conference 2024
♻ ☆ Online Learning Quantum States with the Logarithmic Loss via VB-FTRL
Online learning of quantum states with the logarithmic loss (LL-OLQS) is a quantum generalization of online portfolio selection (OPS), a classic open problem in online learning for over three decades. This problem also emerges in designing stochastic optimization algorithms for maximum-likelihood quantum state tomography. Recently, Jezequel et al. (arXiv:2209.13932) proposed the VB-FTRL algorithm, the first regret-optimal algorithm for OPS with moderate computational complexity. In this paper, we generalize VB-FTRL for LL-OLQS. Let $d$ denote the dimension and $T$ the number of rounds. The generalized algorithm achieves a regret rate of $O ( d^2 \log ( d + T ) )$ for LL-OLQS. Each iteration of the algorithm consists of solving a semidefinite program that can be implemented in polynomial time by, for example, cutting-plane methods. For comparison, the best-known regret rate for LL-OLQS is currently $O ( d^2 \log T )$, achieved by an exponential weight method. However, no explicit implementation is available for the exponential weight method for LL-OLQS. To facilitate the generalization, we introduce the notion of VB-convexity. VB-convexity is a sufficient condition for the volumetric barrier associated with any function to be convex and is of independent interest.
comment: 24 pages. This version explicitly addresses complex quantum density matrices
♻ ☆ GROD: Enhancing Generalization of Transformer with Out-of-Distribution Detection
Transformer networks excel in natural language processing (NLP) and computer vision (CV) tasks. However, they face challenges in generalizing to Out-of-Distribution (OOD) datasets, that is, data whose distribution differs from that seen during training. The OOD detection aims to distinguish data that deviates from the expected distribution, while maintaining optimal performance on in-distribution (ID) data. This paper introduces a novel approach based on OOD detection, termed the Generate Rounded OOD Data (GROD) algorithm, which significantly bolsters the generalization performance of transformer networks across various tasks. GROD is motivated by our new OOD detection Probably Approximately Correct (PAC) Theory for transformer. The transformer has learnability in terms of OOD detection that is, when the data is sufficient the outlier can be well represented. By penalizing the misclassification of OOD data within the loss function and generating synthetic outliers, GROD guarantees learnability and refines the decision boundaries between inlier and outlier. This strategy demonstrates robust adaptability and general applicability across different data types. Evaluated across diverse OOD detection tasks in NLP and CV, GROD achieves SOTA regardless of data format. The code is available at https://anonymous.4open.science/r/GROD-OOD-Detection-with-transformers-B70F.
♻ ☆ Packing Analysis: Packing Is More Appropriate for Large Models or Datasets in Supervised Fine-tuning
Packing, initially utilized in the pre-training phase, is an optimization technique designed to maximize hardware resource efficiency by combining different training sequences to fit the model's maximum input length. Although it has demonstrated effectiveness during pre-training, there remains a lack of comprehensive analysis for the supervised fine-tuning (SFT) stage on the following points: (1) whether packing can effectively enhance training efficiency while maintaining performance, (2) the suitable size of the model and dataset for fine-tuning with the packing method, and (3) whether packing unrelated or related training samples might cause the model to either excessively disregard or over-rely on the context. In this paper, we perform extensive comparisons between SFT methods using padding and packing, covering SFT datasets ranging from 69K to 1.2M and models from 8B to 70B. This provides the first comprehensive analysis of the advantages and limitations of packing versus padding, as well as practical considerations for implementing packing in various training scenarios. Our analysis covers various benchmarks, including knowledge, reasoning, and coding, as well as GPT-based evaluations, time efficiency, and other fine-tuning parameters. We also open-source our code for fine-tuning and evaluation and provide checkpoints fine-tuned on datasets of different sizes, aiming to advance future research on packing methods. Code is available at: https://github.com/ShuheWang1998/Packing-Analysis?tab=readme-ov-file.
♻ ☆ FC-KAN: Function Combinations in Kolmogorov-Arnold Networks
In this paper, we introduce FC-KAN, a Kolmogorov-Arnold Network (KAN) that leverages combinations of popular mathematical functions such as B-splines, wavelets, and radial basis functions on low-dimensional data through element-wise operations. We explore several methods for combining the outputs of these functions, including sum, element-wise product, the addition of sum and element-wise product, quadratic function representation, and concatenation. In our experiments, we compare FC-KAN with multi-layer perceptron network (MLP) and other existing KANs, such as BSRBF-KAN, EfficientKAN, FastKAN, and FasterKAN, on the MNIST and Fashion-MNIST datasets. A variant of FC-KAN, which uses a combination of outputs from B-splines and Difference of Gaussians (DoG) in the form of a quadratic function, outperformed all other models on the average of 5 independent training runs. We expect that FC-KAN can leverage function combinations to design future KANs. Our repository is publicly available at: https://github.com/hoangthangta/FC_KAN.
comment: 15 pages
♻ ☆ TRANSIC: Sim-to-Real Policy Transfer by Learning from Online Correction
Learning in simulation and transferring the learned policy to the real world has the potential to enable generalist robots. The key challenge of this approach is to address simulation-to-reality (sim-to-real) gaps. Previous methods often require domain-specific knowledge a priori. We argue that a straightforward way to obtain such knowledge is by asking humans to observe and assist robot policy execution in the real world. The robots can then learn from humans to close various sim-to-real gaps. We propose TRANSIC, a data-driven approach to enable successful sim-to-real transfer based on a human-in-the-loop framework. TRANSIC allows humans to augment simulation policies to overcome various unmodeled sim-to-real gaps holistically through intervention and online correction. Residual policies can be learned from human corrections and integrated with simulation policies for autonomous execution. We show that our approach can achieve successful sim-to-real transfer in complex and contact-rich manipulation tasks such as furniture assembly. Through synergistic integration of policies learned in simulation and from humans, TRANSIC is effective as a holistic approach to addressing various, often coexisting sim-to-real gaps. It displays attractive properties such as scaling with human effort. Videos and code are available at https://transic-robot.github.io/
comment: 8th Conference on Robot Learning (CoRL 2024), Munich, Germany. Project website: https://transic-robot.github.io/
♻ ☆ Twisting Lids Off with Two Hands
Manipulating objects with two multi-fingered hands has been a long-standing challenge in robotics, due to the contact-rich nature of many manipulation tasks and the complexity inherent in coordinating a high-dimensional bimanual system. In this work, we share novel insights into physical modeling, real-time perception, and reward design that enable policies trained in simulation using deep reinforcement learning (RL) to be effectively and efficiently transferred to the real world. Specifically, we consider the problem of twisting lids of various bottle-like objects with two hands, demonstrating policies with generalization capabilities across a diverse set of unseen objects as well as dynamic and dexterous behaviors. To the best of our knowledge, this is the first sim-to-real RL system that enables such capabilities on bimanual multi-fingered hands.
comment: Project page can be found at https://toruowo.github.io/bimanual-twist
♻ ☆ Ferrari: Federated Feature Unlearning via Optimizing Feature Sensitivity NeurIPS 2024
The advent of Federated Learning (FL) highlights the practical necessity for the 'right to be forgotten' for all clients, allowing them to request data deletion from the machine learning model's service provider. This necessity has spurred a growing demand for Federated Unlearning (FU). Feature unlearning has gained considerable attention due to its applications in unlearning sensitive features, backdoor features, and bias features. Existing methods employ the influence function to achieve feature unlearning, which is impractical for FL as it necessitates the participation of other clients in the unlearning process. Furthermore, current research lacks an evaluation of the effectiveness of feature unlearning. To address these limitations, we define feature sensitivity in the evaluation of feature unlearning according to Lipschitz continuity. This metric characterizes the rate of change or sensitivity of the model output to perturbations in the input feature. We then propose an effective federated feature unlearning framework called Ferrari, which minimizes feature sensitivity. Extensive experimental results and theoretical analysis demonstrate the effectiveness of Ferrari across various feature unlearning scenarios, including sensitive, backdoor, and biased features.
comment: TLDR: The need for a "right to be forgotten" in Federated Learning has led to the development of the Ferrari framework, which efficiently unlearns sensitive features using a Lipschitz continuity-based metric, proven effective in extensive testing. Accepted at NeurIPS 2024
♻ ☆ Offline Inverse Constrained Reinforcement Learning for Safe-Critical Decision Making in Healthcare
Reinforcement Learning (RL) applied in healthcare can lead to unsafe medical decisions and treatment, such as excessive dosages or abrupt changes, often due to agents overlooking common-sense constraints. Consequently, Constrained Reinforcement Learning (CRL) is a natural choice for safe decisions. However, specifying the exact cost function is inherently difficult in healthcare. Recent Inverse Constrained Reinforcement Learning (ICRL) is a promising approach that infers constraints from expert demonstrations. ICRL algorithms model Markovian decisions in an interactive environment. These settings do not align with the practical requirement of a decision-making system in healthcare, where decisions rely on historical treatment recorded in an offline dataset. To tackle these issues, we propose the Constraint Transformer (CT). Specifically, 1) we utilize a causal attention mechanism to incorporate historical decisions and observations into the constraint modeling, while employing a Non-Markovian layer for weighted constraints to capture critical states. 2) A generative world model is used to perform exploratory data augmentation, enabling offline RL methods to simulate unsafe decision sequences. In multiple medical scenarios, empirical results demonstrate that CT can capture unsafe states and achieve strategies that approximate lower mortality rates, reducing the occurrence probability of unsafe behaviors.
♻ ☆ Signature Kernel Conditional Independence Tests in Causal Discovery for Stochastic Processes
Inferring the causal structure underlying stochastic dynamical systems from observational data holds great promise in domains ranging from science and health to finance. Such processes can often be accurately modeled via stochastic differential equations (SDEs), which naturally imply causal relationships via "which variables enter the differential of which other variables". In this paper, we develop conditional independence (CI) constraints on coordinate processes over selected intervals that are Markov with respect to the acyclic dependence graph (allowing self-loops) induced by a general SDE model. We then provide a sound and complete causal discovery algorithm, capable of handling both fully and partially observed data, and uniquely recovering the underlying or induced ancestral graph by exploiting time directionality assuming a CI oracle. Finally, to make our algorithm practically usable, we also propose a flexible, consistent signature kernel-based CI test to infer these constraints from data. We extensively benchmark the CI test in isolation and as part of our causal discovery algorithms, outperforming existing approaches in SDE models and beyond.
♻ ☆ $\textbf{Only-IF}$:Revealing the Decisive Effect of Instruction Diversity on Generalization
Understanding and accurately following instructions is critical for large language models (LLMs) to be effective across diverse tasks. In this work, we rigorously examine the key factors that enable models to generalize to unseen instructions, providing insights to guide the collection of data for instruction-tuning. Through controlled experiments, inspired by the Turing-complete Markov algorithm, we demonstrate that such generalization $\textbf{only emerges}$ when training data is diversified enough across semantic domains. Our findings also reveal that merely diversifying within limited domains fails to ensure robust generalization. In contrast, cross-domain data diversification, even under constrained data budgets, significantly enhances a model's adaptability. We further extend our analysis to real-world scenarios, including fine-tuning of $\textit{$\textbf{specialist}$}$ and $\textit{$\textbf{generalist}$}$ models. In both cases, we demonstrate that 1) better performance can be achieved by increasing the diversity of an established dataset while keeping the data size constant, and 2) when scaling up the data, diversifying the semantics of instructions is more effective than simply increasing the quantity of similar data. Our research provides important insights for dataset collation, particularly when optimizing model performance by expanding training data for both specialist and generalist scenarios. We show that careful consideration of data diversification is key: training specialist models with data extending beyond their core domain leads to significant performance improvements, while generalist models benefit from diverse data mixtures that enhance their overall instruction-following capabilities across a wide range of applications. Our results highlight the critical role of strategic diversification and offer clear guidelines for improving data quality.
comment: Template change
♻ ☆ Neural Collapse Meets Differential Privacy: Curious Behaviors of NoisyGD with Near-perfect Representation Learning ICML 2024
A recent study by De et al. (2022) has reported that large-scale representation learning through pre-training on a public dataset significantly enhances differentially private (DP) learning in downstream tasks, despite the high dimensionality of the feature space. To theoretically explain this phenomenon, we consider the setting of a layer-peeled model in representation learning, which results in interesting phenomena related to learned features in deep learning and transfer learning, known as Neural Collapse (NC). Within the framework of NC, we establish an error bound indicating that the misclassification error is independent of dimension when the distance between actual features and the ideal ones is smaller than a threshold. Additionally, the quality of the features in the last layer is empirically evaluated under different pre-trained models within the framework of NC, showing that a more powerful transformer leads to a better feature representation. Furthermore, we reveal that DP fine-tuning is less robust compared to fine-tuning without DP, particularly in the presence of perturbations. These observations are supported by both theoretical analyses and experimental evaluation. Moreover, to enhance the robustness of DP fine-tuning, we suggest several strategies, such as feature normalization or employing dimension reduction methods like Principal Component Analysis (PCA). Empirically, we demonstrate a significant improvement in testing accuracy by conducting PCA on the last-layer features.
comment: ICML 2024 (oral)
Multimedia 8
☆ Spatial-Aware Efficient Projector for MLLMs via Multi-Layer Feature Aggregation
The projector plays a crucial role in multi-modal language models (MLLMs). The number of visual tokens it outputs affects the efficiency of the MLLM, while the quality of the visual tokens influences the visual understanding capabilities of the MLLM. Current explorations on the projector focus on reducing the number of visual tokens to improve efficiency, often overlooking the inherent spatial discrepancy between the serialized 2-dimensional visual token sequences and natural language token sequences. A Spatial-Aware Efficient Projector (SAEP) is proposed to address this issue. In detail, our SAEP method employs an modified separable depthwise convolution module on multi-layer visual features to enhance the spatial information of visual tokens. As a result, our SAEP method can not only largely reduce the number of visual tokens by 75\%, but also significantly improve the multimodal spatial understanding capability of MLLMs. Moreover, compared to existing projectors, our SAEP gets best performances on massive multimodal evaluation benchmarks, which denotes its effectiveness on bridging the modality gap.
comment: 10 pages, 3 figures
☆ Evaluating Semantic Variation in Text-to-Image Synthesis: A Causal Perspective
Accurate interpretation and visualization of human instructions are crucial for text-to-image (T2I) synthesis. However, current models struggle to capture semantic variations from word order changes, and existing evaluations, relying on indirect metrics like text-image similarity, fail to reliably assess these challenges. This often obscures poor performance on complex or uncommon linguistic patterns by the focus on frequent word combinations. To address these deficiencies, we propose a novel metric called SemVarEffect and a benchmark named SemVarBench, designed to evaluate the causality between semantic variations in inputs and outputs in T2I synthesis. Semantic variations are achieved through two types of linguistic permutations, while avoiding easily predictable literal variations. Experiments reveal that the CogView-3-Plus and Ideogram 2 performed the best, achieving a score of 0.2/1. Semantic variations in object relations are less understood than attributes, scoring 0.07/1 compared to 0.17-0.19/1. We found that cross-modal alignment in UNet or Transformers plays a crucial role in handling semantic variations, a factor previously overlooked by a focus on textual encoders. Our work establishes an effective evaluation framework that advances the T2I synthesis community's exploration of human instruction understanding.
comment: Our benchmark and code are available at https://github.com/zhuxiangru/SemVarBench
☆ GUISE: Graph GaUssIan Shading watErmark
In the expanding field of generative artificial intelligence, integrating robust watermarking technologies is essential to protect intellectual property and maintain content authenticity. Traditionally, watermarking techniques have been developed primarily for rich information media such as images and audio. However, these methods have not been adequately adapted for graph-based data, particularly molecular graphs. Latent 3D graph diffusion(LDM-3DG) is an ascendant approach in the molecular graph generation field. This model effectively manages the complexities of molecular structures, preserving essential symmetries and topological features. We adapt the Gaussian Shading, a proven performance lossless watermarking technique, to the latent graph diffusion domain to protect this sophisticated new technology. Our adaptation simplifies the watermark diffusion process through duplication and padding, making it adaptable and suitable for various message types. We conduct several experiments using the LDM-3DG model on publicly available datasets QM9 and Drugs, to assess the robustness and effectiveness of our technique. Our results demonstrate that the watermarked molecules maintain statistical parity in 9 out of 10 performance metrics compared to the original. Moreover, they exhibit a 100% detection rate and a 99% extraction rate in a 2D decoded pipeline, while also showing robustness against post-editing attacks.
♻ ☆ Detecting Audio-Visual Deepfakes with Fine-Grained Inconsistencies BMVC 2024
Existing methods on audio-visual deepfake detection mainly focus on high-level features for modeling inconsistencies between audio and visual data. As a result, these approaches usually overlook finer audio-visual artifacts, which are inherent to deepfakes. Herein, we propose the introduction of fine-grained mechanisms for detecting subtle artifacts in both spatial and temporal domains. First, we introduce a local audio-visual model capable of capturing small spatial regions that are prone to inconsistencies with audio. For that purpose, a fine-grained mechanism based on a spatially-local distance coupled with an attention module is adopted. Second, we introduce a temporally-local pseudo-fake augmentation to include samples incorporating subtle temporal inconsistencies in our training set. Experiments on the DFDC and the FakeAVCeleb datasets demonstrate the superiority of the proposed method in terms of generalization as compared to the state-of-the-art under both in-dataset and cross-dataset settings.
comment: Accepted in BMVC 2024
♻ ☆ Multi-Track MusicLDM: Towards Versatile Music Generation with Latent Diffusion Model
Diffusion models have shown promising results in cross-modal generation tasks involving audio and music, such as text-to-sound and text-to-music generation. These text-controlled music generation models typically focus on generating music by capturing global musical attributes like genre and mood. However, music composition is a complex, multilayered task that often involves musical arrangement as an integral part of the process. This process involves composing each instrument to align with existing ones in terms of beat, dynamics, harmony, and melody, requiring greater precision and control over tracks than text prompts usually provide. In this work, we address these challenges by extending the MusicLDM, a latent diffusion model for music, into a multi-track generative model. By learning the joint probability of tracks sharing a context, our model is capable of generating music across several tracks that correspond well to each other, either conditionally or unconditionally. Additionally, our model is capable of arrangement generation, where the model can generate any subset of tracks given the others (e.g., generating a piano track complementing given bass and drum tracks). We compared our model with an existing multi-track generative model and demonstrated that our model achieves considerable improvements across objective metrics for both total and arrangement generation tasks.
♻ ☆ Proceedings of The second international workshop on eXplainable AI for the Arts (XAIxArts)
This second international workshop on explainable AI for the Arts (XAIxArts) brought together a community of researchers in HCI, Interaction Design, AI, explainable AI (XAI), and digital arts to explore the role of XAI for the Arts. Workshop held at the 16th ACM Conference on Creativity and Cognition (C&C 2024), Chicago, USA.
comment: Proceedings of The second international workshop on eXplainable AI for the Arts (XAIxArts)
♻ ☆ Improving Multimodal Learning with Multi-Loss Gradient Modulation
Learning from multiple modalities, such as audio and video, offers opportunities for leveraging complementary information, enhancing robustness, and improving contextual understanding and performance. However, combining such modalities presents challenges, especially when modalities differ in data structure, predictive contribution, and the complexity of their learning processes. It has been observed that one modality can potentially dominate the learning process, hindering the effective utilization of information from other modalities and leading to sub-optimal model performance. To address this issue the vast majority of previous works suggest to assess the unimodal contributions and dynamically adjust the training to equalize them. We improve upon previous work by introducing a multi-loss objective and further refining the balancing process, allowing it to dynamically adjust the learning pace of each modality in both directions, acceleration and deceleration, with the ability to phase out balancing effects upon convergence. We achieve superior results across three audio-video datasets: on CREMA-D, models with ResNet backbone encoders surpass the previous best by 1.9% to 12.4%, and Conformer backbone models deliver improvements ranging from 2.8% to 14.1% across different fusion methods. On AVE, improvements range from 2.7% to 7.7%, while on UCF101, gains reach up to 6.1%.
♻ ☆ SceneDreamer360: Text-Driven 3D-Consistent Scene Generation with Panoramic Gaussian Splatting
Text-driven 3D scene generation has seen significant advancements recently. However, most existing methods generate single-view images using generative models and then stitch them together in 3D space. This independent generation for each view often results in spatial inconsistency and implausibility in the 3D scenes. To address this challenge, we proposed a novel text-driven 3D-consistent scene generation model: SceneDreamer360. Our proposed method leverages a text-driven panoramic image generation model as a prior for 3D scene generation and employs 3D Gaussian Splatting (3DGS) to ensure consistency across multi-view panoramic images. Specifically, SceneDreamer360 enhances the fine-tuned Panfusion generator with a three-stage panoramic enhancement, enabling the generation of high-resolution, detail-rich panoramic images. During the 3D scene construction, a novel point cloud fusion initialization method is used, producing higher quality and spatially consistent point clouds. Our extensive experiments demonstrate that compared to other methods, SceneDreamer360 with its panoramic image generation and 3DGS can produce higher quality, spatially consistent, and visually appealing 3D scenes from any text prompt. Our codes are available at \url{https://github.com/liwrui/SceneDreamer360}.
Computation and Language 58
☆ LoRE: Logit-Ranked Retriever Ensemble for Enhancing Open-Domain Question Answering
Retrieval-based question answering systems often suffer from positional bias, leading to suboptimal answer generation. We propose LoRE (Logit-Ranked Retriever Ensemble), a novel approach that improves answer accuracy and relevance by mitigating positional bias. LoRE employs an ensemble of diverse retrievers, such as BM25 and sentence transformers with FAISS indexing. A key innovation is a logit-based answer ranking algorithm that combines the logit scores from a large language model (LLM), with the retrieval ranks of the passages. Experimental results on NarrativeQA, SQuAD demonstrate that LoRE significantly outperforms existing retrieval-based methods in terms of exact match and F1 scores. On SQuAD, LoRE achieves 14.5\%, 22.83\%, and 14.95\% improvements over the baselines for ROUGE-L, EM, and F1, respectively. Qualitatively, LoRE generates more relevant and accurate answers, especially for complex queries.
☆ A Step Towards Mixture of Grader: Statistical Analysis of Existing Automatic Evaluation Metrics
The explosion of open-sourced models and Question-Answering (QA) datasets emphasizes the importance of automated QA evaluation. We studied the statistics of the existing evaluation metrics for a better understanding of their limitations. By measuring the correlation coefficients of each evaluation metric concerning human-like evaluation score, we observed the following: (1) existing metrics have a high correlation among them concerning the question type (e.g., single word, single phrase, etc.), (2) no single metric can adequately estimate the human-like evaluation. As a potential solution, we discuss how a Mixture Of Grader could potentially improve the auto QA evaluator quality.
☆ Safety-Aware Fine-Tuning of Large Language Models NeurIPS 2024
Fine-tuning Large Language Models (LLMs) has emerged as a common practice for tailoring models to individual needs and preferences. The choice of datasets for fine-tuning can be diverse, introducing safety concerns regarding the potential inclusion of harmful data samples. Manually filtering or avoiding such samples, however, can be labor-intensive and subjective. To address these difficulties, we propose a novel Safety-Aware Fine-Tuning (SAFT) framework designed to automatically detect and remove potentially harmful data, by leveraging a scoring function that exploits the subspace information of harmful and benign samples. Experimental results demonstrate the efficacy of SAFT across different LLMs and varying contamination rates, achieving reductions in harmfulness of up to 27.8%. Going beyond, we delve into the mechanism of our approach and validate its versatility in addressing practical challenges in real-world scenarios.
comment: NeurIPS 2024 Workshop on Safe Generative AI
☆ Leveraging Customer Feedback for Multi-modal Insight Extraction NAACL 2024
Businesses can benefit from customer feedback in different modalities, such as text and images, to enhance their products and services. However, it is difficult to extract actionable and relevant pairs of text segments and images from customer feedback in a single pass. In this paper, we propose a novel multi-modal method that fuses image and text information in a latent space and decodes it to extract the relevant feedback segments using an image-text grounded text decoder. We also introduce a weakly-supervised data generation technique that produces training data for this task. We evaluate our model on unseen data and demonstrate that it can effectively mine actionable insights from multi-modal customer feedback, outperforming the existing baselines by $14$ points in F1 score.
comment: NAACL 2024
☆ Collu-Bench: A Benchmark for Predicting Language Model Hallucinations in Code
Despite their success, large language models (LLMs) face the critical challenge of hallucinations, generating plausible but incorrect content. While much research has focused on hallucinations in multiple modalities including images and natural language text, less attention has been given to hallucinations in source code, which leads to incorrect and vulnerable code that causes significant financial loss. To pave the way for research in LLMs' hallucinations in code, we introduce Collu-Bench, a benchmark for predicting code hallucinations of LLMs across code generation (CG) and automated program repair (APR) tasks. Collu-Bench includes 13,234 code hallucination instances collected from five datasets and 11 diverse LLMs, ranging from open-source models to commercial ones. To better understand and predict code hallucinations, Collu-Bench provides detailed features such as the per-step log probabilities of LLMs' output, token types, and the execution feedback of LLMs' generated code for in-depth analysis. In addition, we conduct experiments to predict hallucination on Collu-Bench, using both traditional machine learning techniques and neural networks, which achieves 22.03 -- 33.15% accuracy. Our experiments draw insightful findings of code hallucination patterns, reveal the challenge of accurately localizing LLMs' hallucinations, and highlight the need for more sophisticated techniques.
☆ Evaluating Gender Bias of LLMs in Making Morality Judgements EMNLP
Large Language Models (LLMs) have shown remarkable capabilities in a multitude of Natural Language Processing (NLP) tasks. However, these models are still not immune to limitations such as social biases, especially gender bias. This work investigates whether current closed and open-source LLMs possess gender bias, especially when asked to give moral opinions. To evaluate these models, we curate and introduce a new dataset GenMO (Gender-bias in Morality Opinions) comprising parallel short stories featuring male and female characters respectively. Specifically, we test models from the GPT family (GPT-3.5-turbo, GPT-3.5-turbo-instruct, GPT-4-turbo), Llama 3 and 3.1 families (8B/70B), Mistral-7B and Claude 3 families (Sonnet and Opus). Surprisingly, despite employing safety checks, all production-standard models we tested display significant gender bias with GPT-3.5-turbo giving biased opinions in 24% of the samples. Additionally, all models consistently favour female characters, with GPT showing bias in 68-85% of cases and Llama 3 in around 81-85% instances. Additionally, our study investigates the impact of model parameters on gender bias and explores real-world situations where LLMs reveal biases in moral decision-making.
comment: Accepted by EMNLP Findings 2024
☆ MARS: Multilingual Aspect-centric Review Summarisation EMNLP 2024
Summarizing customer feedback to provide actionable insights for products/services at scale is an important problem for businesses across industries. Lately, the review volumes are increasing across regions and languages, therefore the challenge of aggregating and understanding customer sentiment across multiple languages becomes increasingly vital. In this paper, we propose a novel framework involving a two-step paradigm \textit{Extract-then-Summarise}, namely MARS to revolutionise traditions and address the domain agnostic aspect-level multilingual review summarisation. Extensive automatic and human evaluation shows that our approach brings substantial improvements over abstractive baselines and efficiency to real-time systems.
comment: EMNLP 2024
☆ Self-Data Distillation for Recovering Quality in Pruned Large Language Models NeurIPS 2024
Large language models have driven significant progress in natural language processing, but their deployment requires substantial compute and memory resources. As models scale, compression techniques become essential for balancing model quality with computational efficiency. Structured pruning, which removes less critical components of the model, is a promising strategy for reducing complexity. However, one-shot pruning often results in significant quality degradation, particularly in tasks requiring multi-step reasoning. To recover lost quality, supervised fine-tuning (SFT) is commonly applied, but it can lead to catastrophic forgetting by shifting the model's learned data distribution. Therefore, addressing the degradation from both pruning and SFT is essential to preserve the original model's quality. In this work, we propose self-data distilled fine-tuning to address these challenges. Our approach leverages the original, unpruned model to generate a distilled dataset that preserves semantic richness and mitigates catastrophic forgetting by maintaining alignment with the base model's knowledge. Empirically, we demonstrate that self-data distillation consistently outperforms standard SFT, improving average accuracy by up to 8% on the HuggingFace OpenLLM Leaderboard v1. Specifically, when pruning 6 decoder blocks on Llama3.1-8B Instruct (i.e., 32 to 24 layers, reducing the model size from 8.03B to 6.72B parameters), our method retains 91.2% of the original model's accuracy compared to 81.7% with SFT, while reducing real-world FLOPs by 16.30%. Furthermore, our approach scales effectively across datasets, with the quality improving as the dataset size increases.
comment: Accepted at the NeurIPS 2024 Machine Learning and Compression Workshop
☆ When Neutral Summaries are not that Neutral: Quantifying Political Neutrality in LLM-Generated News Summaries
In an era where societal narratives are increasingly shaped by algorithmic curation, investigating the political neutrality of LLMs is an important research question. This study presents a fresh perspective on quantifying the political neutrality of LLMs through the lens of abstractive text summarization of polarizing news articles. We consider five pressing issues in current US politics: abortion, gun control/rights, healthcare, immigration, and LGBTQ+ rights. Via a substantial corpus of 20,344 news articles, our study reveals a consistent trend towards pro-Democratic biases in several well-known LLMs, with gun control and healthcare exhibiting the most pronounced biases (max polarization differences of -9.49% and -6.14%, respectively). Further analysis uncovers a strong convergence in the vocabulary of the LLM outputs for these divisive topics (55% overlap for Democrat-leaning representations, 52% for Republican). Being months away from a US election of consequence, we consider our findings important.
comment: 12 pages, 3 figures, 4 tables
☆ MisinfoEval: Generative AI in the Era of "Alternative Facts" EMNLP 2024
The spread of misinformation on social media platforms threatens democratic processes, contributes to massive economic losses, and endangers public health. Many efforts to address misinformation focus on a knowledge deficit model and propose interventions for improving users' critical thinking through access to facts. Such efforts are often hampered by challenges with scalability, and by platform users' personal biases. The emergence of generative AI presents promising opportunities for countering misinformation at scale across ideological barriers. In this paper, we introduce a framework (MisinfoEval) for generating and comprehensively evaluating large language model (LLM) based misinformation interventions. We present (1) an experiment with a simulated social media environment to measure effectiveness of misinformation interventions, and (2) a second experiment with personalized explanations tailored to the demographics and beliefs of users with the goal of countering misinformation by appealing to their pre-existing values. Our findings confirm that LLM-based interventions are highly effective at correcting user behavior (improving overall user accuracy at reliability labeling by up to 41.72%). Furthermore, we find that users favor more personalized interventions when making decisions about news reliability and users shown personalized interventions have significantly higher accuracy at identifying misinformation.
comment: EMNLP 2024. Correspondence can be sent to skgabrie@cs.ucla.edu
☆ State of NLP in Kenya: A Survey
Kenya, known for its linguistic diversity, faces unique challenges and promising opportunities in advancing Natural Language Processing (NLP) technologies, particularly for its underrepresented indigenous languages. This survey provides a detailed assessment of the current state of NLP in Kenya, emphasizing ongoing efforts in dataset creation, machine translation, sentiment analysis, and speech recognition for local dialects such as Kiswahili, Dholuo, Kikuyu, and Luhya. Despite these advancements, the development of NLP in Kenya remains constrained by limited resources and tools, resulting in the underrepresentation of most indigenous languages in digital spaces. This paper uncovers significant gaps by critically evaluating the available datasets and existing NLP models, most notably the need for large-scale language models and the insufficient digital representation of Indigenous languages. We also analyze key NLP applications: machine translation, information retrieval, and sentiment analysis-examining how they are tailored to address local linguistic needs. Furthermore, the paper explores the governance, policies, and regulations shaping the future of AI and NLP in Kenya and proposes a strategic roadmap to guide future research and development efforts. Our goal is to provide a foundation for accelerating the growth of NLP technologies that meet Kenya's diverse linguistic demands.
comment: 21 pages
☆ Learning to Rank for Multiple Retrieval-Augmented Models through Iterative Utility Maximization
This paper investigates the design of a unified search engine to serve multiple retrieval-augmented generation (RAG) agents, each with a distinct task, backbone large language model (LLM), and retrieval-augmentation strategy. We introduce an iterative approach where the search engine generates retrieval results for these RAG agents and gathers feedback on the quality of the retrieved documents during an offline phase. This feedback is then used to iteratively optimize the search engine using a novel expectation-maximization algorithm, with the goal of maximizing each agent's utility function. Additionally, we adapt this approach to an online setting, allowing the search engine to refine its behavior based on real-time individual agents feedback to better serve the results for each of them. Experiments on diverse datasets from the Knowledge-Intensive Language Tasks (KILT) benchmark demonstrates that our approach significantly on average outperforms competitive baselines across 18 RAG models. We also demonstrate that our method effectively ``personalizes'' the retrieval process for each RAG agent based on the collected feedback. Finally, we provide a comprehensive ablation study to explore various aspects of our method.
☆ Retrieval Instead of Fine-tuning: A Retrieval-based Parameter Ensemble for Zero-shot Learning
Foundation models have become a cornerstone in deep learning, with techniques like Low-Rank Adaptation (LoRA) offering efficient fine-tuning of large models. Similarly, methods such as Retrieval-Augmented Generation (RAG), which leverage vectorized databases, have further improved model performance by grounding outputs in external information. While these approaches have demonstrated notable success, they often require extensive training or labeled data, which can limit their adaptability in resource-constrained environments. To address these challenges, we introduce Retrieval-based Parameter Ensemble (RPE), a new method that creates a vectorized database of LoRAs, enabling efficient retrieval and application of model adaptations to new tasks. RPE minimizes the need for extensive training and eliminates the requirement for labeled data, making it particularly effective for zero-shot learning. Additionally, RPE is well-suited for privacy-sensitive domains like healthcare, as it modifies model parameters without accessing raw data. When applied to tasks such as medical report generation and image segmentation, RPE not only proved effective but also surpassed supervised fine-tuning methods in certain cases, highlighting its potential to enhance both computational efficiency and privacy in deep learning applications.
☆ Reddit is all you need: Authorship profiling for Romanian
Authorship profiling is the process of identifying an author's characteristics based on their writings. This centuries old problem has become more intriguing especially with recent developments in Natural Language Processing (NLP). In this paper, we introduce a corpus of short texts in the Romanian language, annotated with certain author characteristic keywords; to our knowledge, the first of its kind. In order to do this, we exploit a social media platform called Reddit. We leverage its thematic community-based structure (subreddits structure), which offers information about the author's background. We infer an user's demographic and some broad personal traits, such as age category, employment status, interests, and social orientation based on the subreddit and other cues. We thus obtain a 23k+ samples corpus, extracted from 100+ Romanian subreddits. We analyse our dataset, and finally, we fine-tune and evaluate Large Language Models (LLMs) to prove baselines capabilities for authorship profiling using the corpus, indicating the need for further research in the field. We publicly release all our resources.
comment: 10 pages, 5 tables and 1 figure, submitted to The 19th International Conference on Linguistic Resources and Tools for Natural Language Processing (ConsILR 2024)
☆ RMB: Comprehensively Benchmarking Reward Models in LLM Alignment
Reward models (RMs) guide the alignment of large language models (LLMs), steering them toward behaviors preferred by humans. Evaluating RMs is the key to better aligning LLMs. However, the current evaluation of RMs may not directly correspond to their alignment performance due to the limited distribution of evaluation data and evaluation methods that are not closely related to alignment objectives. To address these limitations, we propose RMB, a comprehensive RM benchmark that covers over 49 real-world scenarios and includes both pairwise and Best-of-N (BoN) evaluations to better reflect the effectiveness of RMs in guiding alignment optimization. We demonstrate a positive correlation between our benchmark and the downstream alignment task performance. Based on our benchmark, we conduct extensive analysis on the state-of-the-art RMs, revealing their generalization defects that were not discovered by previous benchmarks, and highlighting the potential of generative RMs. Furthermore, we delve into open questions in reward models, specifically examining the effectiveness of majority voting for the evaluation of reward models and analyzing the impact factors of generative RMs, including the influence of evaluation criteria and instructing methods. Our evaluation code and datasets are available at https://github.com/Zhou-Zoey/RMB-Reward-Model-Benchmark.
☆ ChroKnowledge: Unveiling Chronological Knowledge of Language Models in Multiple Domains
Large language models (LLMs) have significantly impacted many aspects of our lives. However, assessing and ensuring their chronological knowledge remains challenging. Existing approaches fall short in addressing the accumulative nature of knowledge, often relying on a single time stamp. To overcome this, we introduce ChroKnowBench, a benchmark dataset designed to evaluate chronologically accumulated knowledge across three key aspects: multiple domains, time dependency, temporal state. Our benchmark distinguishes between knowledge that evolves (e.g., scientific discoveries, amended laws) and knowledge that remain constant (e.g., mathematical truths, commonsense facts). Building on this benchmark, we present ChroKnowledge (Chronological Categorization of Knowledge), a novel sampling-based framework for evaluating and updating LLMs' non-parametric chronological knowledge. Our evaluation shows: (1) The ability of eliciting temporal knowledge varies depending on the data format that model was trained on. (2) LLMs partially recall knowledge or show a cut-off at temporal boundaries rather than recalling all aspects of knowledge correctly. Thus, we apply our ChroKnowPrompt, an in-depth prompting to elicit chronological knowledge by traversing step-by-step through the surrounding time spans. We observe that our framework successfully updates the overall knowledge across the entire timeline in both the biomedical domain (+11.9%) and the general domain (+2.8%), demonstrating its effectiveness in refining temporal knowledge. This non-parametric approach also enables knowledge updates not only in open-source models but also in proprietary LLMs, ensuring comprehensive applicability across model types. We perform a comprehensive analysis based on temporal characteristics of ChroKnowPrompt and validate the potential of various models to elicit intrinsic temporal knowledge through our method.
☆ Generating Driving Simulations via Conversation
Cyber-physical systems like autonomous vehicles are tested in simulation before deployment, using domain-specific programs for scenario specification. To aid the testing of autonomous vehicles in simulation, we design a natural language interface, using an instruction-following large language model, to assist a non-coding domain expert in synthesising the desired scenarios and vehicle behaviours. We show that using it to convert utterances to the symbolic program is feasible, despite the very small training dataset. Human experiments show that dialogue is critical to successful simulation generation, leading to a 4.5 times higher success rate than a generation without engaging in extended conversation.
comment: 6 pages, 6 figures, 2 tables
☆ Dynamic and Textual Graph Generation Via Large-Scale LLM-based Agent Simulation
Graph generation is a fundamental task that has been extensively studied in social, technological, and scientific analysis. For modeling the dynamic graph evolution process, traditional rule-based methods struggle to capture community structures within graphs, while deep learning methods only focus on fitting training graphs. This limits existing graph generators to producing graphs that adhere to predefined rules or closely resemble training datasets, achieving poor performance in dynamic graph generation. Given that graphs are abstract representations arising from pairwise interactions in human activities, a realistic simulation of human-wise interaction could provide deeper insights into the graph evolution mechanism. With the increasing recognition of large language models (LLMs) in simulating human behavior, we introduce GraphAgent-Generator (GAG), a novel simulation-based framework for dynamic graph generation. Without training or fine-tuning process of LLM, our framework effectively replicates seven macro-level structural characteristics in established network science theories while surpassing existing baselines in graph expansion tasks by 31\% on specific evaluation metrics. Through node classification task, we validate GAG effectively preserves characteristics of real-world network for node-wise textual features in generated text-rich graph. Furthermore, by incorporating parallel acceleration, GAG supports generating graphs with up to nearly 100,000 nodes or 10 million edges through large-scale LLM-based agent simulation, with a minimum speed-up of 90.4\%. The source code is available at https://anonymous.4open.science/r/GraphAgent-2206.
☆ Simultaneous Computation and Memory Efficient Zeroth-Order Optimizer for Fine-Tuning Large Language Models
Fine-tuning is powerful for adapting large language models to downstream tasks, but it often results in huge memory usages. A promising approach to mitigate this is using Zeroth-Order (ZO) optimization, which estimates gradients to replace First-Order (FO) gradient calculations, albeit with longer training time due to its stochastic nature. By revisiting the Memory-efficient ZO (MeZO) optimizer, we discover that the full-parameter perturbation and updating processes consume over 50% of its overall fine-tuning time cost. Based on these observations, we introduce a novel layer-wise sparse computation and memory efficient ZO optimizer, named LeZO. LeZO treats layers as fundamental units for sparsification and dynamically perturbs different parameter subsets in each step to achieve full-parameter fine-tuning. LeZO incorporates layer-wise parameter sparsity in the process of simultaneous perturbation stochastic approximation (SPSA) and ZO stochastic gradient descent (ZO-SGD). It achieves accelerated computation during perturbation and updating processes without additional memory overhead. We conduct extensive experiments with the OPT model family on the SuperGLUE benchmark and two generative tasks. The experiments show that LeZO accelerates training without compromising the performance of ZO optimization. Specifically, it achieves over 3x speedup compared to MeZO on the SST-2, BoolQ, and Copa tasks.
☆ Reverse Modeling in Large Language Models
Humans are accustomed to reading and writing in a forward manner, and this natural bias extends to text understanding in auto-regressive large language models (LLMs). This paper investigates whether LLMs, like humans, struggle with reverse modeling, specifically with reversed text inputs. We found that publicly available pre-trained LLMs cannot understand such inputs. However, LLMs trained from scratch with both forward and reverse texts can understand them equally well during inference. Our case study shows that different-content texts result in different losses if input (to LLMs) in different directions -- some get lower losses for forward while some for reverse. This leads us to a simple and nice solution for data selection based on the loss differences between forward and reverse directions. Using our selected data in continued pretraining can boost LLMs' performance by a large margin across different language understanding benchmarks.
comment: 13 Pages, 6 Figures, 7 Tables
☆ Single Ground Truth Is Not Enough: Add Linguistic Variability to Aspect-based Sentiment Analysis Evaluation
Aspect-based sentiment analysis (ABSA) is the challenging task of extracting sentiment along with its corresponding aspects and opinions from human language. Due to the inherent variability of natural language, aspect and opinion terms can be expressed in various surface forms, making their accurate identification complex. Current evaluation methods for this task often restrict answers to a single ground truth, penalizing semantically equivalent predictions that differ in surface form. To address this limitation, we propose a novel, fully automated pipeline that augments existing test sets with alternative valid responses for aspect and opinion terms. This approach enables a fairer assessment of language models by accommodating linguistic diversity, resulting in higher human agreement than single-answer test sets (up to 10%p improvement in Kendall's Tau score). Our experimental results demonstrate that Large Language Models (LLMs) show substantial performance improvements over T5 models when evaluated using our augmented test set, suggesting that LLMs' capabilities in ABSA tasks may have been underestimated. This work contributes to a more comprehensive evaluation framework for ABSA, potentially leading to more accurate assessments of model performance in information extraction tasks, particularly those involving span extraction.
comment: Preprint
☆ BlackDAN: A Black-Box Multi-Objective Approach for Effective and Contextual Jailbreaking of Large Language Models
While large language models (LLMs) exhibit remarkable capabilities across various tasks, they encounter potential security risks such as jailbreak attacks, which exploit vulnerabilities to bypass security measures and generate harmful outputs. Existing jailbreak strategies mainly focus on maximizing attack success rate (ASR), frequently neglecting other critical factors, including the relevance of the jailbreak response to the query and the level of stealthiness. This narrow focus on single objectives can result in ineffective attacks that either lack contextual relevance or are easily recognizable. In this work, we introduce BlackDAN, an innovative black-box attack framework with multi-objective optimization, aiming to generate high-quality prompts that effectively facilitate jailbreaking while maintaining contextual relevance and minimizing detectability. BlackDAN leverages Multiobjective Evolutionary Algorithms (MOEAs), specifically the NSGA-II algorithm, to optimize jailbreaks across multiple objectives including ASR, stealthiness, and semantic relevance. By integrating mechanisms like mutation, crossover, and Pareto-dominance, BlackDAN provides a transparent and interpretable process for generating jailbreaks. Furthermore, the framework allows customization based on user preferences, enabling the selection of prompts that balance harmfulness, relevance, and other factors. Experimental results demonstrate that BlackDAN outperforms traditional single-objective methods, yielding higher success rates and improved robustness across various LLMs and multimodal LLMs, while ensuring jailbreak responses are both relevant and less detectable.
☆ Expanding Search Space with Diverse Prompting Agents: An Efficient Sampling Approach for LLM Mathematical Reasoning
Large Language Models (LLMs) have exhibited remarkable capabilities in many complex tasks including mathematical reasoning. However, traditional approaches heavily rely on ensuring self-consistency within single prompting method, which limits the exploration of diverse problem-solving strategies. This study addresses these limitations by performing an experimental analysis of distinct prompting methods within the domain of mathematical reasoning. Our findings demonstrate that each method explores a distinct search space, and this differentiation becomes more evident with increasing problem complexity. To leverage this phenomenon, we applied efficient sampling process that uniformly combines samples from these diverse methods, which not only expands the maximum search space but achieves higher performance with fewer runs compared to single methods. Especially, within the subset of difficult questions of MATH dataset named MATH-hard, The maximum search space was achieved while utilizing approximately 43% fewer runs than single methods on average. These findings highlight the importance of integrating diverse problem-solving strategies to enhance the reasoning abilities of LLMs.
comment: 6 pages, 4 figures
☆ ECIS-VQG: Generation of Entity-centric Information-seeking Questions from Videos EMNLP 2024
Previous studies on question generation from videos have mostly focused on generating questions about common objects and attributes and hence are not entity-centric. In this work, we focus on the generation of entity-centric information-seeking questions from videos. Such a system could be useful for video-based learning, recommending ``People Also Ask'' questions, video-based chatbots, and fact-checking. Our work addresses three key challenges: identifying question-worthy information, linking it to entities, and effectively utilizing multimodal signals. Further, to the best of our knowledge, there does not exist a large-scale dataset for this task. Most video question generation datasets are on TV shows, movies, or human activities or lack entity-centric information-seeking questions. Hence, we contribute a diverse dataset of YouTube videos, VideoQuestions, consisting of 411 videos with 2265 manually annotated questions. We further propose a model architecture combining Transformers, rich context signals (titles, transcripts, captions, embeddings), and a combination of cross-entropy and contrastive loss function to encourage entity-centric question generation. Our best method yields BLEU, ROUGE, CIDEr, and METEOR scores of 71.3, 78.6, 7.31, and 81.9, respectively, demonstrating practical usability. We make the code and dataset publicly available. https://github.com/thePhukan/ECIS-VQG
comment: Accepted in EMNLP 2024, https://openreview.net/forum?id=CriKOn01dI
☆ EasyJudge: an Easy-to-use Tool for Comprehensive Response Evaluation of LLMs
Recently, there has been a growing trend of employing large language models (LLMs) to judge the quality of other LLMs. Many studies have adopted closed-source models, mainly using GPT-4 as the evaluator. However, due to the closed-source nature of the GPT-4 model, employing it as an evaluator has resulted in issues including transparency, controllability, and cost-effectiveness. Some researchers have turned to using fine-tuned open-source LLMs as evaluators. However, existing open-source evaluation LLMs generally lack a user-friendly visualization tool, and they have not been optimized for accelerated model inference, which causes inconvenience for researchers with limited resources and those working across different fields. This paper presents EasyJudge, a model developed to evaluate significant language model responses. It is lightweight, precise, efficient, and user-friendly, featuring an intuitive visualization interface for ease of deployment and use. EasyJudge uses detailed datasets and refined prompts for model optimization, achieving strong consistency with human and proprietary model evaluations. The model optimized with quantitative methods enables EasyJudge to run efficiently on consumer-grade GPUs or even CPUs. We also provide detailed analysis and case studies to further reveal the potential of our method.
☆ A Mixed-Language Multi-Document News Summarization Dataset and a Graphs-Based Extract-Generate Model
Existing research on news summarization primarily focuses on single-language single-document (SLSD), single-language multi-document (SLMD) or cross-language single-document (CLSD). However, in real-world scenarios, news about a international event often involves multiple documents in different languages, i.e., mixed-language multi-document (MLMD). Therefore, summarizing MLMD news is of great significance. However, the lack of datasets for MLMD news summarization has constrained the development of research in this area. To fill this gap, we construct a mixed-language multi-document news summarization dataset (MLMD-news), which contains four different languages and 10,992 source document cluster and target summary pairs. Additionally, we propose a graph-based extract-generate model and benchmark various methods on the MLMD-news dataset and publicly release our dataset and code\footnote[1]{https://github.com/Southnf9/MLMD-news}, aiming to advance research in summarization within MLMD scenarios.
☆ 'Quis custodiet ipsos custodes?' Who will watch the watchmen? On Detecting AI-generated peer-reviews EMNLP
The integrity of the peer-review process is vital for maintaining scientific rigor and trust within the academic community. With the steady increase in the usage of large language models (LLMs) like ChatGPT in academic writing, there is a growing concern that AI-generated texts could compromise scientific publishing, including peer-reviews. Previous works have focused on generic AI-generated text detection or have presented an approach for estimating the fraction of peer-reviews that can be AI-generated. Our focus here is to solve a real-world problem by assisting the editor or chair in determining whether a review is written by ChatGPT or not. To address this, we introduce the Term Frequency (TF) model, which posits that AI often repeats tokens, and the Review Regeneration (RR) model, which is based on the idea that ChatGPT generates similar outputs upon re-prompting. We stress test these detectors against token attack and paraphrasing. Finally, we propose an effective defensive strategy to reduce the effect of paraphrasing on our models. Our findings suggest both our proposed methods perform better than the other AI text detectors. Our RR model is more robust, although our TF model performs better than the RR model without any attacks. We make our code, dataset, and model public.
comment: EMNLP Main, 17 pages, 5 figures, 9 tables
☆ BiDoRA: Bi-level Optimization-Based Weight-Decomposed Low-Rank Adaptation
Parameter-efficient fine-tuning (PEFT) of large language models (LLMs) has gained considerable attention as a flexible and efficient way of adapting LLMs to downstream tasks. Among these methods, weighted decomposed low-rank adaptation (DoRA) has emerged as a promising approach. DoRA bridges the gap between low-rank adaptation (LoRA) and full fine-tuning (FT) by decomposing the weight matrices into magnitude and direction components, thereby maintaining learning behavior similar to FT. Although DoRA shows encouraging performance, it introduces additional parameters compared to LoRA, which potentially increases the risk of overfitting. Moreover, optimizing magnitude and direction simultaneously leads to a coupled gradient updating pattern for both components, limiting its learning capacity. To overcome these limitations, we propose BiDoRA, a bi-level optimization-based PEFT method. In BiDoRA, the direction and magnitude components are optimized on two distinct datasets at different optimization levels, mitigating the risk of overfitting. Additionally, the asynchronous optimization of the two components promotes their decoupling, allowing for more flexible gradient updates suitable for various downstream tasks. Evaluation of BiDoRA on fourteen datasets spanning natural language understanding, natural language generation, and token classification reveals that it significantly outperforms DoRA and other PEFT methods. The superior performance of BiDoRA underscores its effectiveness. The code for BiDoRA is available at https://anonymous.4open.science/r/BiDoRA-5D31.
☆ Empirical Study of Mutual Reinforcement Effect and Application in Few-shot Text Classification Tasks via Prompt
The Mutual Reinforcement Effect (MRE) investigates the synergistic relationship between word-level and text-level classifications in text classification tasks. It posits that the performance of both classification levels can be mutually enhanced. However, this mechanism has not been adequately demonstrated or explained in prior research. To address this gap, we employ empirical experiment to observe and substantiate the MRE theory. Our experiments on 21 MRE mix datasets revealed the presence of MRE in the model and its impact. Specifically, we conducted compare experiments use fine-tune. The results of findings from comparison experiments corroborates the existence of MRE. Furthermore, we extended the application of MRE to prompt learning, utilizing word-level information as a verbalizer to bolster the model's prediction of text-level classification labels. In our final experiment, the F1-score significantly surpassed the baseline in 18 out of 21 MRE Mix datasets, further validating the notion that word-level information enhances the language model's comprehension of the text as a whole.
comment: 10 pagess, 4 figures
♻ ☆ Semantic Parsing with Candidate Expressions for Knowledge Base Question Answering
Semantic parsers convert natural language to logical forms, which can be evaluated on knowledge bases (KBs) to produce denotations. Recent semantic parsers have been developed with sequence-to-sequence (seq2seq) pre-trained language models (PLMs) or large language models, where the models treat logical forms as sequences of tokens. For syntactic and semantic validity, the semantic parsers use grammars that enable constrained decoding. However, the grammars lack the ability to utilize large information of KBs, although logical forms contain representations of KB elements, such as entities or relations. In this work, we propose a grammar augmented with candidate expressions for semantic parsing on a large KB with a seq2seq PLM. The grammar defines actions as production rules, and our semantic parser predicts actions during inference under the constraints by types and candidate expressions. We apply the grammar to knowledge base question answering, where the constraints by candidate expressions assist a semantic parser to generate valid KB elements. In experiments on two benchmarks, KQA Pro and Overnight, the constraints by candidate expressions increased the accuracy of our semantic parser, whether it was trained with strong supervision or weak supervision. Our semantic parser achieved state-of-the-art accuracies on KQA Pro and Overnight, and its implementation is publicly available at https://github.com/daehwannam/candexpr-sp.git.
♻ ☆ LLM-Based Agent Society Investigation: Collaboration and Confrontation in Avalon Gameplay EMNLP 2024
This paper explores the open research problem of understanding the social behaviors of LLM-based agents. Using Avalon as a testbed, we employ system prompts to guide LLM agents in gameplay. While previous studies have touched on gameplay with LLM agents, research on their social behaviors is lacking. We propose a novel framework, tailored for Avalon, features a multi-agent system facilitating efficient communication and interaction. We evaluate its performance based on game success and analyze LLM agents' social behaviors. Results affirm the framework's effectiveness in creating adaptive agents and suggest LLM-based agents' potential in navigating dynamic social interactions. By examining collaboration and confrontation behaviors, we offer insights into this field's research and applications. Our code is publicly available at https://github.com/3DAgentWorld/LLM-Game-Agent.
comment: Accepted by EMNLP 2024
♻ ☆ A Primer on the Inner Workings of Transformer-based Language Models
The rapid progress of research aimed at interpreting the inner workings of advanced language models has highlighted a need for contextualizing the insights gained from years of work in this area. This primer provides a concise technical introduction to the current techniques used to interpret the inner workings of Transformer-based language models, focusing on the generative decoder-only architecture. We conclude by presenting a comprehensive overview of the known internal mechanisms implemented by these models, uncovering connections across popular approaches and active research directions in this area.
♻ ☆ BitDelta: Your Fine-Tune May Only Be Worth One Bit NeurIPS 2024
Large Language Models (LLMs) are typically trained in two phases: pre-training on large internet-scale datasets, and fine-tuning for downstream tasks. Given the higher computational demand of pre-training, it's intuitive to assume that fine-tuning adds less new information to the model, and is thus more compressible. We explore this assumption by decomposing the weights of fine-tuned models into their pre-trained components and an additional delta. We introduce a simple method, BitDelta, which successfully quantizes this delta down to 1 bit without compromising performance. This interesting finding not only highlights the potential redundancy of information added during fine-tuning, but also has significant implications for the multi-tenant serving and multi-tenant storage of fine-tuned models. By enabling the use of a single high-precision base model accompanied by multiple 1-bit deltas, BitDelta dramatically reduces GPU memory requirements by more than 10x, which can also be translated to enhanced generation latency in multi-tenant settings. We validate BitDelta through experiments across Llama-2 and Mistral model families, and on models up to 70B parameters, showcasing minimal performance degradation over all tested settings.
comment: NeurIPS 2024 acceptance
♻ ☆ Cross-Lingual Transfer Learning for Speech Translation
There has been increasing interest in building multilingual foundation models for NLP and speech research. This paper examines how to expand the speech translation capability of these models with restricted data. Whisper, a speech foundation model with strong performance on speech recognition and English translation, is used as the example model. Using speech-to-speech retrieval to analyse the audio representations generated by the encoder, we show that utterances from different languages are mapped to a shared semantic space. This shared embedding space can then be leveraged for zero-shot cross-lingual transfer in speech translation. By fine-tuning the Whisper decoder with only English-to-Chinese speech translation data, improved performance for translation to Chinese can be obtained for multiple languages, in addition to English. Furthermore, for languages related to those seen in training it is possible to perform speech translation, despite the model never seeing the language in training, or being able to perform transcription.
♻ ☆ Decoding Susceptibility: Modeling Misbelief to Misinformation Through a Computational Approach EMNLP 2024
Susceptibility to misinformation describes the degree of belief in unverifiable claims, a latent aspect of individuals' mental processes that is not observable. Existing susceptibility studies heavily rely on self-reported beliefs, which can be subject to bias, expensive to collect, and challenging to scale for downstream applications. To address these limitations, in this work, we propose a computational approach to model users' latent susceptibility levels. As shown in previous research, susceptibility is influenced by various factors (e.g., demographic factors, political ideology), and directly influences people's reposting behavior on social media. To represent the underlying mental process, our susceptibility modeling incorporates these factors as inputs, guided by the supervision of people's sharing behavior. Using COVID-19 as a testbed domain, our experiments demonstrate a significant alignment between the susceptibility scores estimated by our computational modeling and human judgments, confirming the effectiveness of this latent modeling approach. Furthermore, we apply our model to annotate susceptibility scores on a large-scale dataset and analyze the relationships between susceptibility with various factors. Our analysis reveals that political leanings and psychological factors exhibit varying degrees of association with susceptibility to COVID-19 misinformation.
comment: EMNLP 2024
♻ ☆ Efficient Stagewise Pretraining via Progressive Subnetworks
Recent developments in large language models have sparked interest in efficient pretraining methods. Stagewise training approaches to improve efficiency, like gradual stacking and layer dropping (Reddi et al, 2023; Zhang & He, 2020), have recently garnered attention. The prevailing view suggests that stagewise dropping strategies, such as layer dropping, are ineffective, especially when compared to stacking-based approaches. This paper challenges this notion by demonstrating that, with proper design, dropping strategies can be competitive, if not better, than stacking methods. Specifically, we develop a principled stagewise training framework, progressive subnetwork training, which only trains subnetworks within the model and progressively increases the size of subnetworks during training, until it trains the full network. We propose an instantiation of this framework - Random Part Training (RAPTR) - that selects and trains only a random subnetwork (e.g. depth-wise, width-wise) of the network at each step, progressively increasing the size in stages. We show that this approach not only generalizes prior works like layer dropping but also fixes their key issues. Furthermore, we establish a theoretical basis for such approaches and provide justification for (a) increasing complexity of subnetworks in stages, conceptually diverging from prior works on layer dropping, and (b) stability in loss across stage transitions in presence of key modern architecture components like residual connections and layer norms. Through comprehensive experiments, we demonstrate that RAPTR can significantly speed up training of standard benchmarks like BERT and UL2, up to 33% compared to standard training and, surprisingly, also shows better downstream performance on UL2, improving QA tasks and SuperGLUE by 1.5%; thereby, providing evidence of better inductive bias.
♻ ☆ Spotting LLMs With Binoculars: Zero-Shot Detection of Machine-Generated Text
Detecting text generated by modern large language models is thought to be hard, as both LLMs and humans can exhibit a wide range of complex behaviors. However, we find that a score based on contrasting two closely related language models is highly accurate at separating human-generated and machine-generated text. Based on this mechanism, we propose a novel LLM detector that only requires simple calculations using a pair of pre-trained LLMs. The method, called Binoculars, achieves state-of-the-art accuracy without any training data. It is capable of spotting machine text from a range of modern LLMs without any model-specific modifications. We comprehensively evaluate Binoculars on a number of text sources and in varied situations. Over a wide range of document types, Binoculars detects over 90% of generated samples from ChatGPT (and other LLMs) at a false positive rate of 0.01%, despite not being trained on any ChatGPT data.
comment: 20 pages, code available at https://github.com/ahans30/Binoculars
♻ ☆ Differential contributions of machine learning and statistical analysis to language and cognitive sciences
Data-driven approaches have revolutionized scientific research, with machine learning and statistical analysis being commonly used methodologies. Despite their widespread use, these approaches differ significantly in their techniques, objectives and implementations. Few studies have systematically applied both methods to identical datasets to highlight potential differences, particularly in language and cognitive sciences. This study employs the Buckeye Speech Corpus to illustrate how machine learning and statistical analysis are applied in data-driven research to obtain distinct insights on language production. We demonstrate the theoretical differences, implementation steps, and unique objectives of each approach through a comprehensive, tutorial-like comparison. Our analysis reveals that while machine learning excels at pattern recognition and prediction, statistical methods provide deeper insights into relationships between variables. The study highlights how semantic relevance, a novel metric measuring contextual influence on target words, contributes to understanding word duration in speech. We also systematically compare the differences between regression models used in machine learning and statistical analysis, particularly focusing on the training and fitting processes. Additionally, we clarify several common misconceptions that contribute to the confusion between these two approaches. Overall, by elucidating the complementary strengths of machine learning and statistics, this research enhances our understanding of diverse data-driven strategies in language and cognitive sciences, offering researchers valuable guidance on when and how to effectively apply these approaches in different research contexts.
♻ ☆ Towards Robust and Cost-Efficient Knowledge Unlearning for Large Language Models
Large Language Models (LLMs) have demonstrated strong reasoning and memorization capabilities via pretraining on massive textual corpora. However, this poses risk of privacy and copyright violations, highlighting the need for efficient machine unlearning methods that remove sensitive data without retraining from scratch. While Gradient Ascent (GA) is commonly used to unlearn by reducing the likelihood of generating unwanted content, it leads to unstable optimization and catastrophic forgetting of retrained knowledge. We also find that combining GA with low-rank adaptation results in poor trade-offs between computational cost and generative performance. To address these challenges, we propose two novel techniques for robust and efficient unlearning for LLMs. First, we introduce Inverted Hinge loss, which suppresses unwanted tokens while maintaining fluency by boosting the probability of the next most likely token. Second, we develop a data-adaptive initialization for LoRA adapters via low-rank approximation weighted with relative Fisher information, thereby focusing updates on parameters critical for removing targeted knowledge. Experiments on the Training Data Extraction Challenge dataset using GPT-Neo models as well as on the TOFU benchmark with Phi-1.5B and Llama2-7B models demonstrate that our approach effectively removes sensitive information while maintaining reasoning and generative capabilities with minimal impact.
comment: Preprint
♻ ☆ Can pre-trained language models generate titles for research papers?
The title of a research paper communicates in a succinct style the main theme and, sometimes, the findings of the paper. Coming up with the right title is often an arduous task, and therefore, it would be beneficial to authors if title generation can be automated. In this paper, we fine-tune pre-trained language models to generate titles of papers from their abstracts. Additionally, we use GPT-3.5-turbo in a zero-shot setting to generate paper titles. The performance of the models is measured with ROUGE, METEOR, MoverScore, BERTScore and SciBERTScore metrics. We find that fine-tuned PEGASUS-large outperforms the other models, including fine-tuned LLaMA-3-8B and GPT-3.5-turbo, across most metrics. We also demonstrate that ChatGPT can generate creative titles for papers. Our observations suggest that AI-generated paper titles are generally accurate and appropriate.
♻ ☆ LLM Roleplay: Simulating Human-Chatbot Interaction
The development of chatbots requires collecting a large number of human-chatbot dialogues to reflect the breadth of users' sociodemographic backgrounds and conversational goals. However, the resource requirements to conduct the respective user studies can be prohibitively high and often only allow for a narrow analysis of specific dialogue goals and participant demographics. In this paper, we propose LLM Roleplay: a goal-oriented, persona-based method to automatically generate diverse multi-turn dialogues simulating human-chatbot interaction. LLM Roleplay can be applied to generate dialogues with any type of chatbot and uses large language models (LLMs) to play the role of textually described personas. To validate our method, we collect natural human-chatbot dialogues from different sociodemographic groups and conduct a user study to compare these with our generated dialogues. We evaluate the capabilities of state-of-the-art LLMs in maintaining a conversation during their embodiment of a specific persona and find that our method can simulate human-chatbot dialogues with a high indistinguishability rate.
♻ ☆ Extreme Miscalibration and the Illusion of Adversarial Robustness
Deep learning-based Natural Language Processing (NLP) models are vulnerable to adversarial attacks, where small perturbations can cause a model to misclassify. Adversarial Training (AT) is often used to increase model robustness. However, we have discovered an intriguing phenomenon: deliberately or accidentally miscalibrating models masks gradients in a way that interferes with adversarial attack search methods, giving rise to an apparent increase in robustness. We show that this observed gain in robustness is an illusion of robustness (IOR), and demonstrate how an adversary can perform various forms of test-time temperature calibration to nullify the aforementioned interference and allow the adversarial attack to find adversarial examples. Hence, we urge the NLP community to incorporate test-time temperature scaling into their robustness evaluations to ensure that any observed gains are genuine. Finally, we show how the temperature can be scaled during \textit{training} to improve genuine robustness.
♻ ☆ ColBERT Retrieval and Ensemble Response Scoring for Language Model Question Answering
Domain-specific question answering remains challenging for language models, given the deep technical knowledge required to answer questions correctly. This difficulty is amplified for smaller language models that cannot encode as much information in their parameters as larger models. The "Specializing Large Language Models for Telecom Networks" challenge aimed to enhance the performance of two small language models, Phi-2 and Falcon-7B in telecommunication question answering. In this paper, we present our question answering systems for this challenge. Our solutions achieved leading marks of 81.9% accuracy for Phi-2 and 57.3% for Falcon-7B. We have publicly released our code and fine-tuned models.
comment: 7 pages, 2 figures, and 8 tables. This paper has been accepted at the 2024 IEEE Global Communications (GLOBECOM) Workshops
♻ ☆ Self-Powered LLM Modality Expansion for Large Speech-Text Models EMNLP 2024
Large language models (LLMs) exhibit remarkable performance across diverse tasks, indicating their potential for expansion into large speech-text models (LSMs) by integrating speech capabilities. Although unified speech-text pre-training and multimodal data instruction-tuning offer considerable benefits, these methods generally entail significant resource demands and tend to overfit specific tasks. This study aims to refine the use of speech datasets for LSM training by addressing the limitations of vanilla instruction tuning. We explore the instruction-following dynamics within LSMs, identifying a critical issue termed speech anchor bias-a tendency for LSMs to over-rely on speech inputs, mistakenly interpreting the entire speech modality as directives, thereby neglecting textual instructions. To counteract this bias, we introduce a self-powered LSM that leverages augmented automatic speech recognition data generated by the model itself for more effective instruction tuning. Our experiments across a range of speech-based tasks demonstrate that self-powered LSM mitigates speech anchor bias and improves the fusion of speech and text modalities in LSMs. Data, code and scripts are freely available at https://github.com/ytf-philp/Self-powered-LSM.
comment: Accepted to EMNLP 2024
♻ ☆ Less for More: Enhanced Feedback-aligned Mixed LLMs for Molecule Caption Generation and Fine-Grained NLI Evaluation
Scientific language models drive research innovation but require extensive fine-tuning on large datasets. This work enhances such models by improving their inference and evaluation capabilities with minimal or no additional training. Focusing on molecule caption generation, we explore synergies between alignment fine-tuning and model merging in a cross-modal setup. We reveal intriguing insights into the behaviour and suitability of such methods while significantly surpassing state-of-the-art models. Moreover, we propose a novel atomic-level evaluation method leveraging off-the-shelf Natural Language Inference (NLI) models for use in the unseen chemical domain. Our experiments demonstrate that our evaluation operates at the right level of granularity, effectively handling multiple content units and subsentence reasoning, while widely adopted NLI methods consistently misalign with assessment criteria.
♻ ☆ Problem-Solving Guide: Predicting the Algorithm Tags and Difficulty for Competitive Programming Problems
The recent program development industries have required problem-solving abilities for engineers, especially application developers. However, AI-based education systems to help solve computer algorithm problems have not yet attracted attention, while most big tech companies require the ability to solve algorithm problems including Google, Meta, and Amazon. The most useful guide to solving algorithm problems might be guessing the category (tag) of the facing problems. Therefore, our study addresses the task of predicting the algorithm tag as a useful tool for engineers and developers. Moreover, we also consider predicting the difficulty levels of algorithm problems, which can be used as useful guidance to calculate the required time to solve that problem. In this paper, we present a real-world algorithm problem multi-task dataset, AMT, by mainly collecting problem samples from the most famous and large competitive programming website Codeforces. To the best of our knowledge, our proposed dataset is the most large-scale dataset for predicting algorithm tags compared to previous studies. Moreover, our work is the first to address predicting the difficulty levels of algorithm problems. We present a deep learning-based novel method for simultaneously predicting algorithm tags and the difficulty levels of an algorithm problem given. All datasets and source codes are available at https://github.com/sronger/PSG_Predicting_Algorithm_Tags_and_Difficulty.
comment: 7 pages
♻ ☆ Towards a Benchmark for Large Language Models for Business Process Management Tasks
An increasing number of organizations are deploying Large Language Models (LLMs) for a wide range of tasks. Despite their general utility, LLMs are prone to errors, ranging from inaccuracies to hallucinations. To objectively assess the capabilities of existing LLMs, performance benchmarks are conducted. However, these benchmarks often do not translate to more specific real-world tasks. This paper addresses the gap in benchmarking LLM performance in the Business Process Management (BPM) domain. Currently, no BPM-specific benchmarks exist, creating uncertainty about the suitability of different LLMs for BPM tasks. This paper systematically compares LLM performance on four BPM tasks focusing on small open-source models. The analysis aims to identify task-specific performance variations, compare the effectiveness of open-source versus commercial models, and assess the impact of model size on BPM task performance. This paper provides insights into the practical applications of LLMs in BPM, guiding organizations in selecting appropriate models for their specific needs.
comment: Submitted to HICSS (June 15, 2024)
♻ ☆ Turning Up the Heat: Min-p Sampling for Creative and Coherent LLM Outputs
Large Language Models (LLMs) generate text by sampling the next token from a probability distribution over the vocabulary at each decoding step. However, popular sampling methods like top-p (nucleus sampling) often struggle to balance quality and diversity, especially at higher temperatures, leading to incoherent or repetitive outputs. To address this challenge, we propose min-p sampling, a dynamic truncation method that adjusts the sampling threshold based on the model's confidence by scaling according to the top token's probability. We conduct extensive experiments on benchmarks including GPQA, GSM8K, and AlpacaEval Creative Writing, demonstrating that min-p sampling improves both the quality and diversity of generated text, particularly at high temperatures. Moreover, human evaluations reveal a clear preference for min-p sampling in terms of both text quality and diversity. Min-p sampling has been adopted by multiple open-source LLM implementations, highlighting its practical utility and potential impact.
comment: 20 Pages, revised from 8 pages initially. Main additions include: General full rewrite/reformatting, more comparisons with other sampling methods (eta, epsilon, top-k) on 7B parameter models, more benchmarks for >70B parameter models, human evaluation, theoretical explanations, ethics statement, reproducibility and acknowledgements
♻ ☆ Conditional and Modal Reasoning in Large Language Models EMNLP 2024
The reasoning abilities of large language models (LLMs) are the topic of a growing body of research in AI and cognitive science. In this paper, we probe the extent to which twenty-nine LLMs are able to distinguish logically correct inferences from logically fallacious ones. We focus on inference patterns involving conditionals (e.g., 'If Ann has a queen, then Bob has a jack') and epistemic modals (e.g., 'Ann might have an ace', 'Bob must have a king'). These inferences have been of special interest to logicians, philosophers, and linguists, since they play a central role in the fundamental human ability to reason about distal possibilities. Assessing LLMs on these inferences is thus highly relevant to the question of how much the reasoning abilities of LLMs match those of humans. All the LLMs we tested make some basic mistakes with conditionals or modals, though zero-shot chain-of-thought prompting helps them make fewer mistakes. Even the best performing LLMs make basic errors in modal reasoning, display logically inconsistent judgments across inference patterns involving epistemic modals and conditionals, and give answers about complex conditional inferences that do not match reported human judgments. These results highlight gaps in basic logical reasoning in today's LLMs.
comment: Accepted for The 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024). Final version includes additional models and additional inference patterns
♻ ☆ Beyond Scalar Reward Model: Learning Generative Judge from Preference Data
Learning from preference feedback is a common practice for aligning large language models~(LLMs) with human value. Conventionally, preference data is learned and encoded into a scalar reward model that connects a value head with an LLM to produce a scalar score as preference or reward. However, scalar models lack interpretability and are known to be susceptible to biases in datasets. This paper investigates leveraging the generation capability of LLMs to address both limitations in one shot. Specifically, we prompt the pre-trained LLM to generate positive and negative judgments, both supported with rationales in natural language form. The self-generated contrastive judgment pairs are used to train the generative judge with Direct Preference Optimization (DPO). This proposal of training the generative Judge using self-generated Contrastive judgments (Con-J) ensures natural interpretability due to the generated rationales together with the judgments, as well as high robustness against bias without the need for an additional reward head. Experimental results show that the performance of Con-J is comparable to the scalar reward model trained on the same collection of preference data, and demonstrate its superior interpretability and robustness in encoding human preferences.
♻ ☆ "In Dialogues We Learn": Towards Personalized Dialogue Without Pre-defined Profiles through In-Dialogue Learning EMNLP 2024
Personalized dialogue systems have gained significant attention in recent years for their ability to generate responses in alignment with different personas. However, most existing approaches rely on pre-defined personal profiles, which are not only time-consuming and labor-intensive to create but also lack flexibility. We propose In-Dialogue Learning (IDL), a fine-tuning framework that enhances the ability of pre-trained large language models to leverage dialogue history to characterize persona for completing personalized dialogue generation tasks without pre-defined profiles. Our experiments on three datasets demonstrate that IDL brings substantial improvements, with BLEU and ROUGE scores increasing by up to 200% and 247%, respectively. Additionally, the results of human evaluations further validate the efficacy of our proposed method.
comment: Accepted by EMNLP 2024
♻ ☆ A Novel Metric for Measuring the Robustness of Large Language Models in Non-adversarial Scenarios EMNLP
We evaluate the robustness of several large language models on multiple datasets. Robustness here refers to the relative insensitivity of the model's answers to meaning-preserving variants of their input. Benchmark datasets are constructed by introducing naturally-occurring, non-malicious perturbations, or by generating semantically equivalent paraphrases of input questions or statements. We further propose a novel metric for assessing a model robustness, and demonstrate its benefits in the non-adversarial scenario by empirical evaluation of several models on the created datasets.
comment: Published in the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP) findings
♻ ☆ How Far Can We Extract Diverse Perspectives from Large Language Models? EMNLP 2024
Collecting diverse human opinions is costly and challenging. This leads to a recent trend in exploiting large language models (LLMs) for generating diverse data for potential scalable and efficient solutions. However, the extent to which LLMs can generate diverse perspectives on subjective topics is still unclear. In this study, we explore LLMs' capacity of generating diverse perspectives and rationales on subjective topics such as social norms and argumentative texts. We introduce the problem of extracting maximum diversity from LLMs. Motivated by how humans form opinions based on values, we propose a criteria-based prompting technique to ground diverse opinions. To see how far we can extract diverse perspectives from LLMs, or called diversity coverage, we employ a step-by-step recall prompting to generate more outputs from the model iteratively. Our methods, applied to various tasks, show that LLMs can indeed produce diverse opinions according to the degree of task subjectivity. We also find that LLM's performance of extracting maximum diversity is on par with human.
comment: Accepted at EMNLP 2024 Main Conference
♻ ☆ Instruct-Tuning Pretrained Causal Language Models for Ancient Greek Papyrology and Epigraphy
This paper describes an experiment in fine-tuning a pretrained causal language model (Meta's Llama 3.1 8B Instruct) to support three key tasks in philological research: dating, geographic attribution, and restoring missing or illegible characters in ancient Greek inscriptions and documentary papyri. Using an instruction-based approach and a 95%/5% train/test split, the models for papyri achieved a character error rate (CER) of 14.9%, a top-1 accuracy of 73.5%, and a top-20 accuracy of 86.0% in text reconstruction on the test set. For geographic attribution, they achieved a top-1 accuracy of 66.4% and a top-3 accuracy of 79.9%. In chronological attribution, the models showed an average deviation of 21.7 years from the actual terminus post/ante quem, with a median deviation of 0 years. For inscriptions, the models achieved a CER of 20.5%, a top-1 accuracy of 63.7%, and a top-20 accuracy of 83.0% for sequences up to 10 characters. In geographic attribution, they reached a top-1 accuracy of 75.0% and a top-3 accuracy of 83.7%. For dating, they had an average deviation of 37.1 years and a median deviation of 3 years from the actual date. When benchmarked against the state-of-the-art model (Ithaca) on a shared test set and recently edited inscriptions, the instruction-tuned models excelled in text restoration, with the added benefit of ignoring spaces during reconstruction to align with the continuous script of ancient texts. However, the models performed lower than Ithaca in geographic and chronological attribution. These preliminary results suggest that fine-tuning larger pretrained causal language models with instruction templates holds promise for philological research, especially in textual criticism.
comment: 9 pages, 1 table. To be submitted
♻ ☆ J2N -- Nominal Adjective Identification and its Application
This paper explores the challenges posed by nominal adjectives (NAs) in natural language processing (NLP) tasks, particularly in part-of-speech (POS) tagging. We propose treating NAs as a distinct POS tag, "JN," and investigate its impact on POS tagging, BIO chunking, and coreference resolution. Our study shows that reclassifying NAs can improve the accuracy of syntactic analysis and structural understanding in NLP. We present experimental results using Hidden Markov Models (HMMs), Maximum Entropy (MaxEnt) models, and Spacy, demonstrating the feasibility and potential benefits of this approach. Additionally we finetuned a bert model to identify the NA in untagged text.
comment: 7 pages, 4 figures
♻ ☆ Weak-to-Strong Backdoor Attack for Large Language Models
Despite being widely applied due to their exceptional capabilities, Large Language Models (LLMs) have been proven to be vulnerable to backdoor attacks. These attacks introduce targeted vulnerabilities into LLMs by poisoning training samples and full-parameter fine-tuning. However, this kind of backdoor attack is limited since they require significant computational resources, especially as the size of LLMs increases. Besides, parameter-efficient fine-tuning (PEFT) offers an alternative but the restricted parameter updating may impede the alignment of triggers with target labels. In this study, we first verify that backdoor attacks with PEFT may encounter challenges in achieving feasible performance. To address these issues and improve the effectiveness of backdoor attacks with PEFT, we propose a novel backdoor attack algorithm from weak to strong based on feature alignment-enhanced knowledge distillation (W2SAttack). Specifically, we poison small-scale language models through full-parameter fine-tuning to serve as the teacher model. The teacher model then covertly transfers the backdoor to the large-scale student model through feature alignment-enhanced knowledge distillation, which employs PEFT. Theoretical analysis reveals that W2SAttack has the potential to augment the effectiveness of backdoor attacks. We demonstrate the superior performance of W2SAttack on classification tasks across four language models, four backdoor attack algorithms, and two different architectures of teacher models. Experimental results indicate success rates close to 100% for backdoor attacks targeting PEFT.
♻ ☆ Streamlining Redundant Layers to Compress Large Language Models
This paper introduces LLM-Streamline, a pioneer work on layer pruning for large language models (LLMs). It is based on the observation that different layers have varying impacts on hidden states, enabling the identification of less important layers to be pruned.LLM-Streamline comprises two parts: layer pruning, which removes consecutive layers with the lowest importance based on target sparsity, and layer replacement, a novel module that trains a lightweight network to replace the pruned layers to mitigate performance loss. Additionally, a new metric called stability is proposed to address the limitations of the widely used accuracy metric in evaluating model compression. Experiments show that LLM-Streamline outperforms both previous and concurrent state-of-the-art pruning methods in terms of both performance and training efficiency.
♻ ☆ Improving Academic Skills Assessment with NLP and Ensemble Learning
This study addresses the critical challenges of assessing foundational academic skills by leveraging advancements in natural language processing (NLP). Traditional assessment methods often struggle to provide timely and comprehensive feedback on key cognitive and linguistic aspects, such as coherence, syntax, and analytical reasoning. Our approach integrates multiple state-of-the-art NLP models, including BERT, RoBERTa, BART, DeBERTa, and T5, within an ensemble learning framework. These models are combined through stacking techniques using LightGBM and Ridge regression to enhance predictive accuracy. The methodology involves detailed data preprocessing, feature extraction, and pseudo-label learning to optimize model performance. By incorporating sophisticated NLP techniques and ensemble learning, this study significantly improves the accuracy and efficiency of assessments, offering a robust solution that surpasses traditional methods and opens new avenues for educational technology research focused on enhancing core academic competencies.
comment: 5 pages, 2 figures
Computer Vision and Pattern Recognition 19
☆ ChangeMinds: Multi-task Framework for Detecting and Describing Changes in Remote Sensing
Recent advancements in Remote Sensing (RS) for Change Detection (CD) and Change Captioning (CC) have seen substantial success by adopting deep learning techniques. Despite these advances, existing methods often handle CD and CC tasks independently, leading to inefficiencies from the absence of synergistic processing. In this paper, we present ChangeMinds, a novel unified multi-task framework that concurrently optimizes CD and CC processes within a single, end-to-end model. We propose the change-aware long short-term memory module (ChangeLSTM) to effectively capture complex spatiotemporal dynamics from extracted bi-temporal deep features, enabling the generation of universal change-aware representations that effectively serve both CC and CD tasks. Furthermore, we introduce a multi-task predictor with a cross-attention mechanism that enhances the interaction between image and text features, promoting efficient simultaneous learning and processing for both tasks. Extensive evaluations on the LEVIR-MCI dataset, alongside other standard benchmarks, show that ChangeMinds surpasses existing methods in multi-task learning settings and markedly improves performance in individual CD and CC tasks. Codes and pre-trained models will be available online.
☆ GALA: Geometry-Aware Local Adaptive Grids for Detailed 3D Generation
We propose GALA, a novel representation of 3D shapes that (i) excels at capturing and reproducing complex geometry and surface details, (ii) is computationally efficient, and (iii) lends itself to 3D generative modelling with modern, diffusion-based schemes. The key idea of GALA is to exploit both the global sparsity of surfaces within a 3D volume and their local surface properties. Sparsity is promoted by covering only the 3D object boundaries, not empty space, with an ensemble of tree root voxels. Each voxel contains an octree to further limit storage and compute to regions that contain surfaces. Adaptivity is achieved by fitting one local and geometry-aware coordinate frame in each non-empty leaf node. Adjusting the orientation of the local grid, as well as the anisotropic scales of its axes, to the local surface shape greatly increases the amount of detail that can be stored in a given amount of memory, which in turn allows for quantization without loss of quality. With our optimized C++/CUDA implementation, GALA can be fitted to an object in less than 10 seconds. Moreover, the representation can efficiently be flattened and manipulated with transformer networks. We provide a cascaded generation pipeline capable of generating 3D shapes with great geometric detail.
☆ TULIP: Token-length Upgraded CLIP
We address the challenge of representing long captions in vision-language models, such as CLIP. By design these models are limited by fixed, absolute positional encodings, restricting inputs to a maximum of 77 tokens and hindering performance on tasks requiring longer descriptions. Although recent work has attempted to overcome this limit, their proposed approaches struggle to model token relationships over longer distances and simply extend to a fixed new token length. Instead, we propose a generalizable method, named TULIP, able to upgrade the token length to any length for CLIP-like models. We do so by improving the architecture with relative position encodings, followed by a training procedure that (i) distills the original CLIP text encoder into an encoder with relative position encodings and (ii) enhances the model for aligning longer captions with images. By effectively encoding captions longer than the default 77 tokens, our model outperforms baselines on cross-modal tasks such as retrieval and text-to-image generation.
☆ REPeat: A Real2Sim2Real Approach for Pre-acquisition of Soft Food Items in Robot-assisted Feeding
The paper presents REPeat, a Real2Sim2Real framework designed to enhance bite acquisition in robot-assisted feeding for soft foods. It uses `pre-acquisition actions' such as pushing, cutting, and flipping to improve the success rate of bite acquisition actions such as skewering, scooping, and twirling. If the data-driven model predicts low success for direct bite acquisition, the system initiates a Real2Sim phase, reconstructing the food's geometry in a simulation. The robot explores various pre-acquisition actions in the simulation, then a Sim2Real step renders a photorealistic image to reassess success rates. If the success improves, the robot applies the action in reality. We evaluate the system on 15 diverse plates with 10 types of food items for a soft food diet, showing improvement in bite acquisition success rates by 27\% on average across all plates. See our project website at https://emprise.cs.cornell.edu/repeat.
☆ NARAIM: Native Aspect Ratio Autoregressive Image Models NeurIPS
While vision transformers are able to solve a wide variety of computer vision tasks, no pre-training method has yet demonstrated the same scaling laws as observed in language models. Autoregressive models show promising results, but are commonly trained on images that are cropped or transformed into square images, which distorts or destroys information present in the input. To overcome this limitation, we propose NARAIM, a vision model pre-trained with an autoregressive objective that uses images in their native aspect ratio. By maintaining the native aspect ratio, we preserve the original spatial context, thereby enhancing the model's ability to interpret visual information. In our experiments, we show that maintaining the aspect ratio improves performance on a downstream classification task.
comment: Accepted to NeurIPS, see https://openreview.net/forum?id=7Iuh8VWU66
☆ InterMask: 3D Human Interaction Generation via Collaborative Masked Modelling
Generating realistic 3D human-human interactions from textual descriptions remains a challenging task. Existing approaches, typically based on diffusion models, often generate unnatural and unrealistic results. In this work, we introduce InterMask, a novel framework for generating human interactions using collaborative masked modeling in discrete space. InterMask first employs a VQ-VAE to transform each motion sequence into a 2D discrete motion token map. Unlike traditional 1D VQ token maps, it better preserves fine-grained spatio-temporal details and promotes spatial awareness within each token. Building on this representation, InterMask utilizes a generative masked modeling framework to collaboratively model the tokens of two interacting individuals. This is achieved by employing a transformer architecture specifically designed to capture complex spatio-temporal interdependencies. During training, it randomly masks the motion tokens of both individuals and learns to predict them. In inference, starting from fully masked sequences, it progressively fills in the tokens for both individuals. With its enhanced motion representation, dedicated architecture, and effective learning strategy, InterMask achieves state-of-the-art results, producing high-fidelity and diverse human interactions. It outperforms previous methods, achieving an FID of $5.154$ (vs $5.535$ for in2IN) on the InterHuman dataset and $0.399$ (vs $5.207$ for InterGen) on the InterX dataset. Additionally, InterMask seamlessly supports reaction generation without the need for model redesign or fine-tuning.
comment: Project webpage: https://gohar-malik.github.io/intermask
☆ Leveraging Customer Feedback for Multi-modal Insight Extraction NAACL 2024
Businesses can benefit from customer feedback in different modalities, such as text and images, to enhance their products and services. However, it is difficult to extract actionable and relevant pairs of text segments and images from customer feedback in a single pass. In this paper, we propose a novel multi-modal method that fuses image and text information in a latent space and decodes it to extract the relevant feedback segments using an image-text grounded text decoder. We also introduce a weakly-supervised data generation technique that produces training data for this task. We evaluate our model on unseen data and demonstrate that it can effectively mine actionable insights from multi-modal customer feedback, outperforming the existing baselines by $14$ points in F1 score.
comment: NAACL 2024
☆ SlimSeiz: Efficient Channel-Adaptive Seizure Prediction Using a Mamba-Enhanced Network
Epileptic seizures cause abnormal brain activity, and their unpredictability can lead to accidents, underscoring the need for long-term seizure prediction. Although seizures can be predicted by analyzing electroencephalogram (EEG) signals, existing methods often require too many electrode channels or larger models, limiting mobile usability. This paper introduces a SlimSeiz framework that utilizes adaptive channel selection with a lightweight neural network model. SlimSeiz operates in two states: the first stage selects the optimal channel set for seizure prediction using machine learning algorithms, and the second stage employs a lightweight neural network based on convolution and Mamba for prediction. On the Children's Hospital Boston-MIT (CHB-MIT) EEG dataset, SlimSeiz can reduce channels from 22 to 8 while achieving a satisfactory result of 94.8% accuracy, 95.5% sensitivity, and 94.0% specificity with only 21.2K model parameters, matching or outperforming larger models' performance. We also validate SlimSeiz on a new EEG dataset, SRH-LEI, collected from Shanghai Renji Hospital, demonstrating its effectiveness across different patients. The code and SRH-LEI dataset are available at https://github.com/guoruilu/SlimSeiz.
comment: 5 pages, 3 figures
☆ Facial Width-to-Height Ratio Does Not Predict Self-Reported Behavioral Tendencies
A growing number of studies have linked facial width-to-height ratio (fWHR) with various antisocial or violent behavioral tendencies. However, those studies have predominantly been laboratory based and low powered. This work reexamined the links between fWHR and behavioral tendencies in a large sample of 137,163 participants. Behavioral tendencies were measured using 55 well-established psychometric scales, including self-report scales measuring intelligence, domains and facets of the five-factor model of personality, impulsiveness, sense of fairness, sensational interests, self-monitoring, impression management, and satisfaction with life. The findings revealed that fWHR is not substantially linked with any of these self-reported measures of behavioral tendencies, calling into question whether the links between fWHR and behavior generalize beyond the small samples and specific experimental settings that have been used in past fWHR research.
comment: Psychological Science (2017)
☆ Optimizing Waste Management with Advanced Object Detection for Garbage Classification
Garbage production and littering are persistent global issues that pose significant environmental challenges. Despite large-scale efforts to manage waste through collection and sorting, existing approaches remain inefficient, leading to inadequate recycling and disposal. Therefore, developing advanced AI-based systems is less labor intensive approach for addressing the growing waste problem more effectively. These models can be applied to sorting systems or possibly waste collection robots that may produced in the future. AI models have grown significantly at identifying objects through object detection.This paper reviews the implementation of AI models for classifying trash through object detection, specifically focusing on the use of YOLO V5 for training and testing. The study demonstrates how YOLO V5 can effectively identify various types of waste, including \textit{plastic}, \textit{paper}, \textit{glass}, \textit{metal}, \textit{cardboard}, and \textit{biodegradables}}.
☆ Make the Pertinent Salient: Task-Relevant Reconstruction for Visual Control with Distractions
Recent advancements in Model-Based Reinforcement Learning (MBRL) have made it a powerful tool for visual control tasks. Despite improved data efficiency, it remains challenging to train MBRL agents with generalizable perception. Training in the presence of visual distractions is particularly difficult due to the high variation they introduce to representation learning. Building on DREAMER, a popular MBRL method, we propose a simple yet effective auxiliary task to facilitate representation learning in distracting environments. Under the assumption that task-relevant components of image observations are straightforward to identify with prior knowledge in a given task, we use a segmentation mask on image observations to only reconstruct task-relevant components. In doing so, we greatly reduce the complexity of representation learning by removing the need to encode task-irrelevant objects in the latent representation. Our method, Segmentation Dreamer (SD), can be used either with ground-truth masks easily accessible in simulation or by leveraging potentially imperfect segmentation foundation models. The latter is further improved by selectively applying the reconstruction loss to avoid providing misleading learning signals due to mask prediction errors. In modified DeepMind Control suite (DMC) and Meta-World tasks with added visual distractions, SD achieves significantly better sample efficiency and greater final performance than prior work. We find that SD is especially helpful in sparse reward tasks otherwise unsolvable by prior work, enabling the training of visually robust agents without the need for extensive reward engineering.
☆ Improving 3D Few-Shot Segmentation with Inference-Time Pseudo-Labeling
In recent years, few-shot segmentation (FSS) models have emerged as a promising approach in medical imaging analysis, offering remarkable adaptability to segment novel classes with limited annotated data. Existing approaches to few-shot segmentation have often overlooked the potential of the query itself, failing to fully utilize the valuable information it contains. However, treating the query as unlabeled data provides an opportunity to enhance prediction accuracy. Specifically in the domain of medical imaging, the volumetric structure of queries offers a considerable source of valuable information that can be used to improve the target slice segmentation. In this work, we present a novel strategy to efficiently leverage the intrinsic information of the query sample for final segmentation during inference. First, we use the support slices from a reference volume to generate an initial segmentation score for the query slices through a prototypical approach. Subsequently, we apply a confidence-aware pseudo-labeling procedure to transfer the most informative parts of query slices to the support set. The final prediction is performed based on the new expanded support set, enabling the prediction of a more accurate segmentation mask for the query volume. Extensive experiments show that the proposed method can effectively boost performance across diverse settings and datasets.
☆ LongHalQA: Long-Context Hallucination Evaluation for MultiModal Large Language Models
Hallucination, a phenomenon where multimodal large language models~(MLLMs) tend to generate textual responses that are plausible but unaligned with the image, has become one major hurdle in various MLLM-related applications. Several benchmarks have been created to gauge the hallucination levels of MLLMs, by either raising discriminative questions about the existence of objects or introducing LLM evaluators to score the generated text from MLLMs. However, the discriminative data largely involve simple questions that are not aligned with real-world text, while the generative data involve LLM evaluators that are computationally intensive and unstable due to their inherent randomness. We propose LongHalQA, an LLM-free hallucination benchmark that comprises 6K long and complex hallucination text. LongHalQA is featured by GPT4V-generated hallucinatory data that are well aligned with real-world scenarios, including object/image descriptions and multi-round conversations with 14/130 words and 189 words, respectively, on average. It introduces two new tasks, hallucination discrimination and hallucination completion, unifying both discriminative and generative evaluations in a single multiple-choice-question form and leading to more reliable and efficient evaluations without the need for LLM evaluators. Further, we propose an advanced pipeline that greatly facilitates the construction of future hallucination benchmarks with long and complex questions and descriptions. Extensive experiments over multiple recent MLLMs reveal various new challenges when they are handling hallucinations with long and complex textual data. Dataset and evaluation code are available at https://github.com/hanqiu-hq/LongHalQA.
♻ ☆ Fine-Tuning is Fine, if Calibrated NeurIPS 2024
Fine-tuning is arguably the most straightforward way to tailor a pre-trained model (e.g., a foundation model) to downstream applications, but it also comes with the risk of losing valuable knowledge the model had learned in pre-training. For example, fine-tuning a pre-trained classifier capable of recognizing a large number of classes to master a subset of classes at hand is shown to drastically degrade the model's accuracy in the other classes it had previously learned. As such, it is hard to further use the fine-tuned model when it encounters classes beyond the fine-tuning data. In this paper, we systematically dissect the issue, aiming to answer the fundamental question, "What has been damaged in the fine-tuned model?" To our surprise, we find that the fine-tuned model neither forgets the relationship among the other classes nor degrades the features to recognize these classes. Instead, the fine-tuned model often produces more discriminative features for these other classes, even if they were missing during fine-tuning! {What really hurts the accuracy is the discrepant logit scales between the fine-tuning classes and the other classes}, implying that a simple post-processing calibration would bring back the pre-trained model's capability and at the same time unveil the feature improvement over all classes. We conduct an extensive empirical study to demonstrate the robustness of our findings and provide preliminary explanations underlying them, suggesting new directions for future theoretical analysis. Our code is available at https://github.com/OSU-MLB/Fine-Tuning-Is-Fine-If-Calibrated.
comment: The paper has been accepted to NeurIPS 2024. The first three authors contribute equally
♻ ☆ Training Data Protection with Compositional Diffusion Models
We introduce Compartmentalized Diffusion Models (CDM), a method to train different diffusion models (or prompts) on distinct data sources and arbitrarily compose them at inference time. The individual models can be trained in isolation, at different times, and on different distributions and domains and can be later composed to achieve performance comparable to a paragon model trained on all data simultaneously. Furthermore, each model only contains information about the subset of the data it was exposed to during training, enabling several forms of training data protection. In particular, CDMs enable perfect selective forgetting and continual learning for large-scale diffusion models, allow serving customized models based on the user's access rights. Empirically the quality (FID) of the class-conditional CDMs (8-splits) is within 10% (on fine-grained vision datasets) of a monolithic model (no splits), and allows (8x) faster forgetting compared monolithic model with a maximum FID increase of 1%. When applied to text-to-image generation, CDMs improve alignment (TIFA) by 14.33% over a monolithic model trained on MSCOCO. CDMs also allow determining the importance of a subset of the data (attribution) in generating particular samples, and reduce memorization.
♻ ☆ GaitContour: Efficient Gait Recognition based on a Contour-Pose Representation WACV 2025
Gait recognition holds the promise to robustly identify subjects based on walking patterns instead of appearance information. In recent years, this field has been dominated by learning methods based on two principal input representations: dense silhouette masks or sparse pose keypoints. In this work, we propose a novel, point-based Contour-Pose representation, which compactly expresses both body shape and body parts information. We further propose a local-to-global architecture, called GaitContour, to leverage this novel representation and efficiently compute subject embedding in two stages. The first stage consists of a local transformer that extracts features from five different body regions. The second stage then aggregates the regional features to estimate a global human gait representation. Such a design significantly reduces the complexity of the attention operation and improves efficiency and performance simultaneously. Through large scale experiments, GaitContour is shown to perform significantly better than previous point-based methods, while also being significantly more efficient than silhouette-based methods. On challenging datasets with significant distractors, GaitContour can even outperform silhouette-based methods.
comment: Accepted by WACV 2025
♻ ☆ Distillation-guided Representation Learning for Unconstrained Gait Recognition
Gait recognition holds the promise of robustly identifying subjects based on walking patterns instead of appearance information. While previous approaches have performed well for curated indoor data, they tend to underperform in unconstrained situations, e.g. in outdoor, long distance scenes, etc. We propose a framework, termed GAit DEtection and Recognition (GADER), for human authentication in challenging outdoor scenarios. Specifically, GADER leverages a Double Helical Signature to detect segments that contain human movement and builds discriminative features through a novel gait recognition method, where only frames containing gait information are used. To further enhance robustness, GADER encodes viewpoint information in its architecture, and distills representation from an auxiliary RGB recognition model, which enables GADER to learn from silhouette and RGB data at training time. At test time, GADER only infers from the silhouette modality. We evaluate our method on multiple State-of-The-Arts(SoTA) gait baselines and demonstrate consistent improvements on indoor and outdoor datasets, especially with a significant 25.2% improvement on unconstrained, remote gait data.
comment: Accepted by IJCB 2024
♻ ☆ Facial recognition technology and human raters can predict political orientation from images of expressionless faces even when controlling for demographics and self-presentation
Carefully standardized facial images of 591 participants were taken in the laboratory, while controlling for self-presentation, facial expression, head orientation, and image properties. They were presented to human raters and a facial recognition algorithm: both humans (r=.21) and the algorithm (r=.22) could predict participants' scores on a political orientation scale (Cronbach's alpha=.94) decorrelated with age, gender, and ethnicity. These effects are on par with how well job interviews predict job success, or alcohol drives aggressiveness. Algorithm's predictive accuracy was even higher (r=.31) when it leveraged information on participants' age, gender, and ethnicity. Moreover, the associations between facial appearance and political orientation seem to generalize beyond our sample: The predictive model derived from standardized images (while controlling for age, gender, and ethnicity) could predict political orientation (r=.13) from naturalistic images of 3,401 politicians from the U.S., UK, and Canada. The analysis of facial features associated with political orientation revealed that conservatives tended to have larger lower faces. The predictability of political orientation from standardized images has critical implications for privacy, the regulation of facial recognition technology, and understanding the origins and consequences of political orientation.
comment: Now published at American Psychologist (2024)
♻ ☆ System Identification of Neural Systems: Going Beyond Images to Modelling Dynamics
Extensive literature has drawn comparisons between recordings of biological neurons in the brain and deep neural networks. This comparative analysis aims to advance and interpret deep neural networks and enhance our understanding of biological neural systems. However, previous works did not consider the time aspect and how the encoding of video and dynamics in deep networks relate to the biological neural systems within a large-scale comparison. Towards this end, we propose the first large-scale study focused on comparing video understanding models with respect to the visual cortex recordings using video stimuli. The study encompasses more than two million regression fits, examining image vs. video understanding, convolutional vs. transformer-based and fully vs. self-supervised models. Additionally, we propose a novel neural encoding scheme to better encode biological neural systems. We provide key insights on how video understanding models predict visual cortex responses; showing video understanding better than image understanding models, convolutional models are better in the early-mid visual cortical regions than transformer based ones except for multiscale transformers, and that two-stream models are better than single stream. Furthermore, we propose a novel neural encoding scheme that is built on top of the best performing video understanding models, while incorporating inter-intra region connectivity across the visual cortex. Our neural encoding leverages the encoded dynamics from video stimuli, through utilizing two-stream networks and multiscale transformers, while taking connectivity priors into consideration. Our results show that merging both intra and inter-region connectivity priors increases the encoding performance over each one of them standalone or no connectivity priors. It also shows the necessity for encoding dynamics to fully benefit from such connectivity priors.
comment: Updated the simulated experiments to include all models, added our new connectivity module, and added results on the new BOLD Moments Dataset (BMD)
Information Retrieval 15
☆ Leveraging Customer Feedback for Multi-modal Insight Extraction NAACL 2024
Businesses can benefit from customer feedback in different modalities, such as text and images, to enhance their products and services. However, it is difficult to extract actionable and relevant pairs of text segments and images from customer feedback in a single pass. In this paper, we propose a novel multi-modal method that fuses image and text information in a latent space and decodes it to extract the relevant feedback segments using an image-text grounded text decoder. We also introduce a weakly-supervised data generation technique that produces training data for this task. We evaluate our model on unseen data and demonstrate that it can effectively mine actionable insights from multi-modal customer feedback, outperforming the existing baselines by $14$ points in F1 score.
comment: NAACL 2024
☆ Learning to Rank for Multiple Retrieval-Augmented Models through Iterative Utility Maximization
This paper investigates the design of a unified search engine to serve multiple retrieval-augmented generation (RAG) agents, each with a distinct task, backbone large language model (LLM), and retrieval-augmentation strategy. We introduce an iterative approach where the search engine generates retrieval results for these RAG agents and gathers feedback on the quality of the retrieved documents during an offline phase. This feedback is then used to iteratively optimize the search engine using a novel expectation-maximization algorithm, with the goal of maximizing each agent's utility function. Additionally, we adapt this approach to an online setting, allowing the search engine to refine its behavior based on real-time individual agents feedback to better serve the results for each of them. Experiments on diverse datasets from the Knowledge-Intensive Language Tasks (KILT) benchmark demonstrates that our approach significantly on average outperforms competitive baselines across 18 RAG models. We also demonstrate that our method effectively ``personalizes'' the retrieval process for each RAG agent based on the collected feedback. Finally, we provide a comprehensive ablation study to explore various aspects of our method.
☆ The Role of Fake Users in Sequential Recommender Systems
Sequential Recommender Systems (SRSs) are widely used to model user behavior over time, yet their robustness remains an under-explored area of research. In this paper, we conduct an empirical study to assess how the presence of fake users, who engage in random interactions, follow popular or unpopular items, or focus on a single genre, impacts the performance of SRSs in real-world scenarios. We evaluate two SRS models across multiple datasets, using established metrics such as Normalized Discounted Cumulative Gain (NDCG) and Rank Sensitivity List (RLS) to measure performance. While traditional metrics like NDCG remain relatively stable, our findings reveal that the presence of fake users severely degrades RLS metrics, often reducing them to near-zero values. These results highlight the need for further investigation into the effects of fake users on training data and emphasize the importance of developing more resilient SRSs that can withstand different types of adversarial attacks.
comment: 10 pages, 2 figures
☆ Analysis and Design of a Personalized Recommendation System Based on a Dynamic User Interest Model
With the rapid development of the internet and the explosion of information, providing users with accurate personalized recommendations has become an important research topic. This paper designs and analyzes a personalized recommendation system based on a dynamic user interest model. The system captures user behavior data, constructs a dynamic user interest model, and combines multiple recommendation algorithms to provide personalized content to users. The research results show that this system significantly improves recommendation accuracy and user satisfaction. This paper discusses the system's architecture design, algorithm implementation, and experimental results in detail and explores future research directions.
☆ ViFi-ReID: A Two-Stream Vision-WiFi Multimodal Approach for Person Re-identification
Person re-identification(ReID), as a crucial technology in the field of security, plays a vital role in safety inspections, personnel counting, and more. Most current ReID approaches primarily extract features from images, which are easily affected by objective conditions such as clothing changes and occlusions. In addition to cameras, we leverage widely available routers as sensing devices by capturing gait information from pedestrians through the Channel State Information (CSI) in WiFi signals and contribute a multimodal dataset. We employ a two-stream network to separately process video understanding and signal analysis tasks, and conduct multi-modal fusion and contrastive learning on pedestrian video and WiFi data. Extensive experiments in real-world scenarios demonstrate that our method effectively uncovers the correlations between heterogeneous data, bridges the gap between visual and signal modalities, significantly expands the sensing range, and improves ReID accuracy across multiple sensors.
☆ A Comparative Study of PDF Parsing Tools Across Diverse Document Categories
PDF is one of the most prominent data formats, making PDF parsing crucial for information extraction and retrieval, particularly with the rise of RAG systems. While various PDF parsing tools exist, their effectiveness across different document types remains understudied, especially beyond academic papers. Our research aims to address this gap by comparing 10 popular PDF parsing tools across 6 document categories using the DocLayNet dataset. These tools include PyPDF, pdfminer.six, PyMuPDF, pdfplumber, pypdfium2, Unstructured, Tabula, Camelot, as well as the deep learning-based tools Nougat and Table Transformer(TATR). We evaluated both text extraction and table detection capabilities. For text extraction, PyMuPDF and pypdfium generally outperformed others, but all parsers struggled with Scientific and Patent documents. For these challenging categories, learning-based tools like Nougat demonstrated superior performance. In table detection, TATR excelled in the Financial, Patent, Law & Regulations, and Scientific categories. Table detection tool Camelot performed best for tender documents, while PyMuPDF performed superior in the Manual category. Our findings highlight the importance of selecting appropriate parsing tools based on document type and specific tasks, providing valuable insights for researchers and practitioners working with diverse document sources.
comment: 17 pages,11 figures, 5 tables
☆ Generating Driving Simulations via Conversation
Cyber-physical systems like autonomous vehicles are tested in simulation before deployment, using domain-specific programs for scenario specification. To aid the testing of autonomous vehicles in simulation, we design a natural language interface, using an instruction-following large language model, to assist a non-coding domain expert in synthesising the desired scenarios and vehicle behaviours. We show that using it to convert utterances to the symbolic program is feasible, despite the very small training dataset. Human experiments show that dialogue is critical to successful simulation generation, leading to a 4.5 times higher success rate than a generation without engaging in extended conversation.
comment: 6 pages, 6 figures, 2 tables
☆ ContextWIN: Whittle Index Based Mixture-of-Experts Neural Model For Restless Bandits Via Deep RL
This study introduces ContextWIN, a novel architecture that extends the Neural Whittle Index Network (NeurWIN) model to address Restless Multi-Armed Bandit (RMAB) problems with a context-aware approach. By integrating a mixture of experts within a reinforcement learning framework, ContextWIN adeptly utilizes contextual information to inform decision-making in dynamic environments, particularly in recommendation systems. A key innovation is the model's ability to assign context-specific weights to a subset of NeurWIN networks, thus enhancing the efficiency and accuracy of the Whittle index computation for each arm. The paper presents a thorough exploration of ContextWIN, from its conceptual foundation to its implementation and potential applications. We delve into the complexities of RMABs and the significance of incorporating context, highlighting how ContextWIN effectively harnesses these elements. The convergence of both the NeurWIN and ContextWIN models is rigorously proven, ensuring theoretical robustness. This work lays the groundwork for future advancements in applying contextual information to complex decision-making scenarios, recognizing the need for comprehensive dataset exploration and environment development for full potential realization.
☆ ChartKG: A Knowledge-Graph-Based Representation for Chart Images
Chart images, such as bar charts, pie charts, and line charts, are explosively produced due to the wide usage of data visualizations. Accordingly, knowledge mining from chart images is becoming increasingly important, which can benefit downstream tasks like chart retrieval and knowledge graph completion. However, existing methods for chart knowledge mining mainly focus on converting chart images into raw data and often ignore their visual encodings and semantic meanings, which can result in information loss for many downstream tasks. In this paper, we propose ChartKG, a novel knowledge graph (KG) based representation for chart images, which can model the visual elements in a chart image and semantic relations among them including visual encodings and visual insights in a unified manner. Further, we develop a general framework to convert chart images to the proposed KG-based representation. It integrates a series of image processing techniques to identify visual elements and relations, e.g., CNNs to classify charts, yolov5 and optical character recognition to parse charts, and rule-based methods to construct graphs. We present four cases to illustrate how our knowledge-graph-based representation can model the detailed visual elements and semantic relations in charts, and further demonstrate how our approach can benefit downstream applications such as semantic-aware chart retrieval and chart question answering. We also conduct quantitative evaluations to assess the two fundamental building blocks of our chart-to-KG framework, i.e., object recognition and optical character recognition. The results provide support for the usefulness and effectiveness of ChartKG.
☆ Agentic Information Retrieval
What will information entry look like in the next generation of digital products? Since the 1970s, user access to relevant information has relied on domain-specific architectures of information retrieval (IR). Over the past two decades, the advent of modern IR systems, including web search engines and personalized recommender systems, has greatly improved the efficiency of retrieving relevant information from vast data corpora. However, the core paradigm of these IR systems remains largely unchanged, relying on filtering a predefined set of candidate items. Since 2022, breakthroughs in large language models (LLMs) have begun transforming how information is accessed, establishing a new technical paradigm. In this position paper, we introduce Agentic Information Retrieval (Agentic IR), a novel IR paradigm shaped by the capabilities of LLM agents. Agentic IR expands the scope of accessible tasks and leverages a suite of new techniques to redefine information retrieval. We discuss three types of cutting-edge applications of agentic IR and the challenges faced. We propose that agentic IR holds promise for generating innovative applications, potentially becoming a central information entry point in future digital ecosystems.
comment: 11 pages, position paper
☆ Online Digital Investigative Journalism using SociaLens
Media companies witnessed a significant transformation with the rise of the internet, bigdata, machine learning (ML) and AI. Recent emergence of large language models (LLM) have added another aspect to this transformation. Researchers believe that with the help of these technologies, investigative digital journalism will enter a new era. Using a smart set of data gathering and analysis tools, journalists will be able to create data driven contents and insights in unprecedented ways. In this paper, we introduce a versatile and autonomous investigative journalism tool, called {\em SociaLens}, for identifying and extracting query specific data from online sources, responding to probing queries and drawing conclusions entailed by large volumes of data using ML analytics fully autonomously. We envision its use in investigative journalism, law enforcement and social policy planning. The proposed system capitalizes on the integration of ML technology with LLMs and advanced bigdata search techniques. We illustrate the functionality of SociaLens using a focused case study on rape incidents in a developing country and demonstrate that journalists can gain nuanced insights without requiring coding expertise they might lack. SociaLens is designed as a ChatBot that is capable of contextual conversation, find and collect data relevant to queries, initiate ML tasks to respond to queries, generate textual and visual reports, all fully autonomously within the ChatBot environment.
♻ ☆ ColBERT Retrieval and Ensemble Response Scoring for Language Model Question Answering
Domain-specific question answering remains challenging for language models, given the deep technical knowledge required to answer questions correctly. This difficulty is amplified for smaller language models that cannot encode as much information in their parameters as larger models. The "Specializing Large Language Models for Telecom Networks" challenge aimed to enhance the performance of two small language models, Phi-2 and Falcon-7B in telecommunication question answering. In this paper, we present our question answering systems for this challenge. Our solutions achieved leading marks of 81.9% accuracy for Phi-2 and 57.3% for Falcon-7B. We have publicly released our code and fine-tuned models.
comment: 7 pages, 2 figures, and 8 tables. This paper has been accepted at the 2024 IEEE Global Communications (GLOBECOM) Workshops
♻ ☆ Evaluating D-MERIT of Partial-annotation on Information Retrieval EMNLP 2024
Retrieval models are often evaluated on partially-annotated datasets. Each query is mapped to a few relevant texts and the remaining corpus is assumed to be irrelevant. As a result, models that successfully retrieve false negatives are punished in evaluation. Unfortunately, completely annotating all texts for every query is not resource efficient. In this work, we show that using partially-annotated datasets in evaluation can paint a distorted picture. We curate D-MERIT, a passage retrieval evaluation set from Wikipedia, aspiring to contain all relevant passages for each query. Queries describe a group (e.g., "journals about linguistics") and relevant passages are evidence that entities belong to the group (e.g., a passage indicating that "Language" is a journal about linguistics). We show that evaluating on a dataset containing annotations for only a subset of the relevant passages might result in misleading ranking of the retrieval systems and that as more relevant texts are included in the evaluation set, the rankings converge. We propose our dataset as a resource for evaluation and our study as a recommendation for balance between resource-efficiency and reliable evaluation when annotating evaluation sets for text retrieval.
comment: Accepted to EMNLP 2024 main track. Our dataset can be downloaded from https://D-MERIT.github.io
♻ ☆ Fine-Grained Embedding Dimension Optimization During Training for Recommender Systems
Huge embedding tables in modern deep learning recommender models (DLRM) require prohibitively large memory during training and inference. This paper proposes FIITED, a system to automatically reduce the memory footprint via FIne-grained In-Training Embedding Dimension pruning. By leveraging the key insight that embedding vectors are not equally important, FIITED adaptively adjusts the dimension of each individual embedding vector during model training, assigning larger dimensions to more important embeddings while adapting to dynamic changes in data. We prioritize embedding dimensions with higher frequencies and gradients as more important. To enable efficient pruning of embeddings and their dimensions during model training, we propose an embedding storage system based on virtually-hashed physically-indexed hash tables. Experiments on two industry models and months of realistic datasets show that FIITED can reduce DLRM embedding size by more than 65% while preserving model quality, outperforming state-of-the-art in-training embedding pruning methods. On public datasets, FIITED can reduce the size of embedding tables by 2.1x to 800x with negligible accuracy drop, while improving model throughput.
comment: 12 pages, 15 figures
♻ ☆ EHI: End-to-end Learning of Hierarchical Index for Efficient Dense Retrieval
Dense embedding-based retrieval is widely used for semantic search and ranking. However, conventional two-stage approaches, involving contrastive embedding learning followed by approximate nearest neighbor search (ANNS), can suffer from misalignment between these stages. This mismatch degrades retrieval performance. We propose End-to-end Hierarchical Indexing (EHI), a novel method that directly addresses this issue by jointly optimizing embedding generation and ANNS structure. EHI leverages a dual encoder for embedding queries and documents while simultaneously learning an inverted file index (IVF)-style tree structure. To facilitate the effective learning of this discrete structure, EHI introduces dense path embeddings that encodes the path traversed by queries and documents within the tree. Extensive evaluations on standard benchmarks, including MS MARCO (Dev set) and TREC DL19, demonstrate EHI's superiority over traditional ANNS index. Under the same computational constraints, EHI outperforms existing state-of-the-art methods by +1.45% in MRR@10 on MS MARCO (Dev) and +8.2% in nDCG@10 on TREC DL19, highlighting the benefits of our end-to-end approach.
Multimedia 3
☆ Towards Reproducible Learning-based Compression SP 2024
A deep learning system typically suffers from a lack of reproducibility that is partially rooted in hardware or software implementation details. The irreproducibility leads to skepticism in deep learning technologies and it can hinder them from being deployed in many applications. In this work, the irreproducibility issue is analyzed where deep learning is employed in compression systems while the encoding and decoding may be run on devices from different manufacturers. The decoding process can even crash due to a single bit difference, e.g., in a learning-based entropy coder. For a given deep learning-based module with limited resources for protection, we first suggest that reproducibility can only be assured when the mismatches are bounded. Then a safeguarding mechanism is proposed to tackle the challenges. The proposed method may be applied for different levels of protection either at the reconstruction level or at a selected decoding level. Furthermore, the overhead introduced for the protection can be scaled down accordingly when the error bound is being suppressed. Experiments demonstrate the effectiveness of the proposed approach for learning-based compression systems, e.g., in image compression and point cloud compression.
comment: Accepted at MMSP 2024
♻ ☆ VidMuse: A Simple Video-to-Music Generation Framework with Long-Short-Term Modeling
In this work, we systematically study music generation conditioned solely on the video. First, we present a large-scale dataset comprising 360K video-music pairs, including various genres such as movie trailers, advertisements, and documentaries. Furthermore, we propose VidMuse, a simple framework for generating music aligned with video inputs. VidMuse stands out by producing high-fidelity music that is both acoustically and semantically aligned with the video. By incorporating local and global visual cues, VidMuse enables the creation of musically coherent audio tracks that consistently match the video content through Long-Short-Term modeling. Through extensive experiments, VidMuse outperforms existing models in terms of audio quality, diversity, and audio-visual alignment. The code and datasets will be available at https://github.com/ZeyueT/VidMuse/.
comment: The code and datasets will be available at https://github.com/ZeyueT/VidMuse/
♻ ☆ Less for More: Enhanced Feedback-aligned Mixed LLMs for Molecule Caption Generation and Fine-Grained NLI Evaluation
Scientific language models drive research innovation but require extensive fine-tuning on large datasets. This work enhances such models by improving their inference and evaluation capabilities with minimal or no additional training. Focusing on molecule caption generation, we explore synergies between alignment fine-tuning and model merging in a cross-modal setup. We reveal intriguing insights into the behaviour and suitability of such methods while significantly surpassing state-of-the-art models. Moreover, we propose a novel atomic-level evaluation method leveraging off-the-shelf Natural Language Inference (NLI) models for use in the unseen chemical domain. Our experiments demonstrate that our evaluation operates at the right level of granularity, effectively handling multiple content units and subsentence reasoning, while widely adopted NLI methods consistently misalign with assessment criteria.